WO 3 –V 2 O 5 Active Oxides for NO x SCR by NH 3 : Preparation Methods, Catalysts’ Composition, and Deactivation Mechanism—A Review

: Researchers in the ﬁeld of the selective catalytic reduction (SCR) of nitrogen oxides (NO x : NO, NO 2 , or N 2 O) by NH 3 are still greatly challenging the optimization of low-temperature activity and selectivity, high-temperature stability, resistance to alkali metals and other poisoning agents, such as Hg, As, etc. The present study reviews the research progress, related to the latest 20 years, on WO 3 –V 2 O 5 -based catalysts that are expected to overcome the catalytic performances of the current SCR catalytic devices. In details, the e ﬀ ects of the synthesis methods, chemical composition, type of supports (metal oxides, molecular sieves, and ﬁlters), doping elements, or metal oxides added as promoters of WO 3 –V 2 O 5 -based catalysts and, ﬁnally, the inﬂuence of SO 2 and H 2 O in the reaction mixture are addressed. The importance of understanding the deactivation mechanism in the presence of several poisoning agents is also emphasized, which should be taken into consideration for the design of new catalysts. V and W species do not exist independently on the titania surface (V-O-W connectivity were present on the V 2 O 5 –WO 3 / TiO 2 catalysts), which results in a higher activity compared to the corresponding titania-supported single oxides. The studies demonstrate that the activity could be enhanced by grafting V 2 O 5 –WO 3 on TiO 2 . 2.5. Positive / Negative E ﬀ ects of Di ﬀ erent Doping Agents on Catalytic Activity and Selectivity Addition or doping with other metal oxides or elements, such as F [21,58], Nb, Sb [59], Cu, Mn, and Ce [60–62] is the main way to improve catalyst activity, whereas alkaline and alkaline earth metals (Na, K, Ca, and Mg) [63] deactivate catalysts.


NO x Problem, Control Regulations, and Removal Technologies
Agriculture, industries, power generation plants, and transportation are the major sources responsible of air pollution [1]: Ozone depletion, eutrophication, smog, global warming, and acid rain have attracted wide attention from human beings [2,3]. In order to reduce air pollution, new regulations have been introduced in the latest years, such as the Directive 2010/75/EU of the European Parliament and of the Council concerning industrial activities, European standards for vehicles' exhaust gas emissions in Europe, as well as EPA (Environmental Protection Agency) tier regulations in the USA and analogous emission standards developed by the Japanese government.
Naval diesel engines use heavy fuel oil with a high sulfur content, thus most of the pollutants are nitrogen oxides (NO x ), particulate matter (PM), and sulfur oxides (SO x ) [4]. To limit NO x emissions, strict regulations and legislations have been set up by the international maritime organization (IMO) that came into force on 1 January 2016 (MARPOL Annex VI) and will be tighter in the future [5].
Many techniques have been applied for the abatement of NO x emissions, such as pre-combustion and combustion modifications [6,7], as well tail-end control equipment [2,8]. Tail-end control devices are the most efficient in the removal of NO x , including selective noncatalytic reduction Table 1. Summary of the synthesis methods of V 2 O 5 -WO 3 -based SCR catalysts.

Catalysts
Syntheses Method (Cal. Tem./ • C) The Key Facts to the Activity Ref.
V 2 O 5 -WO 3 /TiO 2 Wet impregnation (450 • C) Compared with dry-impregnated catalyst, wet-impregnated catalyst increases the intensity of Brønsted acid sites and the oxidation of the ab-NH 3 .
[24] V 2 O 5 -WO 3 /TiO 2 Dry impregnation (450 • C) V 2 O 5 -WO 3 /TiO 2 Wet impregnation (500 • C) By increasing the precursor solution acidity, more polymeric vanadium species are formed on the catalyst surface; the ratio of V 4+(3+) /V 5+ , surface acidity, and quantity of active sites are increased, whereas the activity of the catalyst is largely improved. [25] The enhanced performance of co-precipitated compared to the impregnated catalysts was associated with the co-precipitated V 2 O 5 -WO 3 /TiO 2 catalysts possess new surface: O = WO 4 (enhanced the adsorption of NH 3 ) and redox surface mono-oxo O = VO 3 sites, because of the surface defects of the TiO 2 . [26] V 2 O 5 -WO 3 /TiO 2 Incipient wetness impregnation (500 • C) V 2 O 5 -WO 3 /TiO 2 Oxalic acid leaching and impregnation (600 • C) Oxalic acid leaching is an effective way to reduce the impurities and increase reducibility of the recovered catalyst. [27] Ce-V 2 O 5 -WO 3 /TiO 2 Deposition-precipitation (500 • C) The higher NO x conversion at low temperature has origin from surface Ce species and can be assigned to more active Lewis acid sites, the weakly adsorbed NO 2 and monodentate nitrate.
[28] Ce-V 2 O 5 -WO 3 /TiO 2 Impregnation method (500 • C) V 2 O 5 -WO 3 /TiO 2 Sol-gel method (400-800 • C) Well dispersed and isolated vanadium oxide species were found to be weakly active for the SCR reaction but with a high selectivity to N 2 . [29] V 2 O 5 -WO 3 /TiO 2 Grafting (300 • C) Grafting mode and metal oxides loadings were significantly affecting the strength of interaction of V 2 O 5 and WO 3 . [30] Catalysts 2019, 9, 527 5 of 30 As shown in Figure 1, Yu et al. [24] investigated the NH 3 -SCR activities of V 2 O 5 -WO 3 /TiO 2 catalysts as fresh and poisoned by potassium chloride, prepared by wet and dry impregnation methods. They found that the wet-impregnated sample, WWT(W), shows a slightly higher activity than the impregnated one (D) at around 200 • C and this was ascribed to the formation of agglomerated V species which increases the intensity of Brønsted acid sites and the oxidation of adsorbed NH 3, while isolated V species were predominant on the dry-impregnated catalyst. After the addition of KCl, both catalysts deactivated due to the decreasing amount of ammonium ions coordinated on the Brønsted acid sites. Isolated vanadia species (DK) appeared to be more reactive with potassium and therefore more severely deactivated.  [24]. Copyright (2013) Elsevier.) As shown in Figure 1, Yu et al. [24] investigated the NH3-SCR activities of V2O5-WO3/TiO2 catalysts as fresh and poisoned by potassium chloride, prepared by wet and dry impregnation methods. They found that the wet-impregnated sample, WWT(W), shows a slightly higher activity than the impregnated one (D) at around 200 °C and this was ascribed to the formation of agglomerated V species which increases the intensity of Brønsted acid sites and the oxidation of adsorbed NH3, while isolated V species were predominant on the dry-impregnated catalyst. After the addition of KCl, both catalysts deactivated due to the decreasing amount of ammonium ions coordinated on the Brønsted acid sites. Isolated vanadia species (DK) appeared to be more reactive with potassium and therefore more severely deactivated.
Dong et al. [25] optimized the preparation conditions (wet impregnation) by focusing on the effect of the pH value of the vanadium precursor solution on the catalytic performance of V2O5-WO3/TiO2 in low temperature NH3-SCR of NOx. According to XPS (X-Ray Photoelectron Spectroscopy), Raman, H2-TPR (Temperature Programmed Reduction), NH3-TPD (Temperature Programmed Desorption), and NH3-DRIFT (Diffuse Reflectance Infrared Fourier Transform Spectroscopy) characterizations, it resulted that the outstanding activity of the catalyst with the decreased pH value could be responsible for the following: More polymeric vanadium species were formed on the catalyst surface; meanwhile, the ratio of V 4+(3+) /V 5+ , surface acidity, and quantity of active sites were increased. Thus, it should be considered that an enhancement of the precursor solution acidity is an effective way to improve SCR catalysts' activity.
The influence of the catalyst synthesis method on the SCR of NO by NH3 over V2O5-WO3/TiO2 catalysts was also investigated by He et al. [26]. In this article, they prepared the catalysts by two different synthesis methods: Co-precipitation of aqueous vanadium and tungsten oxide precursors with TiO(OH)2 and incipient wetness impregnation of the aqueous precursors on a crystalline TiO2 support. The composition difference of the two types of catalysts was confirmed by XRD (X-Ray Dong et al. [25] optimized the preparation conditions (wet impregnation) by focusing on the effect of the pH value of the vanadium precursor solution on the catalytic performance of V 2 O 5 -WO 3 /TiO 2 in low temperature NH 3 -SCR of NO x . According to XPS (X-Ray Photoelectron Spectroscopy), Raman, H 2 -TPR (Temperature Programmed Reduction), NH 3 -TPD (Temperature Programmed Desorption), and NH 3 -DRIFT (Diffuse Reflectance Infrared Fourier Transform Spectroscopy) characterizations, it resulted that the outstanding activity of the catalyst with the decreased pH value could be responsible for the following: More polymeric vanadium species were formed on the catalyst surface; meanwhile, the ratio of V 4+(3+) /V 5+ , surface acidity, and quantity of active sites were increased. Thus, it should be considered that an enhancement of the precursor solution acidity is an effective way to improve SCR catalysts' activity.
The influence of the catalyst synthesis method on the SCR of NO by NH 3 over V 2 O 5 -WO 3 /TiO 2 catalysts was also investigated by He et al. [26]. In this article, they prepared the catalysts by two different synthesis methods: Co-precipitation of aqueous vanadium and tungsten oxide precursors with TiO(OH) 2 and incipient wetness impregnation of the aqueous precursors on a crystalline TiO 2 support. The composition difference of the two types of catalysts was confirmed by XRD (X-Ray Diffraction) and HS-LEIS (High Sensitivity-Low Energy Ion Scattering Spectroscopy) and a distinction of their activity was also revealed by in situ Raman and IR (Infrared) spectroscopy. Co-precipitated catalysts contain oligomerized surface mono-oxo O = VO 3 and O = WO 4 sites on the TiO 2 support. The enhanced performance of co-precipitated catalysts compared to the impregnated ones was associated with the formation of new O = WO 4 species, able to adsorb NH 3 , and to the presence of redox surface mono-oxo O = VO 3 sites.
Qi et al. [27] reported the synthesis of V 2 O 5 -WO 3 /TiO 2 catalysts from waste commercial SCR catalysts through oxalic acid leaching and impregnation with different V 2 O 5 loadings (0.5%, 1.0%, and 1.5 wt %). The NO conversion of the newly synthesized catalyst with 1.0% V 2 O 5 almost recovered the activity of the fresh catalyst (91% of the NO conversion at 300 • C) and showed good resistance to SO 2 and H 2 O.
Meiqing et al. [28] systematically investigated Ce-modified V 2 O 5 -WO 3 /TiO 2 (V/WTi) catalysts by the deposition-precipitation method (V/Ce/WTi-DP and Ce/WTi-DP) and the conventional impregnation method (V/Ce/WTi-IMP and Ce/WTi-IMP). As shown in Figure 2, the results revealed that the activity was mainly dependent on the status of Ce, which not only possessed more surface Ce species but also presented a higher reducibility of the Ce species. The higher NO x conversion can be assigned to more active Lewis acid sites, the weakly adsorbed NO 2 , and the monodentate nitrate. Qi et al. [27] reported the synthesis of V2O5-WO3/TiO2 catalysts from waste commercial SCR catalysts through oxalic acid leaching and impregnation with different V2O5 loadings (0.5%, 1.0%, and 1.5 wt %). The NO conversion of the newly synthesized catalyst with 1.0% V2O5 almost recovered the activity of the fresh catalyst (91% of the NO conversion at 300 °C) and showed good resistance to SO2 and H2O.
Meiqing et al. [28] systematically investigated Ce-modified V2O5-WO3/TiO2 (V/WTi) catalysts by the deposition-precipitation method (V/Ce/WTi-DP and Ce/WTi-DP) and the conventional impregnation method (V/Ce/WTi-IMP and Ce/WTi-IMP). As shown in Figure 2, the results revealed that the activity was mainly dependent on the status of Ce, which not only possessed more surface Ce species but also presented a higher reducibility of the Ce species. The higher NOx conversion can be assigned to more active Lewis acid sites, the weakly adsorbed NO2, and the monodentate nitrate. Djerad et al. [29] prepared V 2 O 5 -WO 3 /TiO 2 catalysts by the sol-gel method with different W and V loadings and calcined at different temperatures. The results are shown in Figure 3, in which it is shown that the V 2 O 5 content strongly influences the catalytic behavior: Weakly active for the SCR reaction but with a high selectivity to N 2 . Moreover, the NO x conversion and N 2 selectivity achieved up to 100% in the temperature range of 250 to 325 • C. Meanwhile, 3 wt % V 2 O 5 -9 wt % WO 3 /TiO 2 was the best catalyst not only for the NO x conversion, but also for the N 2 selectivity.
Meiqing et al. [28] systematically investigated Ce-modified V2O5-WO3/TiO2 (V/WTi) catalysts by the deposition-precipitation method (V/Ce/WTi-DP and Ce/WTi-DP) and the conventional impregnation method (V/Ce/WTi-IMP and Ce/WTi-IMP). As shown in Figure 2, the results revealed that the activity was mainly dependent on the status of Ce, which not only possessed more surface Ce species but also presented a higher reducibility of the Ce species. The higher NOx conversion can be assigned to more active Lewis acid sites, the weakly adsorbed NO2, and the monodentate nitrate. Reiche et al. [30] explored the influence of the grafting sequence and simultaneous grafting of V 2 O 5 and WO 3 on the properties of V 2 O 5 -WO 3 /TiO 2 catalysts. DRIFT and XPS indicated that V and W species do not exist independently on the TiO 2 surface. More importantly, V 2 O 5 -WO 3 /TiO 2 catalysts synthesized by sequential grafting do not result in a hierarchical structure. However, changing the grafting mode and metal oxides loadings significantly affected the strength of interaction of V 2 O 5 and WO 3 .
In Table 1 and from the above-mentioned research, we cannot directly affirm which was the best method to synthesize catalysts with high activity and selectivity in a broad temperature window. New methods able to improve the specific surface area and to increase the active sites of catalysts should be considered in the design of new catalysts. In addition, the activity and selectivity are significantly affected by the active metal oxides loading.

