A Review on Selective Catalytic Reduction of NO_x by NH₃ over Mn–Based Catalysts at Low Temperatures: Catalysts, Mechanisms, Kinetics and DFT Calculations

Abstract

It is a major challenge to develop the low–temperature catalysts (LTC, <250 °C) with excellent efficiency and stability for selective catalytic reduction (SCR) of NO_x by NH₃ from stationary sources. Mn-based LTC have been widely investigated due to its various valence states and excellent redox performance, while the poisoning by H₂O or/and SO₂ is one of the severe weaknesses. This paper reviews the latest research progress on Mn-based catalysts that are expected to break through the resistance, such as modified MnO_x–CeO₂, multi-metal oxides with special crystal or/and shape structures, modified TiO₂ supporter, and novel carbon supporter (ACF, CNTs, GE), etc. The SCR mechanisms and promoting effects of redox cycle are described in detail. The reaction kinetics will be a benefit for the quantitative study of Eley–Rideal (ER) and Langmuir–Hinshelwood (LH) mechanisms. This paper also introduces the applications of quantum-chemical calculation using density functional theory to analyze the physic-chemical properties, explicates the reaction and poisoning mechanisms, and directs the design of functional catalysts on molecule levels. The intensive study of H₂O/SO₂ inhibition effects is by means of the combination analysis of in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and density functional theory (DFT), and the amplification of tolerance mechanisms will be helpful to design an excellent SCR catalyst.

1. Introduction

1.1. NO_x Emissions and Legislations

According to BP’s Energy Outlook (2016), fossil fuels will continue to remain the dominant source of energy powering the world’s economy, supplying 60% of the energy increase out to 2035. It will be increased with oil (+63%), natural gas (+193%) and coal (+5%) accounting for 53% of the growth in demand. Unfortunately, the combustion of fossil fuels inevitably leads to the production of various air pollutants, such as sulfur dioxide, nitrogen oxide (NO_x), particulate matter, heavy metal, volatile organic compounds, etc. [1].

NO_x is a generic term for mono-nitrogen oxides, namely NO and NO₂, which are produced during combustion at high temperatures (above 1350 °C). NO_x (x = 1, 2) emitted to air are largely responsible for the ozone decline in middle to high latitudes from spring to fall, and for the acid rain perturbing the ecosystems and the cause of biological death of lakes and rivers. Peroxyacetylene nitrates (PAN) can also be formed from nitric oxide and contribute significantly to global photo-oxidation pollution [1,2]. In addition, NO_x species are also harmful to the human body, which can diffuse through the alveolar cells and the adjacent capillary vessels of the lungs and damage the alveolar structures and their functions throughout the lungs, provoking both lung infections and respiratory allergies like bronchitis, pneumonia, etc. [1,3].

From 1970 to 2004, America was the largest country for contributing NO_x emissions, followed by Organization for Economic Co-operation and Development (OECD)—Europe regions (see Figure 1A). The NO_x emissions of both countries were decreased by year since the 1990s, which benefited from the legislations and policies of NO_x limitation and technologies of reducing NO_x. For example, the maximum allowable NO_x emission rates of American are set at 0.45 b/mmBtu for tangentially fired boilers, and 0.50 b/mmBtu for dry bottom wall–fired boilers (The Clean Air Act, Sec. 407, USA, 2004). For China, the NO_x emissions were increased by years at that time, which resulted in China becoming the largest contributor in 2008. Since 2011, the NO_x emission of China was also decreased year by year (in Figure 1B), which was attributed to the latest emission standard (GB 13223–2011, China, 2011), in which the emission limits are 100 mg/m³ for new case of coal-fired power plants, 100 mg/m³ for natural gas-fired boilers and 50 mg/m³ for natural gas-fired turbines. The legislations and regulations policies in China for limiting NO_x emissions are getting to be more stringent, such as limiting the ultra-low emission to 50 mg/m³ for thermal-power boilers and the special emission to 150 mg/m³ for sinter, pelletizing and coking furnaces in the following 13th Five Year Plan (2016–2020).

1.2. NO_x Abatement and Demand of LT–SCR Technique

As mentioned above, there is a need to control NO_x emissions due to the adverse environmental and human health effects. The NO_x emissions are mainly from stationary source (combustion, industry and other processes), transportation and miscellaneous source. To reduce NO_x, many control technologies, including combustion control and flue gas denitrification, had been utilized in thermal power plants and industry boilers. Combustion controls are source-control approaches for reducing NO_x generation by altering or modifying the combustion conditions during the combustion process, such as low NO_x burners, fuel re-burning and flue gas recirculation [4]. In spite of this, the NO_x emissions are still too high to meet the emission limits. Selective catalytic reduction (SCR), selective non–catalytic reduction (SNCR), and hybrid SNCR–SCR technologies are the three major post–combustion processes for NO_x abatement [5,6]. Among them, selective catalytic reduction is a method of converting nitrogen oxides, also referred to as NO_x, with the aid of a catalyst into diatomic nitrogen (N₂) and water (H₂O). A gaseous reductant, typically anhydrous ammonia, aqueous ammonia or urea, is added to a stream of flue or exhaust gas and adsorbed onto a catalyst [7]. The NH₃–SCR technique has been widely used for the purification of NO_x from industrial stationary sources due to its outstanding efficiency and strong stability, which can achieve over 90% NO_x conversion [8]. The NH₃–SCR method contains the following reactions: 4NH₃ + 4NO + O₂ → 4N₂ + 6H₂O (standard SCR); 4NH₃ + 2NO₂ + O₂ → 3N₂ + 6H₂O; 4NH₃ + 2NO + 2NO₂ → 4N₂ + 6H₂O (fast SCR) [9,10,11]. In recent years, new De–NO_x technologies, including non–thermal plasma (NTP) [12,13] and pre–oxidation combined absorption [14,15], have been studied in the theoretical stage facing the problems of by–product and energy consumption for NTP, quantitative oxidation and great mass–transfer for pre–oxidation combined absorption, etc.

An overview of the historical development of NH₃–SCR technology and catalysts is available by Smirniotis et al. [3] SCR of NO_x using ammonia was patented in the United States by the Englehard Corporation in 1957. It is well known that the catalyst is the key to the SCR technique. The early SCR catalysts were comprised of platinum or platinum group (noble) metals, such as Co, Ni, Fe and Cr oxides, which were inadequate for the need of a high-temperature range in which explosive ammonium nitrate forms. In the 1960s, the advancements of SCR technology were carried forward in Japan and the United States on the inexpensive and highly durable V₂O₅/TiO₂ catalyst, which demonstrated adequate efficacy at medium temperatures. However, vanadium catalysts have a lack of high thermal durability and a high catalyzing potential to oxidize SO₂ into SO₃, which can be extremely destructive due to its acidic properties. To solve the above shortcomings, mixed oxides of V₂O₅ and WO_x or MoO_x supported on TiO₂ were commonly used as the addition of WO₃ and MoO₃ could improve the thermal stability, durability and hinder the oxidation of SO₂ to SO₃.

At present, the V₂O₅–WO₃(MoO₃)/TiO₂ catalyst has been widely applied in the industry due to the excellent NO_x removal efficiency at a high temperature (300–400 °C) [16,17]. Traditional NH₃–SCR units are operated downstream before the particle removal units. Although the soot-blowers are adopted to clean away the dust from the surface, the catalysts will be inevitably poisoned and deactivated by the toxic material in the dust, such as the alkalis/alkaline–earth metals, phosphorus and heavy metals [16,18]. The following focus of air protection is actively promoting the pollution control of flue gas in steel, cement, glass and other industries. However, the traditional SCR technique is not applicable in these plants due to the burdensome operating costs for heating the low-temperature flue gas (<250 °C). Therefore, it has become a hot topic to explore the development of the catalysts with excellent efficiency and strong stability that can be placed in low-temperature locations behind dust removal and desulfurizer units [4].

1.3. Catalysts for LT–SCR

In recent years, the low-temperature catalysts, including precious metals (Pt, Pd, Rh, etc.) [19,20] and transition metals (Mn, Fe, Ni, Cr, V, Co, Cu, Ce, etc.) [21,22,23,24,25,26,27,28,29], have been widely researched. Precious metal catalysts are usually applied to reducing NO_x from moving sources but are limited in stationary sources due to the shortcomings of high costs, narrow working temperature window, oxygen inhibition effect and sensitivity to SO₂ or other gases. Research on the LT–SCR catalysts in stationary sources mainly focus on the transition metal oxides.

Table 1. Review of catalysts and the catalytic performance in NH₃–SCR of NO_x at low temperatures.