Effect of the Support on Catalytic Activity
To enhance the stability and the activity, V 2 O 5 -WO 3 -based multiple metal oxides and supported catalysts have been widely studied. They can be classified into two types: Metal oxides and molecular sieves or filters as carriers.
2.3.1. Metal Oxides as Carriers V 2 O 5 -WO 3 /TiO 2 is a classical commercial catalyst that has been widely investigated for NO SCR. It has been reported that only the TiO 2 anatase phase has no activity for NO SCR with NH 3 [31]. However, it is well known that TiO 2 anatase improves the catalytic activity by dispersing the V 2 O 5 -WO 3 active phases [32]. At high temperatures, >700 • C, transition from the anatase to rutile phase occurs and such process is strongly influenced by the oxygen vacancies' concentration [21]. As will be discussed in the following Section 2.4, the presence of oxygen vacancies in the support helps the formation of W 5+ species able to active superoxide ions.
Reiche et al. [30,33] prepared V 2 O 5 -WO 3 /TiO 2 by simultaneous grafting of TiO 2 with V 2 O 5 and WO 3 . According to the DRIFT, XPS, and TOF-SIMS (Time-of-Flight secondary ion mass spectrometry) investigations, indicating V and W species do not exist independently on the titania surface (V-O-W connectivity were present on the V 2 O 5 -WO 3 /TiO 2 catalysts), which results in a higher activity compared to the corresponding titania-supported single oxides. The studies demonstrate that the activity could be enhanced by grafting V 2 O 5 -WO 3 on TiO 2 . Zhang et al. [34] elucidated the interaction of V, W, and Ti species on the surface of V 2 O 5 -WO 3 /TiO 2 catalysts for the improvement of the catalytic activity. XRD, UV-vis, PL spectra, DFT theoretical calculations, and in situ DRIFT showed that WO 3 could interact with TiO 2 to improve the electrons transfer due to the V-O-W species, in agreement with Reiche et al. [30,33]. Moreover, the XPS and EPR observations confirmed the presence V 4+ species, which are the active sites for the superoxide ion formation. Such species promoted NO oxidation to NO 3 − and improved the decomposition of NO 3 − to NO 2 , facilitating the reaction rate of "fast SCR". Kompio et al. [23] highlighted a new point on the relations between tungsten and vanadium, studying two series of V 2 O 5 -WO 3 /TiO 2 catalysts with different V 2 O 5 contents (0.5-5 wt %) and 0 or 10 wt % WO 3 impregnated over titania hydrate with an anatase structure. The catalysts were investigated by XRD, specific surface area measurements, Raman and EPR (Electron Paramagnetic Resonance) spectroscopy, temperature-programmed reduction (TPR), and SCR. It was concluded that tungsten plays a promotional effect on the neighboring vanadium oxide species, favoring their dispersion.
The positive role of W was confirmed also by Zhang et al. [35] and Wang et al. [36], who found that WO 3 species play an important role in the formation of superoxide ions (O 2 − ) over V 2 O 5 -WO 3 /TiO 2 catalyst for NH 3 -SCR of NO. This was due to an improved electron transfer between the V-O-W species, which facilitates the formation of reduced V 2 O 5 (V 4+ and V 3+ ). The best activity was registered with a WO3 loading of 6 wt %. However, it should be noticed that WO3 was an electron withdrawal, so higher WO3 loadings might trap electrons to inhibit the formation of superoxide ions. Zhang et al. [37] investigated the promotional roles of ZrO 2 and WO 3 over V 2 O 5 -WO 3 /TiO 2 -ZrO 2 SCR catalysts. They found that ZrO 2 improved the thermally stability, while WO 3 increased the Brønsted acid sites of the catalysts. Furthermore, a combination of WO 3 and ZrO 2 enhanced the dispersion of all metal oxides. As shown in Figure 4, the highest NO x conversion was achieved over V 2 O 5 -9wt.% WO 3 /TiO 2 -ZrO 2 and this finding partially results from the ZrO 2 doping that considerably increased the BET surface area and favored the dispersion of the active components on the surface. However, the mainly promotional role was to change the pathway of NO x reduction and the surface acidity of V 2 O 5 -WO 3 /TiO 2 . There are only Lewis acid sites on the V 2 O 5 -WO 3 supported over TiO 2 -ZrO 2 , while Brønsted and Lewis acid sites coexisted on V 2 O 5 -WO 3 /TiO 2 . Resonance) spectroscopy, temperature-programmed reduction (TPR), and SCR. It was concluded that tungsten plays a promotional effect on the neighboring vanadium oxide species, favoring their dispersion. The positive role of W was confirmed also by Zhang et al. [35] and Wang et al. [36], who found that WO3 species play an important role in the formation of superoxide ions (O2 − ) over V2O5-WO3/TiO2 catalyst for NH3-SCR of NO. This was due to an improved electron transfer between the V-O-W species, which facilitates the formation of reduced V2O5 (V 4+ and V 3+ ). The best activity was registered with a WO3 loading of 6 wt %. However, it should be noticed that WO3 was an electron withdrawal, so higher WO3 loadings might trap electrons to inhibit the formation of superoxide ions. Zhang et al. [37] investigated the promotional roles of ZrO2 and WO3 over V2O5-WO3/TiO2-ZrO2 SCR catalysts. They found that ZrO2 improved the thermally stability, while WO3 increased the Brønsted acid sites of the catalysts. Furthermore, a combination of WO3 and ZrO2 enhanced the dispersion of all metal oxides. As shown in Figure 4, the highest NOx conversion was achieved over V2O5-9wt.% WO3/TiO2-ZrO2 and this finding partially results from the ZrO2 doping that considerably increased the BET surface area and favored the dispersion of the active components on the surface. However, the mainly promotional role was to change the pathway of NOx reduction and the surface acidity of V2O5-WO3/TiO2. There are only Lewis acid sites on the V2O5-WO3 supported over TiO2-ZrO2, while Brønsted and Lewis acid sites coexisted on V2O5-WO3/TiO2.
Herrmann et al. [38] prepared a V2O5-WO3/TiO2 catalyst by supporting the V2O5-WO3 over TiO2 EUROCAT and investigated the electrical properties under different atmospheres, such as pure Herrmann et al. [38] prepared a V 2 O 5 -WO 3 /TiO 2 catalyst by supporting the V 2 O 5 -WO 3 over TiO 2 EUROCAT and investigated the electrical properties under different atmospheres, such as pure oxygen, NO, and NH 3 with and without O 2 . In the absence of oxygen, NO has an electron acceptor character (as NO -) that is able to fill the anionic vacancies, which are mostly present on V 2 O 5 . On the other hand, NH 3 had a strong reduction behavior with the formation of ionized vacancies. In the presence of oxygen, the anionic vacancies of the catalysts were filled by oxygen, therefore, the NO did not exhibit its electron acceptor character anymore, reacting directly with NH 3 , bounded on the acid sites. Under the NO + NH 3 mixture, without oxygen, an increased electrical conductivity was found, ascribed to the formation of anionic vacancies, especially over vanadia, due to dihydroxylation and dehydration of the surface.