Catalysts	Preparation Method (Cal. Tem./°C)	Reaction Conditions	NOx Conversion/% (Tem./°C)	Ref.
MnOx	Precipitation (350 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 25,000 h−1	100% (75–175 °C)	[33,43]
MnOx	Rheological phase (350 °C); Solid phase (60 °C); co–precipitation (100 °C)	[NO] = [NH3] = 500 ppm, [O2] = 3 vol %, N2 balance, GHSV = 47,000 h−1	~100% (80–150 °C)	[40,41,42]
Mn–Ce–Ox	Citric acid method (650 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 2 vol %, He balance, GHSV = 42,000 h−1	~100% (120–150 °C)	[22]
Mn–FeOx	Co–precipitation (500 °C)	[NO] = [NH3] = 800 ppm, [O2] = 5 vol %, N2 balance, GHSV = 24,000 h−1	~100% (120–300 °C)	[25]
MnO2/TiO2	Impregnating (400 °C)	[NO] = [NH3] = 500 ppm, [O2] = 3 vol %, N2 balance, GHSV = 24,000 h−1	~100% (150–200 °C)	[44]
Mn/ZSM–5	Ion–exchange (300 °C)	[NO] = [NH3] = 600 ppm, [O2] = 4.5 vol %, N2 balanced, GHSV = 36,000 h−1	~100% (150–390 °C)	[45]
VOx/CeO2	Hydrothermal (400 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 120,000 h−1	95–100% (250–350 °C)	[46]
Ce10Mo5/TiO2	Impregnating (500 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, He balance, GHSV = 128,000 h−1	~90% (275–400 °C)	[47]
Ce–Sn–Ox	Co–precipitation (400 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 20,000 h–1	90–100% (200–400 °C)	[48]
CuCe–ZSM–5	Ion–exchange (600 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 10 vol %, He balance, GHSV = 15,000 h−1	~90% (148–427 °C)	[49]
Fe0.95Ce0.05Ox	Co–precipitation (400 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 3 vol %, N2 balanced, GHSV = 30,000 h−1	79–100% (175–300 °C)	[50]
Fe0.5WCeOx	Sol-gel method	[NO] = [NH3] = 450 ppm, [O2] = 2.5 vol %, N2 balanced, GHSV = 20,000 h−1	80% at 160 °C; 95–100% (250–500 °C)	[51]

displays the catalytic performance of mixed metal oxides for low-temperature SCR of NO by NH₃. It can be found that Mn-based catalysts have a superior property for SCR at low temperatures that have become the major research objects. It has been accepted that the variable valence states and excellent redox ability of Mn-based catalysts are favorable to the low-temperature SCR activity [30], which are also significantly affected by the crystal structure, crystallinity, specific surface area, oxidation state, active oxygen, active sites and acidity on the surface, etc. [31,32,33,34,35,36,37,38,39,40,41,42].

Table 2. Parameter comparisons of industrial vanadium-based, best Mn-based and zeolite catalysts.

Parameters	Catalysts
	Industrial V2O5–WO3/TiO2	Novel V2O5–WO3/TiO2	MnxCo3−xO4 Nanocages	Mn–W–TiOx	Cu-SSZ or Cu-SAPO
Operation Temperature	300–400 °C	200–400 °C	150–300 °C	125–275 °C	225–400 °C
SCR activity	≥90%	≥90%	~100%	≥98%	≥85%
Degradation	+	+	++	++	++
Ammonia slippage	≤3 ppm	≤5 ppm	−	−	−
Hazardous	Yes	Yes	No	No	No

Noting: (1) ‘-’ means no data; (2) ‘+’ stands for the degradation rate, ‘++’ for faster rate.

gives a comparison of industrial V₂O₅–WO₃/TiO₂, novel low-temperature V₂O₅–WO₃/TiO₂, best Mn-based and zeolite catalysts, including the parameters of operation temperature, SCR activity, degradation performance, and ammonia slippage. It is noteworthy that the V₂O₅–WO₃/TiO₂ catalyst has been applied successfully in commercial plants with the NO_x removal efficiency above 90%, including the novel vanadium-based catalyst for the removal of NO_x in the low-temperature flue gas (200–300 °C) from coking industries. However, the deposition and blockage effects by semi coking tar and ammonium bisulfate have become the problems associated with these industrial low-temperature SCR catalysts for the post-combustion removal of NO_x during the cold start and low idle conditions. In addition, the V₂O₅–WO₃/TiO₂ catalyst is toxic for the environment and humans and will eventually become hazardous wastes after devitalization. Although the degradation performance is inferior to vanadium-based catalysts, the SCR activity of Mn-based catalysts is quite high such that the NO_x removal efficiency at low-temperature regions of 150 to 300 °C is almost 100%, which is promising for application on NO_x purification in the low-temperature flue gases with low-concentration H₂O and SO₂ such as the flue gases from coking after desulfurization, daily-use glass industries and gas-fired boilers.

Li et al. [52] reviewed metal oxide catalysts, (Fe/Cu) zeolite and reaction mechanism of NH₃–SCR and pointed out that the resistance of H₂O and SO₂ on Mn-based metal oxide is a big challenge for its application at low temperatures. Fu et al. [4] focused on the reaction mechanism and kinetics of supported SCR catalysts at 100–300 °C including Mn-based catalysts reported before 2013, and believed that the exploration of surface chemical reaction process may be a direction to develop low-temperature supported catalysts for removing NO_x. Shan et al. [53] proposed that modified V₂O₅ based catalysts have attracted much attention for applications on removing NO_x from stationary sources while using Cu-containing zeolites for diesel vehicles.

Over the last four years, some Mn-based catalysts are expected to break through the resistance to H₂O and SO₂, such as the modified MnO_x–CeO₂, multi-metal oxides with special crystal or/and shape structures, modified TiO₂ supporter, and novel carbon supporter (ACF, CNTs, GE), etc. Fortunately, the application of quantum chemical calculation gives an insight into the molecular and atomic level by density functional theory (DFT) for studying reaction/poisoning mechanisms and designing catalyst with special construction. Therefore, in this article, the research progress on NH₃–SCR of NO_x over Mn-based catalysts at low temperatures is reviewed. We focus on the above catalysts in the latest reports that present a relatively good resistance to H₂O or/and SO₂. The SCR mechanisms and reaction kinetics are also discussed. In addition, the typical applications of quantum chemistry are introduced and the further research directions of Mn-based catalysts are also proposed.

2. Mn-Based Catalysts

Pure MnO_x catalysts suffer significantly from the inhibition or/and poisoning effects of H₂O and SO₂ in real conditions. To enhance the stability and resistance, Mn-based multiple metal oxides and supported catalysts are widely studied.

2.1. Multi–Metal Oxides

2.1.1. Composite Oxides

Although the composite oxides including Cu–Mn [54,55], Sn–Mn [56], Fe–Mn [57,58,59], Nb–Mn [60], Li–Mn [61], Eu–Mn [62] and Ni-Mn [63] catalysts showed good SCR performances (activity and N₂ selectivity), the resistance to H₂O or/and SO₂ were not performed or were not satisfactory at low temperatures below 250 °C. MnO_x–CeO₂ catalysts have an excellent SCR performance due to the synergistic mechanism between Mn and Ce that enhance the quantity, acidity of acid sites and the ability to store/release oxygen [22,64,65,66,67,68,69,70]. Ralph T. Yang’s team obtained about 95% NO conversion at 120 °C in 4 h on Ce–Mn–O_x catalyst prepared by a citric acid method in the presence of 100 ppm SO₂ and 2.5% H₂O at a gas hourly space velocity (GHSV) of 42,000 h⁻¹ [22]. Up to now, this result has not been repeated by others. Mn–Ce–O_x prepared by Liu et al. [26] using a surfactant-template reached a NO_x conversion above 90% at 150–200 °C in the condition of 5 vol % H₂O and 50 ppm SO₂ at 64,000 h⁻¹. However, the effect of high concentration H₂O and SO₂ on catalytic activity needs to be further researched. Chang et al. [71] found that the SCR activity of Mn(0.4)CeO_x catalyst by the co-precipitation method soon decreased to 18% from 86% at 110 °C after introducing 100 ppm SO₂ at 35,000 h⁻¹. Furthermore, multicomponent MnO_x–CeO₂ by Fe [64], Zr [64], Nb [72], Sn [71,73], W [74] or novel catalysts synthesized by template [26,75] were investigated to further improve the SCR performance and resistance to H₂O or/and SO₂. Chang et al. [71,73] found the NO conversion of Sn(0.1)Mn(0.4)CeO_x could be maintained at ca. 70% at 110 °C after introducing 100 ppm SO₂ and 12 vol % H₂O. They concluded that the sulfation of Sn-modified MnO_x–CeO₂ might form Ce(III) sulfate that could enhance the Lewis acidity and improve NO oxidation to NO₂ during NH₃–SCR at T > 200 °C. Ma et al. [74] reported that W_0.1Mn_0.4Ce_0.5 mixed oxides have an excellent low-temperature activity and N₂ selectivity, which might benefit from the weakened reducibility and increased number of acid sites. They proposed that the decreased SO₂ oxidation activity, as well as the reduced formation of ammonium/manganese sulfates, were the main reasons for the high SO₂ resistance of this catalyst (60% NO_x conversion with 60 ppm SO₂ at 150 °C and 300,000 h⁻¹). In our previous study [76], we found that the NO_x conversion of Ni₁Mn₄O₅ and Co₁Mn₄O₅ can be 78% in the presence of 150 ppm SO₂ at 175 °C, which benefited from the suppression of surface ammonium bisulfate and metal-sulphation, the trifling impacts of SO₂ on the reaction pathways of bidentate nitrate species and the ER mechanisms in which adsorbed NH₃-species were activated and reacted with gaseous NO or NO₂.

2.1.2. Spinel Crystal Catalysts

Recently, the specific structural multi-metal oxides have been investigated to removal NO_x. Zhang et al. [77] found that the NO conversion of perovskite-type BiMnO₃ catalyst decreased from 85.5% to 82.7% during the following 9 h after introducing 5% H₂O and 100 ppm SO₂ and proposed that the more Lewis acid sites and a high concentration of surface oxygen on perovskite-type BiMnO₃ were responsible for its better performance. Perovskite-type NiMnO₃ catalyst prepared by Wan et al. [24] kept the NO conversion at 90% for the following 20 h when 10 vol % H₂O was added, and remained at 82% for 5 h in the case of 100 ppm SO₂ at 230 °C at 64,000 h⁻¹.

Chen et al. [78] proposed that the efficient electron transfer between Cr and Mn in the spinel crystal of CrMn_1.5O₄ was thought to be the reason for a good SCR performance of Cr(0.4)–MnO_x catalyst, which lost 15% of its initial activity in 4 h after introducing 100 ppm SO₂ under 120 °C at 30,000 h⁻¹. The last two years, the Mn_xCo_3−xO₄ spinel (i.e., Mn₂CoO₄ and MnCo₂O₄) catalyst has been widely studied for SCR, which showed an excellent SCR activity, N₂ selectivity and H₂O/SO₂ durability [79,80,81,82,83]. Qiu et al. [80] successfully synthesized the porous bimetallic Mn_xCo_3−xO_x catalysts by the nano–casting method, and found that the NO conversion of MnCo₂O₄ spinel was near 98% with less than 50 ppm N₂O at 150–300 °C. It also showed a high H₂O and SO₂ resistance that the NO_x conversion was stabilized at 86% in 12 h at 200 °C at 50,000 h⁻¹ after introducing 5% H₂O and 100 ppm SO₂. Qiao et al. [79] prepared porous Mn₂Co₁O_x spinel catalyst by a one-step combustion method, and they found that the NO conversion was almost unaffected by 10 vol % H₂O and kept at 92% in the presence of 100 ppm SO₂ at 200 °C at 30,000 h⁻¹. It is accepted that the excellent SCR performance and H₂O or/and SO₂ resistance were mainly attributed to the orderly mesoporous spinel structures with strong interaction of Mn and Co cations, which can provide a larger surface area, abundant active surface oxygen species, and Lewis acid sites for the adsorption and activation of reaction gases [80,81].

It is worth introducing that Zhang et al. [83] synthesized the hollow porous Mn_xCo_3−xO₄ nanocages and nanoparticles as SCR catalysts that derived from Mn₃[Co(CN)₆]₂·nH₂O with a structure of nanocube-like metal-organic frameworks via a self-assemble method. Compared to Mn_xCo_3−xO₄ nanoparticles (seen Figure 2), Mn_xCo_3−xO₄ nanocages presented a much better catalytic activity at low–temperature regions, higher N₂ selectivity, more extensive operating-temperature window and higher H₂O or/and SO₂ tolerance. The NO conversion of Mn_xCo_3−xO₄ nanocages was 99% in the absence of 8 vol %H₂O and decreased slightly after introducing 200 ppm SO₂ at 175 °C, demonstrating that this catalyst was the best catalyst of H₂O or SO₂-resistant, and the inhibitions were reversible. The coexistence of H₂O and SO₂ induced a 10% decrease in the NO conversion and recovered to 96% after cutting off H₂O and SO₂. They concluded that the uniform structure/distribution and strong interaction of Mn and Co provided a larger surface area and more active sites, and enhanced the catalytic cycle. However, for all this, the 5 h of SO₂ and steam poisoning studies are not good enough to understand the tolerance effect. The stability and tolerance to H₂O and SO₂ in the long-term and large-scale of this catalyst need to be further investigated.

Liu et al. [23] had rationally designed and originally developed three-dimensional hierarchical MnO₂@NiCo₂O₄ core-shell nanowire arrays with nickel foam, which took the advantages of the high surface area of Ni-Co nanowires for achieving high loading and dispersion of manganese oxides, as well as the synergistic catalytic effect between Ni, Co and Mn multiple oxides, resulting in an excellent low-temperature catalytic performance and superior H₂O resistance (promoting effect) in a long-term stable operation.

2.1.3. Specific Structure/Shape Catalysts

Metal-organic frameworks (MOFs) are very attractive in separation and catalysis due to the advantages of high porosity, large surface areas, and versatility of their structures, such as MnCe@MOF [84], MIL-100 (Fe-Mn) [85], MOF-74(Mn, Co) [86], etc. Jiang et al. [86] found that, in their reaction process, NO conversion on Mn-MOF-74 decreased with the introduction of 5 vol % H₂O and 100 ppm SO₂ and almost recovered when gas was cut off. However, for Co-MOF-74, SO₂ has almost no effect on the catalytic activity (65% NO conversion) at 220 °C.

Meng et al. [87] found that the novel fluffy structural Co–Mn–O, Fe–Mn–O, Ni–Mn–O catalysts had high degree crystal splitting, high surface areas, abundant surface acid sites and large active surface oxygen, which were essential for the enhancement of their catalytic activities. Novel 3D flower-like NiMnFe mixed oxides prepared by Li et al. [88] using an in situ hydrothermal method showed an excellent catalytic performance, which was attributed to the synergistic effect between Ni, Mn and Fe species. Chen et al. [89] found that the Fe_(0.4)MnO_x catalyst with a crystal phase of Fe₃Mn₃O₈ showed excellent SCR performances (98.8% NO_x conversion with 100% N₂ selectivity) and good tolerance to 5% H₂O and 100 ppm SO₂ (a 13% decrease in the NO_x conversion in 4 h) at 120 °C at 30,000 h⁻¹. MnO₂ doped Fe₂O₃ hollow nanofibers were prepared using an electrospinning method by Zhan et al. [90], which reached an NO conversion of 82% in 9 h at 150 °C after introducing 8 vol % H₂O and 200 ppm SO₂.

It is noteworthy that Li et al. [91] synthesized the Mn₂O₃-doped Fe₂O₃ hexagonal microsheets, which obtained a 98% NO conversion at 200 °C at 30,000 h⁻¹. Meanwhile, good tolerance to 15% H₂O or 100 ppm SO₂ were obtained, maintaining the NO conversion at 92% and 85% in 100 h, respectively. It suggests that coupled Mn₂O₃ nanocrystals played a key role and produced a possible redox mechanism in the low-temperature SCR process.

2.2. Metal Oxides as the Carrier

It has been reported that the SCR activity was followed in the order: MnO_x/TiO₂ > MnO_x/γ-Al₂O₃ > MnO_x/SiO₂ > MnO_x/Y-ZrO₂ [92]. Moreover, the higher surface MnO₂ concentration it is, the higher the Lewis acidity will be. The redox properties of TiO₂-supported MnO_x catalyst were important factors in achieving better DeNO_x performance at low temperatures [93]. Kang et al. [92] concluded the deactivation caused by H₂O and/or SO₂ was significantly affected by these kind of supports. They found that TiO₂ contributed to the resistance to SO₂, but the NO_x conversion also decreased to 68% in 4 h after introducing 100 ppm SO₂ and 10 vol % H₂O at 150 °C. Smirniotis et al. [93] compared TiO₂-supported, Al₂O₃-supported and SiO₂-supported manganese oxide catalysts for the low-temperature SCR reaction and indicated the presence of primarily Lewis acid sites in Mn/TiO₂. However, there is only Brönsted acidity in Al₂O₃-supported and SiO₂-supported catalysts. Wu et al. [94] reported that SO₂ has an obvious poisoning-effect on SCR activity of Mn/TiO₂, which was decreased from 93% to 30% in the presence of 100 ppm SO₂ for 6.5 h at 150 °C. TiO₂ has been widely used as the support of Mn-based catalysts due to its stable physical and chemical properties. The SCR activity, selectivity and resistance are greatly affected by the precursors, preparation method, amorphous phase, surface dispersity and acidity, and the interaction between Mn and Ti, etc. [44,95,96,97,98].

For enhancing the resistance to H₂O and SO₂, the composite-oxides, such as Mn confined titania nanotubes (Mn/TNT) [99], M-Mn/TiO₂ (M = Fe [100,101,102,103,104], Co [105,106,107], Ce [105,108], Ni [109], W [110] etc.), Fe–Mn/Ti–Zr [111], MnO_x/Fe–Ti [112], MnO_x/CeO₂–TiO₂ [113], Mn–Ce/PG (palygorskite) [114], MnO_x/Ce_0.8M_0.2O₂ (M = Ti, Sn) [115], MnO_x(0.6)/Ce_0.5Zr_0.5O₂, [116] MnO₂/Ce_(1−x)Zr_xO₂–TiO₂, [117] MnO₂–Ce_(1−x)Zr_xO₂/TiO₂ [118], SO₄²⁻–Mn–Co–Ce/Ti–Si [119], etc., were investigated. Among these reports, Zhu et al. [104] reported that Fe_0.3Ho_0.1Mn_0.4/TiO₂ catalyst presented a broad operating temperature window in the range of 60–200 °C and exhibited superior sulfur-poisoning resistance with 80% NO_x conversion in the presence of 200 ppm SO₂ and 15% H₂O at 120 °C. Shen et al. [116] found that MnO_x(0.6)/Ce_0.5Zr_0.5O₂ displayed a good resistance to 3% H₂O and 100 ppm SO₂ that remained 90% NO conversion at 180 °C over the course of 5 h. Qiu et al. [119] found that the SO₄²⁻–Mn–Co–Ce/Ti–Si catalyst has an excellent SO₂ durability at low temperatures. However, these catalysts show a relatively good resistance for a short time. Because of its pre-sulfating of dopants for protecting MnO_x, the downward trend of NO_x conversion makes it difficult to determine whether the activity will be reduced or not as the reaction time goes on. Huang et al. [107] prepared MnO₂–Co₃O₄ decorated TiO₂ nanorods heterostructures (MnO₂-(Co₃O₄)/TiO₂ hybrids) and found that the SCR activity of this catalyst was barely affected by 8 vol % H₂O and exhibited an over 90% NO conversion at 150 °C during the test.