Jung et al. [39] supported V 2 O 5 -WO 3 on TiO 2 -SO 4 2− and characterized the catalysts by BET, XRD, XPS, TPD, and H 2 -TPR analyses. Compared with the V 2 O 5 -WO 3 /TiO 2 catalyst, in the V 2 O 5 -WO 3 /TiO 2 -SO 4 2− , the sulfate deposited over titania did not affect the physical properties (in terms of the BET surface area and phase transformation). Conversely, in TPD and H 2 -TPR, it was found that the sulfate stability strongly depended on the loaded metal oxides. In addition, the reduction property of V 2 O 5 also changed due to the interaction between the V-O-W species. In Figure 5, the SCR activity in terms of the NO conversion over Catalysts 2019, 9, x FOR PEER REVIEW 8 of 30 SCR activity in terms of the NO conversion over TiO2-SO4 2− , V2O5/TiO2-SO4 2− , WO3/TiO2-SO4 2− , and V2O5-WO3/TiO2-SO4 2− , is compared. The similar activity between V2O5/TiO2-SO4 2− and V2O5-WO3/TiO2-SO4 2− at low temperatures was explained by focusing on the dispersion between acid and redox sites. At high temperatures, the redox property modified by the V-O-W species contributed to the high activity. Sun et al. [40] prepared a V2O5-WO3/Ti0.5Sn0.5O2 (TS) SCR catalyst by supporting V2O5-WO3 on Ti0.5Sn0.5O2. The results revealed that the activity of the NH3-SCR reaction was closely correlated to the Brønsted acid sites, and they were in direct proportion with the amount of WO3 up to the highly-dispersed particles formed. Indeed, when the WO3 exist in a dispersed state, the tungsten atoms are adjacent to the vanadium atoms forming the V-O-W bonds that are responsible for the increased Brønsted acidity. However, further increases of WO3 (loadings ≥1.0 mmol/100 m2 TS) results in the formation of crystalline WO3 that covers the active sites, leading to a decrease of the activities.
Camposeco et al. [41] synthesized a series of titanic acid nanotubes (TANs) by the conventional hydrothermal method. Then, V2O5/WO3 was added by wet impregnation. The catalytic activity of the V2O5 (5 wt %)-WO3 (3 wt %)/TAN catalysts was strongly improved for the SCR-NH3 process in a wide range of temperatures in comparison with both samples, V2O5 (10 wt %)/TAN and V2O5-WO3/TiO2 (see Figure 6a). Interestingly, the above mentioned V2O5 (5 wt %)-WO3 (3 wt %)/TAN catalyst showed better NO conversion values than V2O5 (10 wt %)/TAN in the presence of water (10 vol %) and SO2 (50 ppm) (Figure 6b). The reason for the higher activity can be explained by the following factors: (1) The excellent physical properties that TANs possess, such as tubular structure, good thermal stability (460 °C), and high specific surface area (314 m 2 /g); and (2) the presence of Brønsted and Lewis acid sites that promote NO conversion in a wide range of temperatures.
Analogous effects of TANs were reported by Mejía-Centeno et al. [42] (Figure 6c,d). Aguilar-Romero et al. [22] also prepared TANs and employed these as support for V2O5-WO3. They investigated: The effect of WO3 and V2O5 loading on the TiO2-NT surface acidity, the effect of WO3 The results revealed that the activity of the NH 3 -SCR reaction was closely correlated to the Brønsted acid sites, and they were in direct proportion with the amount of WO 3 up to the highly-dispersed particles formed. Indeed, when the WO 3 exist in a dispersed state, the tungsten atoms are adjacent to the vanadium atoms forming the V-O-W bonds that are responsible for the increased Brønsted acidity. However, further increases of WO 3 (loadings ≥ 1.0 mmol/100 m 2 TS) results in the formation of crystalline WO 3 that covers the active sites, leading to a decrease of the activities.
Camposeco et al. [41] synthesized a series of titanic acid nanotubes (TANs) by the conventional hydrothermal method. Then, V 2 O 5 /WO 3 was added by wet impregnation. The catalytic activity of the V 2 O 5 (5 wt %)-WO 3 (3 wt %)/TAN catalysts was strongly improved for the SCR-NH 3 process in a wide range of temperatures in comparison with both samples, V 2 O 5 (10 wt %)/TAN and V 2 O 5 -WO 3 /TiO 2 (see Figure 6a). Interestingly, the above mentioned V 2 O 5 (5 wt %)-WO 3 (3 wt %)/TAN catalyst showed better NO conversion values than V 2 O 5 (10 wt %)/TAN in the presence of water (10 vol %) and SO 2 (50 ppm) (Figure 6b). The reason for the higher activity can be explained by the following factors: (1) The excellent physical properties that TANs possess, such as tubular structure, good thermal stability (460 • C), and high specific surface area (314 m 2 /g); and (2) the presence of Brønsted and Lewis acid sites that promote NO conversion in a wide range of temperatures.
Catalysts 2019, 9, x FOR PEER REVIEW 9 of 30 nanotubes. Moreover, they concluded that the morphology of the nanotubes and the metal loading play important roles for the catalyst activity. Boudali et al. [43] prepared a WO3-V2O5/STi-PILC (Sulfated Titanium Pillared Clay) catalyst for NH3-SCR. The results showed that the highest NO conversion can reach up to 100% at 300 °C. Vanadia enhanced the NO conversion over STi-PILC, while no significant effect of tungsten was observed for WO3-V2O5/STi-PILC catalyst. When both tungsten and sulfate were simultaneously present on the surface of vanadia supported STi-PILC, the sulfate species seemed to play a more important role for NO abatement than tungsten.
Najbar et al. [44] synthesized Ti-Sn-rutile for deposition of WO3-V2O5. The SCR activity and surface species structure were determined for the freshly prepared catalyst, for the catalyst previously used in NO reduction by ammonia (320 ppm NO, 335 ppm NH3, and 2.35 vol % O2) at 300 °C as well as for the catalyst previously annealed at 300 °C under 2.35 vol % O2/He. A significant evolution of the active species occurred during the use of the catalyst at 300 °C, especially the increase of the oxidation state of vanadium and the decrease of the tungsten content in the surface layers of the catalyst. Changes in the SCR activity occurring upon the different treatments were discussed in terms of dissociative adsorption of water and the OH group's Brønsted acidity. In detail, the dissociative adsorption of water on V 5+ OW 6+ sites appeared to be mainly responsible for Analogous effects of TANs were reported by Mejía-Centeno et al. [42] (Figure 6c,d). Aguilar-Romero et al. [22] also prepared TANs and employed these as support for V 2 O 5 -WO 3 . They investigated: The effect of WO 3 and V 2 O 5 loading on the TiO 2 -NT surface acidity, the effect of WO 3 on the V 4+ /V 5+ ratio, and its correlation with the catalytic SCR activity. V 2 O 5 -WO 3 /TiO 2 -NT displayed a higher NO conversion than WO 3 /TiO 2 -NT and V 2 O 5 /TiO 2 -NT. In fact, the addition of both V 2 O 5 and WO 3 is responsible for the increased surface acidity (Brønsted and Lewis) of the TiO 2 -nanotubes. Moreover, they concluded that the morphology of the nanotubes and the metal loading play important roles for the catalyst activity.
Boudali et al. [43] prepared a WO 3 -V 2 O 5 /STi-PILC (Sulfated Titanium Pillared Clay) catalyst for NH 3 -SCR. The results showed that the highest NO conversion can reach up to 100% at 300 • C. Vanadia enhanced the NO conversion over STi-PILC, while no significant effect of tungsten was observed for WO 3 -V 2 O 5 /STi-PILC catalyst. When both tungsten and sulfate were simultaneously present on the surface of vanadia supported STi-PILC, the sulfate species seemed to play a more important role for NO abatement than tungsten.
Najbar et al. [44] synthesized Ti-Sn-rutile for deposition of WO 3 -V 2 O 5 . The SCR activity and surface species structure were determined for the freshly prepared catalyst, for the catalyst previously used in NO reduction by ammonia (320 ppm NO, 335 ppm NH 3 , and 2.35 vol % O 2 ) at 300 • C as well as for the catalyst previously annealed at 300 • C under 2.35 vol % O 2 /He . A significant evolution of the active species occurred during the use of the catalyst at 300 • C, especially the increase of the oxidation state of vanadium and the decrease of the tungsten content in the surface layers of the catalyst. Changes in the SCR activity occurring upon the different treatments were discussed in terms of dissociative adsorption of water and the OH group's Brønsted acidity. In detail, the dissociative adsorption of water on V 5+ OW 6+ sites appeared to be mainly responsible for the activity at 200 • C, while water adsorption on both the V 5+ OW 6+ and V 4+ OW 6+ sites determined the activity at 250 • C.