Smirniotis et al. [21,109,120,121,122,123] found that Mn/TiO₂ anatase (Hombikat), Mn-Ni(0.4)/TiO₂ and Mn-Ce(5.1)/TiO₂ catalysts, with external surface manganese oxide clusters, promoted Mn⁴⁺, nano-size metal oxide crystallites, and surface labile oxygen species seemed to be the reasons for high de-NO_x efficiency at low temperatures. These catalysts also demonstrated impressive catalytic activity by exhibiting stable conversions within hundreds of hours to 7 and 10 vol % H₂O vapors at 175 and 200 °C. The addition of H₂O in the feed gas does not seem to create an eternal alteration to the surface manganese species most probably due to the existence of high redox potential pairs of Mn and Ni or Ce in the structure, which represent an efficient active catalyst. Furthermore, Smirniotis et al. [99,122] designed Mn loaded elevated surface texture hydrated titania and MnO_x confined interweaved titania nanotubes to improve SCR activity in the temperature regime of 100–300 °C. These studies indicated that the novel titania nanotubes prepared by alkaline hydrothermal treatment could provide enormous surface area and a unique nano-tubular structure, which can be beneficial for the higher dispersion of the active species and greatly promote the deNO_x potential of MnO_x-based formulations with less pronounced coagulation under the tested reaction temperatures.

Deng et al. [102] investigated MnO_x/TiO₂ (anatase) nano-sheets (NS) with a preferentially exposed (001) facet, in which it was possible to enhance the low-temperature SCR activity of the catalysts by tailoring the preferentially exposed facet of TiO₂. It is worthwhile to note that Wang et al. [110] reported that the W(0.25)–Mn(0.25)–Ti(0.5) catalyst presented an obvious improvement in the sulfur-resistant, obtaining a NO_x conversion of 100% from 140 °C to 260 °C in the presence of 100 ppm SO₂ and 10 vol % H₂O, which may be favorable in practical application. Guo et al. [124] found that the modification of TiO₂ support by W would lead to lower crystallinity, higher reducibility and surface acidity, accompanied by the presence of more chemisorbed oxygen over its surface and resulting in promoting the NH₃–SCR reaction over the Mn/TiWO_x catalyst.

2.3. Molecular Sieves as the Carrier

According to the literature, many ion (Fe, Cu, Ce, Mn) exchanged zeolites were reported to be active in NH₃–SCR reaction. Among them, Fe, Cu, Ce exchanged ZSM-5, SAPO, CHA, SSZ, SBA seem to be particularly interesting and have been extensively studied to remove NO_x from moving sources. In the early years, the MnO_x/NaY [125], M–Mn/USY (Fe, Ce) [126,127], M–Mn/ZSM–5 (Fe, Ce) [128,129] and Mn/SAPO-34 [130] were reported for SCR. The NO_x conversion of Mn–Ce/ZSM–5 by Carja et al. [128] reached above 75% in a broader temperature range (244–549 °C) with a GHSV of 332,000 h⁻¹ and the catalytic activity was stable at 275 °C even in the presence of H₂O and SO₂. Lin et al. [127] reported that the NO conversion of a 10%Mn–8%Fe/USY catalyst began to decline rapidly and eventually stabilized at around 60% 180 °C as 10% H₂O and 100 ppm SO₂ were added to the reaction gas. Qi et al. [126] found that the NO conversion on 14%Ce–6%Mn/USY decreased slowly to 80% from 98% at 150 °C in 4 h. However, regrettably, the NO conversion showed a downward trend when the SO₂ and H₂O were turning off. Yu et al. [131] prepared the MnO_x/SAPO-34 by conventional impregnation and an improved molecularly designed dispersion method for the low-temperature SCR, which obtained the 90% NO_x conversion at above 200 °C. Furthermore, the resistance to SO₂ and H₂O was not ideal. After nearly two years, attention on Mn–based molecular sieves for SCR in stationary sectors is not adequate, which may be due to the relatively high-temperature windows and unsatisfactory resistance to H₂O or/and SO₂.

2.4. Carbon Materials as the Carrier

Carbon materials including active carbon (AC), active carbon fiber (ACF) and carbon nanotubes (CNTs), and graphene (GE) have been widely utilized in air purification and separation due to its large number of well-distributed microspores, high adsorption speed and high surface area. Catalysts such as MnO_x/AC [132,133,134], MnO_x/AC/C [135], Mn–Ce/ACH (honeycomb) [136,137], Mn–Ce/ACF [138] and triple-shell hollow structured CeO₂-MnO_x/carbon-spheres [139] on had been reported for SCR, but they were significantly suffered from the poisoning of H₂O and SO₂.

2.4.1. Carbon Nanotubes (CNTs)

Reports have shown that CNTs were a good carrier to support active metal oxides for SCR reactions, such as MnO_x/MWCNTs (multi-walled) [140,141], Mn–Ce/CNTs [142,143,144,145], Mn–Fe/CNTs [146], MnO₂–Fe₂O₃–CeO₂–Ce₂O₃/CNT [147], Mn–Ce/TiO₂–CNT [148], etc. More importantly, the CNT-based catalysts presented good capacity for H₂O and SO₂ resistance. Pourkhalil et al. [141] found that the NO_x conversion over MnO_x/FMWNTs decreased from 97% to 92% within 6 h after adding 100 ppm SO₂ and 2.5 vol % H₂O at 200 °C at 30,000 h⁻¹. Zhang et al. [144] reported that the MnCe/CNTs catalyst has a stable NO conversion of 78% with 100 ppm SO₂ and 4 vol % H₂O at 300 °C. Furthermore, Zhang et al. [149] demonstrated a core-shell structural de–NO_x catalyst with a high SO₂-tolerance and stability, which was that CNT-supported MnO_x and CeO_x NPs coated with mesoporous TiO₂ sheaths (mesoTiO₂@MnCe/CNTs). In this design, the meso–TiO₂ sheaths could prevent the generation of ammonium/manganese sulfate species from blocking the active sites. It was indicated that CNT supported Mn–based catalysts can be investigated to achieve higher stability and better activity in the presence of SO₂ and H₂O at low temperatures. Fang et al. [150] reported the nano-flaky MnO_x on carbon nanotubes (nf-MnO_x@CNTs) synthesized by a facile chemical bath deposition route that has an excellent performance in the low-temperature SCR of NO to N₂ with NH₃ and also presented a favorable stability and H₂O resistance. Cai et al. [151] further synthesized a multi-shell Fe₂O₃@MnO_x@CNTs. They found that the formation of a multi-shell structure induced the enhancement of the active oxygen species, reducible species as well as adsorption of the reactants, which brought about excellent de-NO_x performance. Moreover, the Fe₂O₃ shell could effectively suppress the formation of the surface sulfate species, leading to the desirable SO₂ resistance to the multi-shell catalyst. Hence, the synthesis protocol could provide guidance for the preparation and elevation of Mn-based catalysts.

2.4.2. Graphene (GE)

Graphene, a new carbon nanomaterial, has the exceptional properties of a large theoretical specific surface area, flexible structure, high electron mobility and excellent conductivity. Thus, it was used as a carrier for SCR to promote the electron transfer between various Mn cations because of the efficient electron gain and loss in GE bringing the improvement in the redox performance [152,153,154,155]. Xiao et al. [152] found that MnO_x–CeO₂/Graphene (0.3 wt %) was an environmentally-benign catalyst for controlling NO_x, which exhibited excellent NH₃–SCR activity and strong resistance against H₂O and SO₂. We found that the MnO_x/TiO₂–GE catalyst showed high NH₃–SCR activity and N₂ selectivity, and good stability during low-temperature SCR at a high GHSV of 67,000 h⁻¹ [153]. To improve the SCR activity and resistance, we further investigated a series of Ce-Mn/TiO₂-GE catalysts by the sol-gel and ultrasonic impregnation methods [154]. It was found that the 7 wt % (Ce(0.3)–MnO_x)/TiO₂–0.8 wt % GE catalyst exhibited an improved resistance to H₂O and SO₂ at 180 °C comparing with the markedly decreased NO_x conversion within the first 90 min but maintained at about 80% after adding 10 vol % H₂O and 200 ppm SO₂ to the system. In particular, hydrophobic groups on the surface of GE were not conducive to H₂O adsorption. Although this catalyst displays a relatively good stability, H₂O and SO₂ tolerance in our small-scale lab, the long-term and large-scale sulfur-tolerance properties of the catalyst need to be further investigated.

In recent years, as shown in

Table 3. Research results of SO₂ and H₂O resistance over kinds of manganese based catalysts in literature.