Molecular Sieves or Filters as the Carrier
Li et al. [45] reported an enhanced activity of NO SCR and mechanical stability of V 2 O 5 -WO 3 /TiO 2 catalyst supported on cordierite after acid pre-treatment of the cordierite substrate and the addition of silica. The acid pre-treatment of cordierite significantly increased the BET surface area and pore volume of the catalysts.
Silica increased the surface acidity of the V 2 O 5 -WO 3 /TiO 2 catalyst, and the strongest surface acidity was registered at the Si/Ti molar ratio of 5/5. Moreover, by XPS analysis, it was shown that the incorporation of Si shifted the V valence from V 5+ to V 4+ , increasing the amount of chemisorbed oxygen, with a positive effect on SCR activity. The optimal catalyst showed a high NO conversion above 90% at 300 to 500 • C, high mechanical stability, and resistance to sulfur and hydrothermal aging.
Kim et al. [46] fabricated three kinds of V 2 O 5 -WO 3 /TiO 2 supported-SiC catalytic filters by using the rotational coating method and investigated their NO conversion performances. Three different fabrication routes were used in order to find the effective method for the preparation of the best V 2 O 5 -WO 3 /TiO 2 supported-SiC catalytic filters. Method-III among the three different methods used appeared to be the most effective one. This method deposited the active component, V2O5-WO3, by the wet impregnation method over titania, then the powder was dried and calcined at 450 • C for 5 h. The resulting preformed catalyst was ground in wet solution conditions with a ball mill and then used for coating the SiC filter. It was demonstrated that by grinding the catalyst with ball milling, the catalyst particle size was lowered and the particles were dispersed in the coating solution (to produce the high catalytic surface area), which diffused into the pores of the SiC filters, favoring the catalyst dispersion and increasing the exposed active sites. The catalytic filter prepared by method-III (with a 50 nm catalyst particle confirmed with SEM) showed the 99.9% NO conversion with the Nx-slip concentration below 15 ppm in the optimal temperature range from 230 to 350 • C. Figure 7a,b display the NO conversions and N 2 O concentration, respectively, over V 2 O 5 -WO 3 /TiO 2 supported-SiC catalysts prepared with the three different methods.
Choi et al. [47] investigated the effect of Pt on the V 2 O 5 -WO 3 /TiO 2 supported-SiC catalytic filter (Pt-V 2 O 5 -WO 3 -TiO 2 /SiC). The Pt-added catalytic filter lowered the optimum working temperature from 280 to 330 • C for the non-Pt impregnated catalyst (V 2 O 5 -WO 3 -TiO 2 /SiC) to 180 to 230 • C for the Pt-V 2 O 5 -WO 3 -TiO 2 /SiC. Pt promoted catalytic activity at low temperatures, but increased ammonia oxidation. The Pt promotional effect was believed to result from a high electron transfer achieved on the Pt-catalytic system, as shown in Figure 7c,d.
area and pore volume of the catalysts.
Silica increased the surface acidity of the V2O5-WO3/TiO2 catalyst, and the strongest surface acidity was registered at the Si/Ti molar ratio of 5/5. Moreover, by XPS analysis, it was shown that the incorporation of Si shifted the V valence from V 5+ to V 4+ , increasing the amount of chemisorbed oxygen, with a positive effect on SCR activity. The optimal catalyst showed a high NO conversion above 90% at 300 to 500 °C, high mechanical stability, and resistance to sulfur and hydrothermal aging. Kim et al. [46] fabricated three kinds of V2O5-WO3/TiO2 supported-SiC catalytic filters by using the rotational coating method and investigated their NO conversion performances. Three different fabrication routes were used in order to find the effective method for the preparation of the best V2O5-WO3/TiO2 supported-SiC catalytic filters. Method-III among the three different methods used appeared to be the most effective one. This method deposited the active component, V2O5-WO3, by the wet impregnation method over titania, then the powder was dried and calcined at 450 °C for 5 h. The resulting preformed catalyst was ground in wet solution conditions with a ball mill and then used for coating the SiC filter. It was demonstrated that by grinding the catalyst with ball milling, the catalyst particle size was lowered and the particles were dispersed in the coating solution (to Najbar et al. [48] focused on the competition between NO reduction and decomposition over V 2 O 5 -WO 3 /Ti(Sn)O 2 catalyst on a Cr-Al steel monolith. Furthermore, they also revealed that the NO conversion was higher than NH 3 over the reduced catalyst. An increase of the pre-reduction temperature was in favor of improving the activity and stability in direct NO decomposition. The reason for the high selectivity of NO decomposition to N 2 was attributed to tungsten nitrosyl complexes via the W-N bond.
Zhao et al. [49] obtained promising catalytic performance over monolithic V 2 O 5 -WO 3 /TiO 2 catalysts by impregnating the active components (V 2 O 5 , WO 3 , and TiO 2 ) and optimizing the relative amounts and the deposition procedure. The active coating layers were stably deposited over the monolithic substrate as revealed by ultrasonic treatment. XPS analysis confirmed the co-existence of V 5+ and V 4+ species, while the BET surface area was significantly improved after the coating of the monolith due to its porous structure.
According to the data summarized in Table 2, it reveals that SiC or Pt doped SiC filters are the best carriers for promotion, at low-temperatures, of the activity of V 2 O 5 -WO 3 -based catalysts. Indeed, SiC and Pt-doped Si-C filters play an important role in increasing the BET surface area and the dispersion of active components, as well as in the promotion of electron transfer due to Pt introduction. All these factors should be taken into account in the design of new catalysts in the future. The well dispersed and isolated vanadium oxide species were found to be weakly active for the SCR reaction but with a high selectivity to N 2 .
[29] Activity was closely correlated to the amounts of the Brønsted acid sites: they were in direct proportion to highly dispersed WO 3 . However, too much WO 3 will cover the active sites and lead to the decrease of the activities; About the reduction temperature, which becomes higher due to the V-O-W species.
[40] The higher activity can be explained by the following factors: (1) The excellent physics properties that titanic acid nanotubes possess tubular structure, good thermal stability (460 • C) and high specific surface area (314 m 2 /g).
(2) TAN as carriers not only significantly increase Brønsted acid sites that light-off conversion at low temperatures, but also increase Lewis acid sites that promote the NO reduction at high temperature. [41,42] WO 3 -V 2 O 5 /STi-PILC STi-PILC 90-100% (250-400 • C)~S ulfate species seems to play a more important role for NO removal activity than WO 3 .
[43] It was important to facilitate the NH 3 -SCR reaction that XPS confirmed the co-existence of V 5+ and V 4+ and the BET surface area was significantly improved due to the pore structure. [49]