Catalysts	Preparation Method	Reaction Conditions	Original/Deactivated/Recovered Activity (Loss of Activity)	Refs.
H2O
MnOx	Low-temperature solid phase	[NO] = [NH3] = 500 ppm, [O2] = 3 vol %, N2 balance, GHSV = 47,000 h−1, 10% H2O, at 80 °C	98%/87%/96% (11.2%)	[40]
MnO2	Hydrothermal (400 °C)	[NO] = [NH3] = 500 ppm, [O2] = 3 vol %, N2 balance, 10% H2O, at 200 °C	92%/70%/- (23.9%)	[38]
Mn–Ce–Ox	CTAB template (500 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, He balance, GHSV = 64,000 h−1, 5% H2O, at 100 °C	100%/47%/- (53.0%)	[26]
Mn–Ni–Ox	Co-precipitation (400 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 64,000 h−1, 10% H2O, at 230 °C	100%/94%/- (6.0%)	[24]
Mn–Co–Ox	Metal−organic frameworks template (450 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 38,000 h−1, 8% H2O, at 175 °C	99%/98%/99% (slightly)	[83]
SO2
MnOx	Low-temperature solid phase	[NO] = [NH3] = 500 ppm, [O2] = 3 vol %, N2 balance, GHSV = 47,000 h−1, 100 ppm SO2, at 80 °C	98%/25%/- (74.5%)	Our work
MnOx		[NO] = [NH3] = 1000 ppm, [O2] = 3 vol %, N2 balance, GHSV = 30,000 h−1, 100 ppm SO2, at 120 °C	80%/10%/50% (87.5%)	[19]
Mn–Cr–Ox	Citric acid (650 °C)		99%/84%/96% (15.2%)	[89]
Mn–Ni–Ox	Co-precipitation (400 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 64,000 h−1, 100 ppm SO2, at 230 °C	100%/82%/98% (18.0%)	[24]
Mn–Co–Ox	Metal−organic frameworks template (450 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 38,000 h−1, 100 ppm SO2, at 175 °C	decreases slightly	[83]
Mn–Ce–Sn–Ox	Co-precipitation (500 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 2 vol %, N2 balance, GHSV = 35,000 h−1, 100 ppm SO2, at 250 °C	100%/96%/-	[71,159]
Mn–Ce–Ni–Ox	Co-precipitation (500 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 48,000 h−1, 150 ppm SO2, at 175 °C	91%/78%/88% (14.3%)	Our work [76]
Mn–Ce–Co–Ox			90%/76%/88% (15.6%)
H2O+SO2
Mn–Cu–Ox	Co-precipitation (350 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 30,000 h−1, 11% H2O + 100 ppm SO2, at 125 °C	95%/64%/90% (32.6%)	[54]
Mn–Ce–Ox	Co-precipitation (500 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 2 vol %, He balance, GHSV = 42,000 h−1, 2.5% H2O + 100 ppm SO2, at 150 °C	94%/83%/94% (11.7%)	[64]
Mn–Ce–Ox	CTAB template (500 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, He balance, GHSV = 64,000 h−1, 5% H2O + 50 ppm SO2, at 100 °C	100%/35%/- (65.1%)	[26]
Mn–Ce–Sn–Ox	Co-precipitation (500 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 2 vol %, N2 balance, GHSV = 35,000 h−1, 12% H2O + 100 ppm SO2, at 100 °C	100%/70%/90% (30.0%)	[71]
Mn–Co–Ox	Metal−organic frameworks template (450 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 38,000 h−1, 8% H2O + 100 ppm SO2, at 175 °C	99%/89%96% (10.1%)	[83]
Mn–Co–Ox	KIT-6 template (450 °C)	[NO] = [NH3] = 500 ppm, [O2] = 5 vol %, N2 balance, GHSV = 50,000 h−1, 5% H2O + 100 ppm SO2, at 200 °C	100%/86%/93% (14.1%)	[80]
Fe0.3Ho0.1Mn0.4 /TiO2	Impregnation (450 °C)	[NO] = [NH3] = 800 ppm, [O2] = 5 vol %, N2 balance, GHSV = 50,000 h−1, 15% H2O + 200 ppm SO2, at 120 °C	95%/80%/85% (15.8%)	[104]
Mn(0.25)–W(0.25)–TiO2(0.5)	One-pot co-precipitation (400 °C)	[NO] = [NH3] = 1000 ppm, [O2] = 5 vol %, He balance, GHSV = 100,000 h−1, 10% H2O + 100 ppm SO2, at 125–250 °C	100%/98%/- (slightly)	[110]

, elements doped composite (active component/support) catalyst, which is in order to protect the active component by pre-sulfurization, has been extensively studied and achieved good low-temperature SCR activity, as well as a certain effect on water resistance and sulfur resistance, such as Sn [71,73], Cr [78,156], Ni [24,76], Fe [154,157], Ce [22,64,66], and Co [21,79,83]. In addition, many synthesis methods have been widely used in the preparation of non-supported Mn-based catalysts such as co-precipitation method [54,64], citrate method [78], redox method [40,56], hydrolysis method [158], combustion method [68,158], and (inorganic/organic) template synthesis method [26,80,83]. These catalysts had achieved good low-temperature SCR activity and a certain effect on water resistance and sulfur resistance. Currently, the catalysts with high H₂O and SO₂ resistant capacity in the reports are better prepared by unconventional methods, such as homogeneous MnO_x–CeO₂ pellets prepared by a one-step hydrolysis process [158], Sn–MnO_x–CeO₂ catalyst prepared by ultrasound-assisted co-precipitation method [71,73], CrMn_1.5O₄ catalyst with spinel structure prepared by citrate method [78], and Mn_xCo_3−xO₄ catalyst prepared by the template synthesis method [79,80,83]. It can be learned that the synthesis method has a great impact on the structure of the catalyst, while the structure of the catalyst is closely related to the physical and chemical properties of the catalyst. Therefore, regarding the synthesis method and the structure of the catalyst as starting points, it is one of the research directions of low-temperature SCR catalyst—that of preparing special structured Mn-based catalyst with high low-temperature SCR activity and high H₂O and SO₂ resistant capacity.

3. SCR Mechanisms

3.1. Adsorption Behavior of Reactants

It is well-known that the adsorption and activation behaviors of gaseous on the surface of catalyst are the important processes in the gas-solid phase catalytic reaction [160]. In particular, it is difficult for gaseous NO to be adsorbed without O₂ due to the non-activated sites. The adsorption capacity of NO with O₂ has a qualitative increase, resulting in the final form of nitrite, nitrate, and NO₂-congaing species. Kijlstra et al. [161] investigated the co-adsorption behaviors of NO with O₂ over MnOx/Al₂O₃ catalyst using the Fourier Transform infrared spectroscopy (FTIR) technique and summarized the thermal stabilities of various NO_x-absorbed species in the following order: nitrosyl < linear nitrites, monodentate nitrites, bridged nitrites < bridged nitrates < bidentate nitrates, as shown in

Table 4. Thermal stabilities of various NO_x-absorbed species.

Species	Split V3 (cm–1)	Corresponding V3 (cm−1)	Form	Desorption/Decomposition
Nitrosyl	1835		${\mathrm{M}}^{\mathrm{n}+}–{\mathrm{N}=\mathrm{O}}^{\delta -}$	50 °C
Bridged nitrate	1620	1220	$\begin{array}{l}{\mathrm{M}}^{\mathrm{n}+}{-\mathrm{O}}_{{\ddots }_{\mathrm{N}-\mathrm{O}}}\\ {\mathrm{M}}^{\mathrm{n}+}{-\mathrm{O}}^{⋰}\end{array}$	150–300 °C
Bidentate nitrate	1290	1555	${\mathrm{M}}^{\mathrm{n}+}{}_{{\ddots }_{\mathrm{O}}}^{{⋰}^{\mathrm{O}}}{}_{⋰}{}^{\ddots }\mathrm{N}-\mathrm{O}$	300–425 °C
Linear nitrite	1466	1075 (V1)	${\mathrm{M}}^{\mathrm{n}+}–\mathrm{O}–\mathrm{N}=\mathrm{O}$	50–200 °C
Monodentate nitrite	1415	1322 (V1)	${\mathrm{M}}^{\mathrm{n}+}-{\mathrm{N}}_{{\ddots }_{\mathrm{O}}}^{{⋰}^{\mathrm{O}}}$	50–200 °C
Bridged nitrite		1230	$\begin{array}{l}{\mathrm{M}}^{\mathrm{n}+}{-\mathrm{O}}_{{\ddots }_{\mathrm{N}}}\\ {\mathrm{M}}^{\mathrm{n}+}{-\mathrm{O}}^{⋰}\end{array}$	50–250 °C

. Here, the temperature for thermal stability of absorbed species was defined by the temperature range as one’s infrared spectra reduced with in situ increasing of temperature. The author found that, as for NO adsorption, nitrosylic species formed at Mn³⁺ sites in the absence of O₂ were very unstable in its presence, which were no active reactive intermediates over the catalyst. In contrast, bidentate nitrates were formed during heating at the expense of less stable nitrites and nitrates. Wherein the nitrites with lower thermal stability can sustainably react with adsorbed NH₃, it is conducive to low-temperature activity of the catalyst while the nitrates with higher thermal stability will be sustained and steady on the surface of catalysts resulting in the activity decline.

Qi et al. [64,66] proposed the NO adsorption over MnO_x–CeO₂ catalyst, which is in the form of initiative anionic nitrosyl (NO⁻) to nitrite or nitrate species, whose transformation rate on MnO_x was much faster than that on CeO₂. Compared to the nitrate and nitrite, NO₂ was easy to desorb from the catalyst. The nitrite and nitrate transformed into more stable forms (>80 °C) and weakened at the temperature >150 °C.