Effect of Chemical Composition on Catalytic Activity
Shen et al. [50] prepared V 2 O 5 -WO 3 /TiO 2 and V 2 O 5 -WO 3 /TiO 2 -SiO 2 SCR catalysts and investigated the resistance toward K and SO 2 poisoning. The calcination temperature of V 2 O 5 -WO 3 /TiO 2 catalysts was varied between 400 and 750 • C and played an important role for the activity (Figure 8a). Three different weight ratios of SiO 2 /TiO 2 were used, 3 wt %, 6 wt %, and 9 wt %, respectively, and the loading of K poison was fixed as a K/V molar ratio of 0.5. A different catalytic behavior was observed for silica doped and non-doped catalysts: Over K-poisoned V 2 O 5 -WO 3 /TiO 2 -SiO 2 catalysts, (Figure 8b), the NO conversion values reached up to 90% to 98% (280-410 • C), with a positive effect observed by increasing the silica loading. On the contrary, the K-poisoned catalyst without SiO 2 showed lower activity with a maximum NO conversion of~70% between 350 and 420 • C. The outstanding performances of V 2 O 5 -WO 3 /TiO 2 -SiO 2 SCR catalysts were explained by the catalysts' increased acidity due to the silica presence that prevents K poisoning, forming Si-OK scavenger species. As it concerns the SO 2 poisoning, the authors investigated the effect on the catalyst, V 2 O 5 -WO 3 /TiO 2 , calcined at 750 • C, by adding 1500 ppm to the flue gas. They found that the NO conversion declined sharply by 25% in half an hour, then, by decreasing the SO 2 concentration to 750 ppm over 4 h, the catalyst partially recovered the activity and maintained the 45% NO conversion for the following 3 h. plays different roles: Inhibition of the shrinkage of the catalyst BET surface area due to the phase transition from anatase to rutile and the growth of the TiO2 size; and maintenance of the highly-dispersed W species. Dong et al. [52] prepared Mn-Ce/TiO2 (denoted as M) and Cu-Ce/TiO2 (denoted as C) by the sol-gel method, then a series of V2O5-WO3/M/cordierite honeycomb ceramics and V2O5-WO3/C/cordierite honeycomb ceramics complex catalysts were obtained by coating V2O5-WO3, M, and C onto cordierite using the impregnation method. The results indicated that the higher activity originated from the high V 4+ /V 5+ ratio and the large amount of surface chemisorbed oxygen, as well as the high specific surface area increased by the M and C composites. Li et al. [53] prepared V2O5-WO3/TiO2 and CeO2-WO3/TiO2 catalysts by the impregnation method. In comparison with V2O5-WO3/TiO2 catalyst, the novel CeO2-WO3/TiO2 catalyst exhibited a much better resistance to hydrothermal aging. After aging at 670 °C in 5% H2O/air for 64 h, the NO conversion was equal to to 80% (238-573 °C) over CeO2-WO3/TiO2 catalyst. Such result was attributed to the well-dispersed CeO2 and W species, as well as to the formation of a new species, Ce2(WO4)3, due to the interaction of CeO2 and W. Ce2(WO4)3 prevented the W species' crystallization and TiO2 phase transition (anatase to rutilo) and sintering. Meanwhile, the existence of the Ce 4+ -Ce 3+ Liu et al. [51] also studied the effect of SiO 2 additive on the NH 3 -SCR activity and thermal stability of V 2 O 5 -WO 3 /TiO 2 -SiO 2 catalysts. Compared with V 2 O 5 -WO 3 /TiO 2 catalyst, the SCR activity and selectivity of V 2 O 5 -WO 3 /TiO 2 -SiO 2 catalyst was improved after aging at 750 • C in 10 vol % H 2 O/air for 24 h (in Figure 8c,d). The different behavior was explained by the addition of SiO 2 , which plays different roles: Inhibition of the shrinkage of the catalyst BET surface area due to the phase transition from anatase to rutile and the growth of the TiO 2 size; and maintenance of the highly-dispersed W species.
Dong et al. [52] prepared Mn-Ce/TiO 2 (denoted as M) and Cu-Ce/TiO 2 (denoted as C) by the sol-gel method, then a series of V 2 O 5 -WO 3 /M/cordierite honeycomb ceramics and V 2 O 5 -WO 3 /C/cordierite honeycomb ceramics complex catalysts were obtained by coating V 2 O 5 -WO 3 , M, and C onto cordierite using the impregnation method. The results indicated that the higher activity originated from the high V 4+ /V 5+ ratio and the large amount of surface chemisorbed oxygen, as well as the high specific surface area increased by the M and C composites.
Li et al. [53] prepared V 2 O 5 -WO 3 /TiO 2 and CeO 2 -WO 3 /TiO 2 catalysts by the impregnation method. In comparison with V 2 O 5 -WO 3 /TiO 2 catalyst, the novel CeO 2 -WO 3 /TiO 2 catalyst exhibited a much better resistance to hydrothermal aging. After aging at 670 • C in 5% H 2 O/air for 64 h, the NO conversion was equal to to 80% (238-573 • C) over CeO 2 -WO 3 /TiO 2 catalyst. Such result was attributed to the well-dispersed CeO 2 and W species, as well as to the formation of a new species, Ce 2 (WO 4 ) 3 , due to the interaction of CeO 2 and W. Ce 2 (WO 4 ) 3 prevented the W species' crystallization and TiO 2 phase transition (anatase to rutilo) and sintering. Meanwhile, the existence of the Ce 4+ -Ce 3+ couple provided more oxygen vacancies and improved the surface Lewis acid of the catalyst.
Shi et al. [54] added zirconia to V 2 O 5 -WO 3 /TiO 2 catalyst and investigated the performance before and after aging treatment (at 750 • C for 12 h). Compared with V 2 O 5 -WO 3 /TiO 2 catalyst, ZrO 2 -V 2 O 5 -WO 3 /TiO 2 catalyst without aging was slightly more active at high temperatures, but less active at low temperatures. After aging treatment, the NH 3 -SCR activity of the catalysts was significantly enhanced and exhibited the highest NO conversions and N 2 selectivity (150-550 • C). The following key factors were invoked as being responsible for the improved catalytic behavior: (1) Addition of ZrO 2 improves the thermal stability of ZrO 2 -V 2 O 5 -WO 3 /TiO 2 catalysts and preserves the surface area from sintering, resulting in a better dispersion of V sites after aging; (2) ZrO 2 enhances the strength of the surface acidic groups. Regarding the non-aged ZrO 2 -V 2 O 5 -WO 3 /TiO 2 catalyst, the decreased activity at low temperatures was attributed to the inhibitions due to ZrO 2 , which decrease the amount of Brønsted acid sites and V 5+ = O active sites, inhibiting ammonia absorption.
Tao et al. [55] focused on the relationship between the loadings of V 2 O 5 and WO 3 and the catalytic performance of V 2 O 5 -WO 3 /TiO 2 catalysts for the SCR of NO with NH 3 . The catalytic activity was enhanced by increasing the V 2 O 5 loading less than 2 wt %, in agreement with the presence of TiO 2 as the anatase phase (by XRD) and to the high dispersion on the catalyst surface of all the active species (by XPS). Conversely, when the V 2 O 5 loading exceeded 2 wt %, the activity began to decline because the high V 2 O 5 loading on TiO 2 favored the phase transition from anatase to rutile. On the other hand, the NO conversion was also improved by moderate WO 3 loadings up to 6 wt %. Exceeding such a percentage, WO 3 inhibited the formation of superoxide ions, decreasing the activity.
As shown in Figure 9, Lietti et al. [56] investigated the NO SCR over commercial and model V 2 O 5 -WO 3 /TiO 2 and V 2 O 5 -MoO 3 /TiO 2 catalysts. The results demonstrated that WO 3 and MoO 3 increased the activity as "structural" and "chemical" promoters for the catalysts, providing better mechanical, structural, and morphological properties. In Figure 9A,B, the NO conversion and N 2 selectivity values are compared for V 2 O 5 -WO 3 /TiO 2 and V 2 O 5 -MoO 3 /TiO 2 catalysts at different V 2 O 5 loadings and constant WO 3 or MoO 3 contents, respectively, 9 and 6 wt %. In both cases, the activity was increased by increasing the vanadia loading, so that the "temperature window" of the reaction was widened and shifted towards lower temperatures. However, the V 2 O 5 (x)-WO 3 (9)/TiO 2 samples exhibited higher selectivity in a larger range of temperatures. mechanical, structural, and morphological properties. In Figure 9a and 9b, the NO conversion and N2 selectivity values are compared for V2O5-WO3/TiO2 and V2O5-MoO3/TiO2 catalysts at different V2O5 loadings and constant WO3 or MoO3 contents, respectively, 9 and 6 wt %. In both cases, the activity was increased by increasing the vanadia loading, so that the "temperature window" of the reaction was widened and shifted towards lower temperatures. However, the V2O5(x)-WO3(9)/TiO2 samples exhibited higher selectivity in a larger range of temperatures. Xu et al. [57] investigated the low-temperature NH3-SCR activity of V2O5-Sb2O3/TiO2 and V2O5-WO3/TiO2 monolithic catalysts in the presence of SO2. V2O5-Sb2O3/TiO2 monolithic catalyst showed a higher NO conversion and broader temperature windows as well as a higher SO2 and H2O Xu et al. [57] investigated the low-temperature NH 3 -SCR activity of V 2 O 5 -Sb 2 O 3 /TiO 2 and V 2 O 5 -WO 3 /TiO 2 monolithic catalysts in the presence of SO 2 . V 2 O 5 -Sb 2 O 3 /TiO 2 monolithic catalyst showed a higher NO conversion and broader temperature windows as well as a higher SO 2 and H 2 O resistance. The excellent performance was suppressed by the extent of NH 4 HSO 4 accumulated on the catalyst surface. Table 3 summarizes the main results discussed so far on the chemical composition effects of V 2 O 5 -WO 3 /TiO 2 based catalysts.
The introduction of SiO 2 significantly increased the catalysts acidity and leaded to more V and W species were exposure on the surface of catalyst by forming Si-OK that was the result of the reaction between enriched K and Si. [50] The addition of SiO 2 on the one hand, inhibited the shrinkage of catalyst BET surface area due to the phase transition from anatase to rutile and the growth of TiO 2 size. On the other hands, the W species remained highly dispersion. [51] The higher activity origin from high V 4+ /V 5+ ratio and large amount of surface chemisorbed oxygen as well as BET surface area caused by the M and C composites.

Positive/Negative Effects of Different Doping Agents on Catalytic Activity and Selectivity
Addition or doping with other metal oxides or elements, such as F [21,58], Nb, Sb [59], Cu, Mn, and Ce [60][61][62] is the main way to improve catalyst activity, whereas alkaline and alkaline earth metals (Na, K, Ca, and Mg) [63] deactivate catalysts.
Zhang et al. [21] found that NO conversion was improved by F doping. As shown in Figure 10, V 2 O 5 -WO 3 /TiO 2 -F 1.35 (the ratio of Ti/F was 1.35) exhibited the highest NO removal efficiency (82.8% at 210 • C), owing to the improved interaction of WO 3 with TiO 2 forming oxygen vacancies and increasing the number of reduced W species (W 5+ ), which play an important role for superoxide ion formation. To further investigate the mechanism, molecular modeling was employed using density functional theory (DFT) and first principles molecular dynamics simulations [58]. catalyst surface. Table 3 summarizes the main results discussed so far on the chemical composition effects of V2O5-WO3/TiO2 based catalysts.

Positive/Negative Effects of Different Doping Agents on Catalytic Activity and Selectivity
Addition or doping with other metal oxides or elements, such as F [21,58], Nb, Sb [59], Cu, Mn, and Ce [60][61][62] is the main way to improve catalyst activity, whereas alkaline and alkaline earth metals (Na, K, Ca, and Mg) [63] deactivate catalysts.
Zhang et al. [21] found that NO conversion was improved by F doping. As shown in Figure 10, V2O5-WO3/TiO2-F1.35 (the ratio of Ti/F was 1.35) exhibited the highest NO removal efficiency (82.8% at 210 °C), owing to the improved interaction of WO3 with TiO2 forming oxygen vacancies and increasing the number of reduced W species (W 5+ ), which play an important role for superoxide ion formation. To further investigate the mechanism, molecular modeling was employed using density functional theory (DFT) and first principles molecular dynamics simulations [58]. Ye et al. [59] investigated the enhanced activity of Nb and Sb doped V2O5-WO3/TiO2 SCR catalysts (V2O5-WO3-Nb2O5/TiO2 and V2O5-WO3-Sb2O5/TiO2) that was assigned to the increase of reactive electrophilic oxygen species on the surface of V2O5-WO3-Nb2O5/TiO2 and V2O5-WO3-Sb2O5/TiO2. Moreover, the catalysts' deactivation by SO2 and H2O was inhibited through the addition of Nb2O5 and Sb2O5.
Tian et al. [60] prepared Cu, Mn, and Ce doped V2O5-WO3/TiO2 catalysts and studied their performances. The activity was improved because Cu, Mn, and Ce can make the surface of the catalyst moderately acidic, increasing the V 4+ /V 5+ ratio and enhancing the redox performance of the catalyst.
Chen et al. [61] also investigated Ce doped V2O5-WO3/TiO2 catalysts with low vanadium loadings in the NH3-SCR of the NOx, and the optimized catalyst showed an above 90% NOx conversion from 200 to 450 °C. Additionally, the selectivity could be significantly increased at high temperatures, even at a high GHSV (Gas Hourly Space Velocity) of 113,000 h −1 , and NOx conversion nearly reached 100% in the temperature range of 250 to 350 °C. According to XPS, analysis cerium was present as Ce 3+ , causing a charge imbalance and therefore oxygen vacancies' formation. Such defects increased the amount of chemisorbed oxygen, indicating that more NO can be oxidized to Tian et al. [60] prepared Cu, Mn, and Ce doped V 2 O 5 -WO 3 /TiO 2 catalysts and studied their performances. The activity was improved because Cu, Mn, and Ce can make the surface of the catalyst moderately acidic, increasing the V 4+ /V 5+ ratio and enhancing the redox performance of the catalyst.
Chen et al. [61] also investigated Ce doped V 2 O 5 -WO 3 /TiO 2 catalysts with low vanadium loadings in the NH 3 -SCR of the NO x , and the optimized catalyst showed an above 90% NO x conversion from 200 to 450 • C. Additionally, the selectivity could be significantly increased at high temperatures, even at a high GHSV (Gas Hourly Space Velocity) of 113,000 h −1 , and NO x conversion nearly reached 100% in the temperature range of 250 to 350 • C. According to XPS, analysis cerium was present as Ce 3+ , causing a charge imbalance and therefore oxygen vacancies' formation. Such defects increased the amount of chemisorbed oxygen, indicating that more NO can be oxidized to NO 2 .
The effect of Ce doping of TiO 2 support on NH 3 -SCR activity over V 2 O 5 -WO 3 /CeO 2 -TiO 2 catalyst was investigated by Cheng et al. [62]. The catalytic activity of V 2 O 5 -WO 3 /TiO 2 was greatly enhanced by Ce doping, the best NO conversion being achieved at the molar ratio of Ce/Ti = 1/10 (see Figure 11). The crystalline structure also played an important role on the activity; indeed, the catalysts showing predominantly anatase TiO 2 showed better catalytic performance than the catalysts with predominantly a fluorite CeO 2 structure. Moreover, the Ce additive enhanced the amount of oxygen adsorbed on the surface with beneficial effects on the SCR reaction.
Catalysts 2019, 9, x FOR PEER REVIEW 15 of 30 enhanced by Ce doping, the best NO conversion being achieved at the molar ratio of Ce/Ti = 1/10 (see Figure 11). The crystalline structure also played an important role on the activity; indeed, the catalysts showing predominantly anatase TiO2 showed better catalytic performance than the catalysts with predominantly a fluorite CeO2 structure. Moreover, the Ce additive enhanced the amount of oxygen adsorbed on the surface with beneficial effects on the SCR reaction. The activity of alkaline and alkaline earth metals (Na, K, Ca, and Mg) doped V2O5-WO3/TiO2 catalysts was investigated by Chen et al. [63]. The activity downward trend is: K > Na > Ca > Mg. Several effects must be taken into account in order to explain these findings: (1) Na and K decrease the quantity and stability of the Brønsted acid sites more than Mg and Ca; (2) the surface chemisorbed oxygen decreased in the same order; (3) the reducibility of vanadium species on the Na and K doped catalysts is lower than that on the Ca and Mg doped catalysts with the same alkali metal oxides loadings; (4) Na and K ions also affect the reduction degree of tungsten species, while no effect was experienced by Ca and Mg; and (5) the poisoning by alkali metals of vanadium species appears to be the main factors for V2O5-WO3/TiO2 catalysts' deactivation.
In Table 4, the NH3-SCR performances of V2O5-WO3/TiO2 catalysts doped by various elements (F, Nb, Sb, Na, K, Ca, Mg, Cu, Mn, Ce), as discussed above, are summarized. Based on such results, simultaneous doping with F and Ce could be the most effective way to optimize the performance of new V2O5-WO3/TiO2 based catalysts due to the positive roles played individually by F and Ce in enhancing the NOx conversion and selectivity.