Ramis et al. [162] found that NH_2(ads) formed by H-abstraction of coordinated NH₃ on Lewis acid sites was the intermediate in both SCR reaction and NH₃ oxidation to N₂, and –NH_(ads) and –N_(ads) formed by depth sequential oxidation of NH_2(ads) were the intermediates to N₂O and NO. Furthermore, Kijlstra et al. [31] found gaseous NH₃ can be absorbed on Brønsted acid sites (Mn–O–H) to NH₄⁺ ions and Lewis acid sites (Mn³⁺) to coordinated NH₃ for the subsequent transformation to NH₂ species, and O₂ appeared to be important in assisting in the formation of reactive NO_x–absorbed species and for H-abstraction from adsorbed NH₃.

3.2. Reaction Pathways

The adsorption and further H-abstraction of NH₃ (sequential oxidation) are regarded as the key factors in the NH₃–SCR reaction. Ramis et al. [162] proposed various reaction pathways of NH₃ at NO + O₂ atmosphere, as shown in Figure 3. Here, no more detailed description is needed. In the NH₃–SCR system, it is accepted that the function of both Lewis and Brönsted acid sites are affected significantly by temperature, and that NH₄⁺ adsorbed on Brönsted acid sites play a major role in NH₃–SCR reaction at higher temperature, while coordinated NH₃ linked to Lewis acid sites play a decisive role at a slightly lower temperature [163].

It has become a consensus that there are two mechanisms of low-temperature NH₃–SCR reaction [163]. One is the Eley–Rideal (ER) mechanism that the gaseous NO first reacts with activated NH₃-absorbed species to form intermediates and then decomposes into N₂ and H₂O. Another is the Langmuir–Hinshelwood (LH) mechanism that the gaseous NO are absorbed on basic sites and further combine with the adjacent activated NH₃ species to form N₂ and H₂O. Qi et al. [64,66] studied the NH₃–SCR mechanism on the MnO_x–CeO₂ catalyst and proposed that a gaseous NH₃ molecule was first adsorbed to form activated NH₃ (coordinated NH₃ and –NH₂). The ER mechanism was considered as a typical mechanism that the –NH₂ species reacted with gaseous NO to generate nitrosamine (NH₂NO) and ultimately decomposed into N₂ and H₂O. In addition, the LH mechanism was also observed in which the absorbed NO was oxidized to NO₂ and HNO₂, which further reacted with coordinated NH₃ to form NH₄NO₂ and ultimately decomposed to N₂ and H₂O. Many researchers [26,64,66,83,112,164,165] also confirmed that the coexistence of the LH mechanism and the ER mechanism are prevalent in the NH₃–SCR reaction at low temperatures.

3.2.1. Langmuir–Hinshelwood Mechanism

The LH mechanism can be approximately expressed as: gaseous NH₃ and NO are absorbed on the surface through Reactions (1) and (2), respectively. It is generally agreed that the SCR reaction starts with the adsorption of NH₃. Physically adsorbed NO can be oxidized by Mⁿ⁺ to form nitrite and nitrate through Reactions (3) and (4), respectively. Furthermore, these species react with adsorbed NH₃ to form NH₄NO₂ and NH₄NO₃ by Reactions (5) and (6) then decomposed to N₂ and N₂O, respectively. Finally, the reduced Mⁿ⁺ can be rapidly regenerated by O₂ (i.e., Reaction (7)):

$${\mathrm{NH}}_3{}_{\left(\mathrm{gas}\right)}\to {\mathrm{NH}}_3{}_{\left(\mathrm{ads}\right)},$$

(1)

$${\mathrm{NO}}_{\left(\mathrm{gas}\right)}\to {\mathrm{NO}}_{\left(\mathrm{ads}\right)},$$

(2)

$${\mathrm{M}}^{\mathrm{n}+}{=\mathrm{O}+\mathrm{NO}}_{\left(\mathrm{ads}\right)}\to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–\mathrm{NO},$$

(3)

$${\mathrm{M}}^{\mathrm{n}+}{=\mathrm{O}+\mathrm{NO}}_{\left(\mathrm{gas}\right)}+\frac{1}{2}{\mathrm{O}}_2\to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2,$$

(4)

$$\begin{array}{ll}{\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}+\mathrm{NH}}_{3(\mathrm{ads})}& \to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–\mathrm{NO}–{\mathrm{NH}}_3\\ & \to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{H}+\mathrm{N}}_2{+\mathrm{H}}_2\mathrm{O},\end{array}$$

(5)

$$\begin{array}{ll}{\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{+\mathrm{NH}}_{3(\mathrm{ads})}& \to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2–{\mathrm{NH}}_3\\ & \to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{H}+\mathrm{N}}_2{\mathrm{O}+\mathrm{H}}_2\mathrm{O},\end{array}$$

(6)

$${\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–\mathrm{H}+\frac{1}{4}{\mathrm{O}}_2\to {\mathrm{M}}^{\mathrm{n}+}=\mathrm{O}+\frac{1}{2}{\mathrm{H}}_2\mathrm{O}.$$

(7)

3.2.2. Eley–Rideal Mechanism

The SCR reaction through the ER mechanism can be approximately described as that the adsorbed NH₃ can be activated to NH₂ (through Reaction (8)) and subsequently oxidized to NH by Mⁿ⁺. Generally, the reaction products of gaseous NO with NH₂ and NH (i.e., Reactions (10) and (11)) are N₂ and N₂O, respectively:

$${\mathrm{NH}}_3{}_{\left(\mathrm{gas}\right)}\to {\mathrm{NH}}_3{}_{\left(\mathrm{ads}\right)},$$

(1)

$${\mathrm{NH}}_3{}_{\left(\mathrm{ads}\right)}{+\mathrm{M}}^{\mathrm{n}+}=\mathrm{O}\to {\mathrm{NH}}_2{}_{\left(\mathrm{ads}\right)}+{\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–\mathrm{H},$$

(8)

$${\mathrm{NH}}_2{}_{\left(\mathrm{ads}\right)}{+\mathrm{M}}^{\mathrm{n}+}=\mathrm{O}\to \mathrm{NH}{}_{\left(\mathrm{ads}\right)}+{\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–\mathrm{H},$$

(9)

$${\mathrm{NH}}_2{}_{\left(\mathrm{ads}\right)}{+\mathrm{NO}}_{(\mathrm{gas})}\to {\mathrm{N}}_2{+\mathrm{H}}_2\mathrm{O},$$

(10)

$${\mathrm{NH}}_{\left(\mathrm{ads}\right)}{+\mathrm{NO}}_{(\mathrm{gas})}\to {\mathrm{N}}_2{\mathrm{O}+\mathrm{H}}^{+}.$$

(11)

Afterwards, the dual LH–ER mechanism was proposed by Chen et al. [166] on Ti_0.9Mn_0.05Fe_0.05O_2−δ catalyst. Monodentate nitrate M⁽ⁿ⁻¹⁾–O–NO₂ is involved in the reaction with adsorbed NH₄⁺ (i.e., Reaction (12)) or NH₃ on neighboring acid sites to produce an active intermediate M⁽ⁿ⁻¹⁾–O–NO₂[NH₄⁺]₂ (i.e., Reaction (13)) or M⁽ⁿ⁻¹⁾–O–NO₂[NH₃]₂ (i.e., Reaction (14)). The adsorbed NO_x–NH₃ complex finally reacts with gaseous NO to form N₂ and H₂O through Reactions (15) and (16). Furthermore, Hadjiivanov et al. [167] and Jiang et al. [168] found that bidentate nitrate could be transformed to monodentate nitrates, producing new Brønsted acid sites for more NH₄⁺. Long et al. [169] and Liu et al. [170] also reported that monodentate nitrate species could be converted to the intermediate complex with coordinated NH₃, which further decomposed or reacted with gas NO via ER mechanism to N₂. The latter can be considered as the association mechanisms of ER and LH [76]:

$${\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{H}+\mathrm{NH}}_3{}_{\left(\mathrm{gas}\right)}\to {\mathrm{M}}^{(\mathrm{n}-1)+}–{\mathrm{O}+\mathrm{NH}}_4{}^{+},$$

(12)

$${\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{+2\mathrm{NH}}_4{}^{+}\to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{[{\mathrm{NH}}_4{}^{+}]}_2,$$

(13)

$${\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{+2\mathrm{NH}}_{3\left(\mathrm{gas}\right)}\to {\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{[{\mathrm{NH}}_3]}_2,$$

(14)

$${\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{[{\mathrm{NH}}_4{}^{+}]}_2{+\mathrm{NO}}_{\left(\mathrm{gas}\right)}\to {\mathrm{M}}^{(\mathrm{n}-1)+}–{\mathrm{O}+2\mathrm{N}}_2{+3\mathrm{H}}_2{\mathrm{O}+2\mathrm{H}}^{+}$$

(15)

$${\mathrm{M}}^{(\mathrm{n}-1)+}–\mathrm{O}–{\mathrm{NO}}_2{[{\mathrm{NH}}_3]}_2{+\mathrm{NO}}_{\left(\mathrm{gas}\right)}\to {\mathrm{M}}^{(\mathrm{n}-1)+}{\text{–}\mathrm{O}+2\mathrm{N}}_2{+3\mathrm{H}}_2\mathrm{O}.$$

(16)

The better and potential performances of TiO₂ supported metal oxides in the NH₃–SCR reaction were investigated in-depth by the group of Smirniotis from the early 2000s [21,109,120,121,122,123]. They gave a clear picture of the SCR reaction pathways drawn by exploring the surface interactions of isotopic-labeled reactants, using transient response analysis and in situ FTIR studies, as shown in Figure 4 [171]. As for the surface of MnO₂/TiO₂, they proposed that the reactions followed a Mrks-van-Krevelen-like mechanism through the formation of nitrosamide and azoxy intermediates, in which the lattice oxygen has a direct effect on the mechanism for the instantaneous oxidation of NH₃, and the products and active intermediates were mainly due to a coupling of one nitrogen atom from ammonia and another one from nitric oxide (2¹⁵NO + 2NH₃ + 0.5O₂ → 2¹⁴N-¹⁵N +3H₂O).