Deactivation or Poisoning Mechanism
In practical applications of NO-SCR by NH3, the V2O5-WO3/TiO2 catalysts can be poisoned by several species, such as impurities of biodiesel (K, Na, P), urea solution (K, Na, Ca, Mg), and abrasion of the engine (Cr, Cu). As introduced in Section 2.5, alkaline and alkaline earth metals have The activity of alkaline and alkaline earth metals (Na, K, Ca, and Mg) doped V 2 O 5 -WO 3 /TiO 2 catalysts was investigated by Chen et al. [63]. The activity downward trend is: K > Na > Ca > Mg. Several effects must be taken into account in order to explain these findings: (1) Na and K decrease the quantity and stability of the Brønsted acid sites more than Mg and Ca; (2) the surface chemisorbed oxygen decreased in the same order; (3) the reducibility of vanadium species on the Na and K doped catalysts is lower than that on the Ca and Mg doped catalysts with the same alkali metal oxides loadings; (4) Na and K ions also affect the reduction degree of tungsten species, while no effect was experienced by Ca and Mg; and (5) the poisoning by alkali metals of vanadium species appears to be the main factors for V 2 O 5 -WO 3 /TiO 2 catalysts' deactivation.
In Table 4, the NH 3 -SCR performances of V 2 O 5 -WO 3 /TiO 2 catalysts doped by various elements (F, Nb, Sb, Na, K, Ca, Mg, Cu, Mn, Ce), as discussed above, are summarized. Based on such results, simultaneous doping with F and Ce could be the most effective way to optimize the performance of new V 2 O 5 -WO 3 /TiO 2 based catalysts due to the positive roles played individually by F and Ce in enhancing the NO x conversion and selectivity.

Deactivation or Poisoning Mechanism
In practical applications of NO-SCR by NH 3 , the V 2 O 5 -WO 3 /TiO 2 catalysts can be poisoned by several species, such as impurities of biodiesel (K, Na, P), urea solution (K, Na, Ca, Mg), and abrasion of the engine (Cr, Cu). As introduced in Section 2.5, alkaline and alkaline earth metals have a negative effect. Following the study of Chen et al. [63], further investigations on alkali ions as well as on other poisoning species have drawn great attention: Hg [64][65][66], alkali metals [66][67][68][69][70], P [69], Cr [69], Cu [69], Mg [69], Ca [69,71], and As [72]. The main poisoning elements and corresponding reasons for catalysts' deactivation are listed in Table 5. Hg Transform V 5+ species to V 4+ species and consumed the lattice oxygen. [65] KCl KCl could react with V-OH, leading the active sites for NH 3 absorption inactive. [66] Hg The gaseous Hg adsorbed on the vanadia sites reduce the active sites.

KCl and Hg
There is the competition for active sites that are partially reduced by gaseous Hg, while others can be increased by delaying the deactivation caused by KCl.

KCl
Lowering the acidity at different extent by different forms of V species; NH 3 adsorption temperature allows to evaluate the acid sites. [67] Alkali metal Decrease the ability of NH 3 adsorption but increase in the NH 3 desorption rate. [68] P, Cr and Cu Moderate poisoning of the catalysts due to competition between increase the N 2 O production and lowering the number of acid sites. [69] K, Na, Mg and Ca Reduced capacity of ab-NH 3. [69,70] Ca Can passivate surface acid sites of fresh and bulk W species. [71] As As 2 O 5 dense layers derived from the ab-As 2 O 3 oxidized of by chemisorption oxygen on catalyst surface prevent NH 3 adsorption and active sites recovery. [72] Kong et al. [64] investigated the HgCl 2 deactivation mechanism over V 2 O 5 -WO 3 /TiO 2 SCR catalysts. The results showed that HgCl 2 caused an agglomeration of active components, the BET surface area decreased significantly, and the reaction of HgCl 2 with Brønsted acid sites (V-OH) occurred. From NH 3 -TPD, NH 3 chemisorption, FTIR, and XPS characterization data, it was found that V-OH and V = O bonds were affected by HgCl 2 . NH 3 was adsorbed on new sites (Cl-V-O-H) that were not active for NO SCR, and HgCl 2 existed in a stable form bonded to the bridge sites.
Yang et al. [65] evaluated the activity of mercury exposed catalysts by XPS and found that mercury oxidation led to the transformation of the V 5+ to V 4+ species and consumed the lattice oxygen of catalyst, thus decreasing the activity. In another article, Kong et al. [66] introduced KCl by impregnation of the V 2 O 5 -WO 3 /TiO 2 catalysts and added Hg by the dry-gas adsorption method. The catalysts modified with KCl were deactivated due to the reaction of KCl with V-OH acidic groups forming -V-O-K and Cl-V-O-K inactive sites. Hg had a weaker negative effect, and it did not cause catalyst agglomeration and increased the amount of chemisorbed NH 3 , thus delaying the deactivation caused by KCl. However, Hg reacted with Cl (from KCl), forming HgCl or HgCl 2 . Then, inactive species, such as -V-O~Hg or -V-O-Hg-Cl, were formed that were able to adsorb NH 3 in competition with the active sites.
The deactivation of V 2 O 5 -WO 3 /TiO 2 catalyst induced by KCl was investigated systematically via three methods (wet impregnation, solid diffusion, and vapor deposition) by Lei et al. [67]. The deactivation rate occurred in the following order: Vapor deposition >> solid diffusion > wet impregnation, and was mainly attributed to the presence of different forms of V species that reduced the acidity to different extents as previously found by Yu et al. [24]. When KCl was added by wet impregnation, the activity of V 2 O 3 species was only slightly reduced, possibly due to the intimate interaction between K and oxygen in vanadium oxides. On the contrary, by the solid diffusion method, the formation of V 2 O 5 -K 2 S 2 O 7 eutectic occurred, leading to a very small surface area and a too strong oxidation ability, which were the main reasons for the catalyst deactivation.
Xie et al. [68] assessed the deactivation mechanism by investigating a commercial V 2 O 5 -WO 3 /TiO 2 catalyst through exposure to the flue gas of a coal-fired power plant by a transient kinetic analysis that focused on the distinction between the deactivation behaviors of adsorption sites and redox sites. The results showed that the alkali element preferentially poisons the active sites associated with vanadium (V 5+-OH and/or V 5+ = O) rather than the sites associated with Ti and W. Thus, it was concluded that the poisoning of vanadium-linked species caused decreased NH 3 adsorption and increased NH 3 desorption, negatively affecting the SCR reaction.
The deactivation mechanism of V 2 O 5 -WO 3 /TiO 2 ceramic honeycomb catalysts induced by P, K, Na, Mg, Ca, Cr, and Cu was assayed by Klimczak et al. [69]. The catalysts were poisoned by two different methods: (i) wet impregnation of dilute aqueous solution of the corresponding nitrate or ammonium salts; (ii) deposition of the inorganic aerosol particles. The second approach is more realistic as it uses reaction conditions close to the mobile application and it seems to be more harmful. Both impregnated and aerosol deactivated catalysts showed a strong poisoning effect of alkaline metals due to a reduced capacity of absorbed NH 3 . P, Cr, and Cu moderately poisoned the catalysts because there was competition between the increased N 2 O production and reduced acid sites. In Figure 12, the NO x conversion curves over V 2 O 5 -WO 3 /TiO 2 (400 cpsi monoliths) are compared [69]. Figure 12a shows fresh, hydrothermal aged and K, Na and Ca poisoned samples; Figure 12b displays fresh, hydrothermal aged and Cr and Cu poisoned catalysts. In both cases, a molar throughput of 3 mmol of poisoning was used and added by aerosol. Li et al. [71] studied the activity of Ca poisoning V2O5-WO3/ TiO2 catalysts based on the modification of activity, structure, reducibility, acidity, and reaction properties. The results revealed that the bulk W species and surface acid sites could be passivated with high contents of CaO, leading to CaWO4 formation. Additionally, the reducibility and surface oxygen species were also inhibited by CaO species.
Kong et al. [72] studied the poisoning of V2O5-WO3/TiO2 catalysts by arsenic (As). In Figure  13a,b, the NOx conversion curves over fresh and poisoned catalysts and the deactivation mechanism are displayed. The main reason for catalytic deactivation originates from the following: As2O3 would be adsorbed on the surface of catalyst, then oxidized into As2O5 by surface chemisorption oxygen. The As2O5 dense layers deposited on the catalyst surface prevent NH3 adsorption and the active sites' recovery. Deng et al. [70] studied alkali metals' deactivation process of V 2 O 5 -WO 3 /TiO 2 SCR catalyst by using an entrained-flow combustion system. The conclusions can be described as follows, and the activity of the catalyst sample was affected by the fly ash. The order of the poisoning effects of different alkali salts on the activity was Na 2 CO 3 < K 2 CO 3 < Na 2 SO 4 < K 2 SO 4 < NaCl < KCl. Both the decrease of the specific surface area and the masking of V 2 O 5 caused the activity reduction, which can be attributed to mechanical deactivation. More importantly, the deposited particles of alkali salts can both coordinate to the -OH on the surface and interact with the lattice oxygen in the V species, which can be assigned to chemical deactivation. All of them can be ascribed to the reduction of the capacity of ab-NH 3 .
Li et al. [71] studied the activity of Ca poisoning V 2 O 5 -WO 3 / TiO 2 catalysts based on the modification of activity, structure, reducibility, acidity, and reaction properties. The results revealed that the bulk W species and surface acid sites could be passivated with high contents of CaO, leading to CaWO 4 formation. Additionally, the reducibility and surface oxygen species were also inhibited by CaO species.
Kong et al. [72] studied the poisoning of V 2 O 5 -WO 3 /TiO 2 catalysts by arsenic (As). In Figure 13a,b, the NOx conversion curves over fresh and poisoned catalysts and the deactivation mechanism are displayed. The main reason for catalytic deactivation originates from the following: As 2 O 3 would be adsorbed on the surface of catalyst, then oxidized into As 2 O 5 by surface chemisorption oxygen. The As 2 O 5 dense layers deposited on the catalyst surface prevent NH 3 adsorption and the active sites' recovery.