3.2.3. Effects of H₂O and SO₂

Kijlstra et al. [31,32] found that H₂O has the effects of reversible physical and irreversible chemical competitive adsorption with NO and NH₃. Xiong et al. [172] proposed that the effects of H₂O were not only attributed to the competition adsorption, but also to the decrease of oxidation ability and the inhibition of interface reaction over the Mn-Fe spinel.

Park et al. [173] found that SO₂ was easily oxidized to SO₃ and eventually formed (NH₄)₂SO₄ and NH₄HSO₄ or other ammonium nitrates via the reaction with NH₃ and H₂O at low temperatures (<200 °C). Kijlstra et al. [32] proposed that the deactivation of Mn-based catalysts by SO₂ was mainly due to the formation of MnSO₄ on the surface. Many reports have demonstrated the above arguments [31,32,40,157,174].

In our study, we found that the adsorption of NO_x to monodentate nitrite species over Mn-based catalysts were significantly inhibited by SO₂, while the bidentate nitrate species had no influences from SO₂, which further transformed to monodentate nitrate producing new Brønsted acid sites for adsorbing NH⁴⁺ [76]. Liu et al. [157,174] have made a careful investigation of the SCR mechanism and H₂O/SO₂ inhibition effects over the Fe-Mn-TiO₂ catalyst using in situ DRIFTS. They found the impacts of H₂O on the adsorption of NH₃ and NO gaseous were moderate and reversible such that SO₂ significantly reduced the adsorption capacity of NH₃, suppressing the formation of nitrate/nitrite species (NH₄NO₃ and NH₄NO₂), which were the active intermediates in the SCR reaction. In addition, the accumulative deposition of sulfate species on the surface further interrupted the Langmuir–Hinshelwood reaction pathway, resulting in an unrecoverable reduction of SCR activity.

3.3. By–Product of N₂O

$${4\mathrm{NO}+4\mathrm{NH}}_3{+3\mathrm{O}}_2\to {4\mathrm{N}}_2{\mathrm{O}+6\mathrm{H}}_2\mathrm{O}.$$

(17)

N₂O mainly originates from the reaction of NO and NH₃ (Reaction (17)) in which gaseous NO react with adsorbed NH_x and N species to give N₂ and N₂O via the ER mechanism [165,175]. The postulation that one nitrogen atom of N₂O comes from NO and another from NH₃ has been directly substantiated by the isotopic labeling experiments by Singoredio et al. [176] and Janssen et al. [177] Ramis et al. [162], Tang et al. [165] and Yang et al. [164] believed that the depth dehydrogenation oxidation of NH₃ to the form of NH₂ intermediate could react with gaseous NO to form N₂, while NH or N species could react with NO to only give N₂O. It has been proven that gaseous NH₃ molecules could be adsorbed to the active Mn sites through coordination bond of Mn-NH₃ without cleavage of N–H bonds (Reactions (8) and (9)) [175], while adsorbing dissociatively on surface active oxygen species successively strip all hydrogen atoms to form adsorbed nitrogen atom species, which are responsible for the generations of N₂, N₂O and NO [178]. Thus, the selectivity to N₂O, predominantly derived from the activation of NH₃ and three N–H bonds, must be cleaved by surface oxygen species of to form N₂O.

Furthermore, it has been reported that the Mn–O bond energies of manganese oxides play an important role in the SCR reaction [165,179] in which the active oxygen species of MnO_x with lower Mn–O bond energy can facilitate the cleavage of more N–H bonds in NH₃ molecule to form more adsorbed nitrogen atom species, which lead to greater amounts of deep oxidation by-product [165]. Therefore, this is helpful to develop novel catalysts for NH₃–SCR of NO_x with low selectivity to N₂O.

3.4. Promoting Effect of Redox Cycles

Ce can enhance the low–temperature activity of MnO_x catalyst due to its redox shift between Ce⁴⁺ and Ce³⁺ [26,163,180]. It is accepted that the combination of CeO₂ and MnO_x significantly promotes the catalytic activity of low-temperature NH₃–SCR reaction, contributing to the following reactions [173]: (1) 2MnO₂ → Mn₂O₃ + O*; (2) Mn₂O₃ + 2CeO₂ → 2MnO₂ + Ce₂O₃; and (3) Ce₂O₃ + 1/2O₂ → 2CeO₂. The higher Ce⁴⁺/Ce³⁺ ratio may result in the higher SCR activity due to the intensified oxygen storage and release between Ce⁴⁺ and Ce³⁺ via the following equations: (1) 2CeO₂ → Ce₂O₃ + O* and (2) Ce₂O₃ +1/2 O₂ → 2CeO₂, which can promote the oxidation of NO to NO₂ [73]. It is widely accepted that O_α (chemisorbed oxygen) species are more active than O_β (lattice oxygen) species due to the higher mobility [73,157]. Moreover, the higher percentage of O_α species is conducive to the prior oxidation of NO to NO₂, leading to the enhancement of the “fast-SCR” reaction (4NH₃ + 2NO + 2NO₂ → 4N₂ + 6H₂O).

In many kinds of literature [24,54,78,181], it has been proposed that the electron transfer between the variable valence elements is beneficial to improve the redox ability of the multi-metal oxides catalyst and further promote the catalytic performances. Therefore, in the SCR process of the catalyst, such a cycle system for enhancing activity can be expressed as Cr⁵⁺ + 2Mn³⁺ ↔ Cr³⁺ + 2Mn⁴⁺ [78], Co³⁺ + Mn³⁺ ↔ Co²⁺ + Mn⁴⁺ [83], Ni³⁺ + Mn³⁺ ↔ Ni²⁺ + Mn⁴⁺ [24], Cu²⁺ + Mn³⁺ ↔ Cu⁺ + Mn⁴⁺ [54], and the dual redox cycles (Mn⁴⁺ + Ce³⁺ ↔ Mn³⁺ + Ce⁴⁺, Mn⁴⁺ + Ti³⁺ ↔ Mn³⁺ + Ti⁴⁺) [181].

In general, Mn–based catalysts have variable valence states and good redox capability at low temperatures, which are associated with the crystal structure, crystallinity, specific surface area, oxidation state, active oxygen, active sites and acidity on the surface, etc. [52]. As mentioned above, many reports [3,9,10,20,21,22,23] have indicated that the ER mechanism and LH mechanism are prevalent in low-temperature NH₃–SCR reaction on Mn–based catalysts. However, the study of active sites remains contentious between Mn³⁺ active sites [31,55,73,182] and Mn⁴⁺ active sites [26,43,54,79,158,183]. The synergistic effect of promoter about the pending interactions and synergies between Mn and promoter M elements needs to be further studied, including the surface acidity and oxidation ability of adsorption sites under multi-element effect. In addition, the contribution of ER mechanism and LH mechanism of NH₃–SCR reaction system needs to be further defined.

4. Reaction Kinetics

4.1. Macro–Kinetics

Assuming the reaction was free of diffusion limitation, the intrinsic rate of NO conversion (R_NO) as a function of reactant concentrations can be expressed as follows [22]:

$$R_{\mathrm{NO}}=k\times {[\mathrm{NO}]}^x\times {[{\mathrm{NH}}_3]}^y\times {[{\mathrm{O}}_2]}^z,$$

(18)

where R_NO is the rate of NO conversion, k is the rate constant, and x, y, z are the reaction orders of NO, NH₃ and O₂, respectively.

It has become a consensus that the SCR reaction is approximately zero–order with respect to NH₃ and first order with respect to NO over MnO_x-CeO₂ [22], MnO_x/TiO₂ [184], and Mn₂O₃-WO₃/γ-Al₂O₃ catalysts [185], etc. In spite of the different reaction-order of O₂, it is accepted that oxygen plays an important role in the SCR reaction, and the NO_x conversion shows nearly no change when O₂ concentration is more than 1–2%.

4.2. Micro–Kinetics

Both LH and ER mechanism exist in the low temperature SCR system over Mn-based catalysts as previously described. In recent years, Junhua Li’s groups [112,164,186] made several breakthroughs about kinetics with the identification of LH and ER mechanisms of low–temperature SCR catalysts. The detailed summaries are described as follows.

It is well known that the concentration of NH₃ and NO in the gas phase is sufficiently high for the surface to be saturated with adsorbed NH₃ and NO, so the concentration of NH₃ [NH_3(ads)] and NO [NO_(ads)] adsorbed on the surface at a specific temperature is invariable and can be described as Reactions (19) and (20), respectively:

$${[\mathrm{NH}}_{3\left(\mathrm{ads}\right)}]=k_1\times c_{acid},$$

(19)

$$-\frac{d[{\mathrm{NO}}_{\left(\mathrm{ads}\right)}]}{dt}=-\frac{d[{\mathrm{M}}^{\mathrm{n}+}]}{dt}=k_2\times [{\mathrm{NO}}_{\left(\mathrm{ads}\right)}]\times [{\mathrm{M}}^{\mathrm{n}+}],$$

(20)

where, k₁, k₂ and c_acid are a constant and the acidity on the catalyst, respectively. k₁ and k₂ would rapidly decrease with the increase of reaction temperature.