SO2 and H2O Effects
A discussion about deactivation and poisoning effects on V2O5-WO3-based catalysts cannot exclude the influence of SO2 and/or H2O for realist applications. Indeed, flue gas emissions from power plants always contain SO2 (200−1500 ppm) that can be further oxidized to SO3 by the supported V2O5-WO3/TiO2 catalysts. Moreover, SO3 is corrosive to stainless steel SCR reactor Figure 13. NO x conversion over fresh and poisoned V 2 O 5 -WO 3 /TiO 2 catalysts (a); picture for deactivation mechanism (b). (Reprinted with permission from [72]. Copyright (2015) Elsevier).

SO 2 and H 2 O Effects
A discussion about deactivation and poisoning effects on V 2 O 5 -WO 3 -based catalysts cannot exclude the influence of SO 2 and/or H 2 O for realist applications. Indeed, flue gas emissions from power plants always contain SO 2 (200−1500 ppm) that can be further oxidized to SO 3 by the supported V 2 O 5 -WO 3 /TiO 2 catalysts. Moreover, SO 3 is corrosive to stainless steel SCR reactor components and readily reacts with ammonia to form ammonium sulfate/bisulfate that deposit on the catalysts, causing deactivation, and on the walls of heat exchangers to reduce their efficiency. Regarding the effects of water, it must be pointed out that significant amounts of moisture, 10% to 30%, are present in the flue gas exiting from industrial combustion processes. Moreover, moisture interacts with the surface of the catalysts and modifies the surface's active sites and the distribution of Lewis and Brønsted acid sites. However, the role of water has not been extensively investigated, so it deserves further attention.
In the 1990s, preliminary investigations about SO 2 effects were reported by Chen et al. [73], who found that SO 2 promotes the activity by increasing the Brønsted acidity, in accordance with the assumption that Brønsted acid sites are the active sites for the reaction. A few years later, the same authors [74] investigated SO 4 2− /TiO 2 superacid for NO SCR with NH 3 . Since SO 4 2− /TiO 2 is formed on TiO 2 under SCR reaction conditions when SO 2 is present, the results elucidated the role of TiO 2 support (and SO 2 ) in the SCR reaction. XPS and chemisorption measurements along with SCR kinetic results indicated that the reaction takes place by an Eley-Rideal mechanism, on Brønsted acid sites on SO 4 2− /TiO 2 surface. The catalytic results showed NO conversion is in the range of 96% to 100% at temperatures between 400 and 425 • C and the good performances were explained considering that at 400 • C, about 70% of the SO 4 2− ions are occupied by ammonia that is active for reactions with NO.
The promoting effect of the sulphate content on the catalytic activity of titania was also reported by Ciambelli et al. [75], who found an enhanced NO conversion without decreased selectivity to N 2 , which was attributed to the enhanced chemisorption of ammonia.
The coupled effect of SO 2 and H 2 O was investigated by Amiridis et al [76] at 350 • C over V 2 O 5 /TiO 2 catalysts with variable vanadia loadings. Independently on the vanadia surface coverage, the addition of H 2 O to the reaction mixture resulted in a decrease of the SCR turnover frequency of approximately 40% to 50% due to a competitive adsorption of H 2 O on the active vanadia sites. The presence of SO 2 in the gas phase resulted in a significant increase of the turnover frequency at low vanadia surface coverage due to the formation of surface sulfate species that in the presence of H 2 O, act as strong Brønsted acid sites. However, an excess of water had a negative effect in the SCR by slowing the reoxidation of vanadia.
The reported results on the effect of SO 2 and water thus far were confirmed by Magnusson et al. [16], who investigated how SO 2 affects the NO x activity of a V 2 O 5 -WO 3 /TiO 2 urea-SCR catalyst for marine applications. Surprisingly, the authors found that the addition of SO 2 in the absence of H 2 O promoted NO x reduction at 350 • C, while only H 2 O gave rise to a decrease in the NO x abatement and also an inhibition of the N 2 O formation. No promotional effect by SO 2 was observed at temperatures below 230 • C. Further, long term effects of SO 2 and H 2 O were investigated and the NOx reduction remained stable, also after long term exposure of SO 2 . Finally, it was concluded that in the presence of both H 2 O and SO 2 , the catalyst did not show any deactivation at temperatures above 300 • C and fairly low space velocities (below 12,200 h −1 ). However, at lower temperatures (250 • C) and/or higher space velocities, the catalytic performance for NOx reduction decreased with time. Therefore, for V 2 O 5 -WO 3 -based catalysts used at low temperatures, the effects of water and sulfur dioxide must be carefully addressed.
A recent detailed investigation about H 2 O and SO 2 effects has been carried out by Lai and Wachs [77]. This is a perspective study addressing the current fundamental understanding and advances on V 2 O 5 -WO 3 /TiO 2 catalysts taking into consideration the main parameters affecting the activity, such as the catalyst synthesis, molecular structures of titania supported vanadium and tungsten oxide species, surface acidity, catalytic active sites, surface reaction intermediates, reaction mechanism, rate determining-step, and reaction kinetics.
As remarked in [77] and as well in the references therein, water inhibits the SCR reaction rate at low (∼1%) but does not affect the SCR reaction at high concentrations (>5%). Such an effect arises from (1) competitive adsorption with the reactants (NO or NH 3 ) and/or (2) inhibition of the reaction between NO and adsorbed NH 3 . H 2 O increased the concentration of surface NH 4 + * species on Brønsted acid sites and such surface NH 4 + * intermediates arise from the reaction of NH 3 * species with moisture.
Regarding the effect of SOx, the effects of SO 2 and SO 3 on TiO 2 and supported V 2 O 5 /TiO 2 has received much attention, while less investigation has focused on tungsten catalysts. For V 2 O 5 -WO 3 /TiO 2 , both surface vanadia and tungsten oxide sites contribute equally to the overall SO 2 oxidation reaction to SO 3 .

Conclusions and Perspectives
In summary, a fundamental understanding and recent advances on the respective role of V 2 O 5 -WO 3 /TiO 2 catalysts were reviewed in detail, with special concern to the literature published in the last 20 years. It is generally accepted that WO 3 increases the Lewis acidity and inhibits the initial sintering of TiO 2 as well as helping to form superoxide ions. Concerning V 2 O 5 , it increases the surface Brønsted and Lewis acidity and enhances N 2 selectivity. The importance of interaction between V 2 O 5 and WO 3 was also revealed. In addition, the relationships between the catalytic activity and synthesis methods, supports, composition, and doping agents were discussed as well. It can be pointed out that alkaline and alkaline earth metals (K, Na, Ca, Mg) as well other elements, such as Cr, As, and Hg, deactivate the V 2 O 5 -WO 3 -based catalysts. Especially, alkali metals decrease the ability of NH 3 adsorption, which is a disadvantage for low-temperature activity and selectivity.
The specific surface area of titania plays an important role for the dispersion and interaction of V 2 O 5 and WO 3 active oxides. In addition, the support composition influences the activity. For example, the addition of SiO 2 inhibits the shrinkage of the catalyst's specific surface area, avoiding the phase transition from anatase to rutile and growth of the TiO 2 size.
A Pt doped SiC filter is a promising carrier that not only improves the surface acid sites, but also significantly enlarges the BET surface area, the pore volume of the catalysts, and favors high electron transfer, enhancing the absorption of NH 3 and lighting-off NO conversion at low temperatures.
The presence of SO 2 and/or H 2 O on the flue gases must be carefully addressed for practical applications, paying special attention to the experimental conditions that can affect positively or negatively the catalytic performances of V 2 O 5 -WO 3 based catalysts.
Considering that V 2 O 5 -WO 3 based catalysts are inactive during the engine cold-start, the enhancement of NO SCR at low temperatures (<200 • C), maintaining high selectivity to N 2 , is still challenging for a successful application of such systems for automotive and naval usage. Such improvement could be achieved by doping V 2 O 5 -WO 3 based catalysts with new oxides. According with recent data, MnEuO X , MnSmO X, MnGdOx, MnCeOx represent promising co-catalysts.
Author Contributions: W.Z. and S.Q. prepared the original draft of the manuscript, G.P. and L.F.L. equally contributed to the writing, review and editing.