4.2.1. LH Mechanism

The kinetic equations of N₂ and N₂O formation through the LH mechanism can be approximately described as:

$$\frac{d[{\mathrm{N}}_2]}{dt}\mathsf{丨}\mathrm{LH}=k_3\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}-\mathrm{NH}}_3] (\text{from Reactions} (3) \text{and} (5)),$$

(21)

$$\frac{d[{\mathrm{N}}_2\mathrm{O}]}{dt}\mathsf{丨}\mathrm{LH}=k_4\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}}_2{-\mathrm{NH}}_3] (\text{from Reactions} (4) \text{and} (6)),$$

(22)

where k₃, k₄, [M⁽ⁿ⁻¹⁾⁺−O−NO−NH₃], and [M⁽ⁿ⁻¹⁾⁺−O−NO₂−NH₃] are the decomposition rate constants of NH₄NO₂ and NH₄NO₃, and the concentrations of NH₄NO₂ and NH₄NO₃ on the surface, respectively.

The kinetic equations of NH₄NO₂ and NH₄NO₃ formation (i.e., Reactions (5) and (6)) can be approximately described as Reactions (23) and (24), respectively:

$$\frac{d{[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}-\mathrm{NH}}_3]}{dt}\mathsf{丨}\mathrm{LH}=k_5\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}-\mathrm{NH}}_3]\times {[\mathrm{NH}}_{3\left(\mathrm{ads}\right)}],$$

(23)

$$\frac{d{[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}}_2{-\mathrm{NH}}_3]}{dt}\mathsf{丨}\mathrm{LH}=k_6\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}}_2{-\mathrm{NH}}_3]\times {[\mathrm{NH}}_{3\left(\mathrm{ads}\right)}],$$

(24)

where k₅, k₆, [M⁽ⁿ⁻¹⁾⁺−O−NO], [M⁽ⁿ⁻¹⁾⁺−O−NO₂], and [NH_3(ads)] are the reaction kinetic constants of Reactions (5) and (6), and the concentrations of nitrite, nitrate and NH₃ adsorbed on the surface, respectively.

The kinetic equations of nitrite and nitrate formation (i.e., Reactions (3) and (4)) can be approximately described as Reactions (25) and (26):

$$\frac{d{[\mathrm{M}}^{(\mathrm{n}-1)+}-\mathrm{O}-\mathrm{NO}]}{dt}\mathsf{丨}\mathrm{LH}=k_7\times {[\mathrm{M}}^{\mathrm{n}+}=\mathrm{O}]\times {[\mathrm{NO}}_{\left(\mathrm{ads}\right)}],$$

(25)

$$\frac{d{[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}}_2]}{dt}\mathsf{丨}\mathrm{LH}=k_8\times {[\mathrm{M}}^{\mathrm{n}+}=\mathrm{O}]\times {[\mathrm{NO}}_{\left(\mathrm{ads}\right)}][{\mathrm{O}}_2{]}^{\frac{1}{2}},$$

(26)

where k₇, k₈, [Mⁿ⁺=O] and [NO_(ads)] are the reaction kinetic constants of Reactions (3) and (4), and the concentrations of Mⁿ⁺ and NO adsorbed on the surface, respectively.

Meanwhile, the concentration of Mⁿ⁺ on the surface can be regarded as a constant at the steady state, as it can be rapidly recovered through Reaction (7). It can be implied from Reactions (23)–(26) that the formation of NH₄NO₂ and NH₄NO₃ is approximately not related to the concentrations of gaseous NO and NH₃. Hinted at by Reactions (24) and (25), the formations of N₂ and N₂O via LH mechanism are approximately independent of gaseous NO concentration.

4.2.2. ER Mechanism

The kinetic equations of N₂ and N₂O formation through the ER mechanism (i.e., Reactions (10) and (11)) can be described as Reactions (27) and (28):

$$\frac{d[{\mathrm{N}}_2]}{dt}\mathsf{丨}\mathrm{ER}=-\frac{d[{\mathrm{NO}}_{\left(\mathrm{gas}\right)}]}{dt}=k_9\times [{\mathrm{NH}}_2]\times [{\mathrm{NO}}_{(\mathrm{gas})}] (\text{From Reactions} (8) \text{and} (10)),$$

(27)

$$\frac{d[{\mathrm{N}}_2\mathrm{O}]}{dt}\mathsf{丨}\mathrm{ER}=-\frac{d[{\mathrm{NO}}_{\left(\mathrm{gas}\right)}]}{dt}=-\frac{d[\mathrm{NH}]}{dt}=k_{10}\times [\mathrm{NH}]\times [{\mathrm{NO}}_{(\mathrm{gas})}] (\text{From Reactions} (9) \text{and} (11)),$$

(28)

where k₉, k₁₀, [NH₂], [NH] and [NO_(g)] were the reaction rate constants of Reactions (10) and (11), the concentrations of NH₂ and NH on the surface, and gaseous NO concentration, respectively.

The reaction kinetic equations of NH₂ and NH formation can be described as Reactions (29) and (30):

$$\frac{d[{\mathrm{NH}}_2]}{dt}\mathsf{丨}\mathrm{ER}=k_{11}\times [{\mathrm{NH}}_{3\left(\mathrm{ads}\right)}]\times [{\mathrm{M}}^{\mathrm{n}+}=\mathrm{O}],$$

(29)

$$\frac{d[\mathrm{NH}]}{dt}\mathsf{丨}\mathrm{ER}=k_{12}\times [{\mathrm{NH}}_2]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}],$$

(30)

where k₁₁, k₁₂, [Mⁿ⁺=O] and [Mn⁴⁺=O] were the reaction kinetic constant of Reactions (8) and (9), and the concentrations of Mⁿ⁺ and Mn⁴⁺ on the surface, respectively.

At the steady state of reaction, NH concentration will not be varied. According to Reactions (28) and (30), the variation of NH concentration can be described as follows:

$$\frac{d[\mathrm{NH}]}{dt}\mathsf{丨}\mathrm{ER}=k_{12}\times [{\mathrm{NH}}_{2\left(\mathrm{ads}\right)}]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}]-k_{10}\times [\mathrm{NH}]\times {[\mathrm{NO}}_{\left(\mathrm{gas}\right)}]=0.$$

(31)

Thus,

$$[\mathrm{NH}]\mathsf{丨}\mathrm{ER}=\frac{k_{12}\times [{\mathrm{NH}}_2]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}]}{k_{10}\times {[\mathrm{NO}}_{\left(\mathrm{gas}\right)}]},$$

(32)

and

$$\begin{array}{ll}\frac{d[{\mathrm{N}}_2\mathrm{O}]}{dt}\mathsf{丨}\mathrm{ER}& =k_{10}\times [{\mathrm{NO}}_{(\mathrm{gas})}]\times \frac{k_{12}\times [{\mathrm{NH}}_2]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}]}{k_{10}\times {[\mathrm{NO}}_{\left(\mathrm{gas}\right)}]}\\ & =k_{12}\times [{\mathrm{NH}}_2]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}]\end{array}$$

(33)

4.2.3. Total Reaction Kinetic Equations

Taking the contributions of both LH and ER mechanism into account, the kinetic equations of NO reduction and N₂O formation can be approximately described as follows:

$$\begin{array}{ll}{k}_{\mathrm{NO}}& =-\frac{d[{\mathrm{NO}}_{\left(\mathrm{gas}\right)}]}{dt}=-\frac{d[{\mathrm{NO}}_{\left(\mathrm{gas}\right)}]}{dt}丨\mathrm{ER}-\frac{d[{\mathrm{NO}}_{\left(\mathrm{gas}\right)}]}{dt}丨\mathrm{LH}\\ & =(k_9\times [{\mathrm{NH}}_2]\times [{\mathrm{NO}}_{(\mathrm{gas})}]+k_{12}\times [{\mathrm{NH}}_2]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}])\\ & +(k_3\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}-\mathrm{NH}}_3]+k_4\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}}_2{-\mathrm{NH}}_3])\\ & =k_{\mathrm{SCR}-\mathrm{ER}}[{\mathrm{NO}}_{\left(\mathrm{gas}\right)}]+k_{\mathrm{SCR}-\mathrm{LH}}+k_{\mathrm{NSCR}}\end{array},$$

(34)

$$k_{\mathrm{SCR}-\mathrm{ER}}=k_9\times [{\mathrm{NH}}_2],$$

(35)

$$k_{\mathrm{SCR}-\mathrm{LH}}=k_3\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}-\mathrm{NH}}_3],$$

(36)

$$\begin{array}{ll}{k}_{\mathrm{NSCR}}& =\frac{d[{\mathrm{N}}_2\mathrm{O}]}{dt}=-\frac{d[{\mathrm{N}}_2\mathrm{O}]}{dt}\mathsf{丨}\mathrm{ER}-\frac{d[{\mathrm{N}}_2\mathrm{O}]}{dt}\mathsf{丨}\mathrm{LH}\\ & =k_{12}\times [{\mathrm{NH}}_2]\times [{\mathrm{Mn}}^{4+}=\mathrm{O}]+k_4\times {[\mathrm{M}}^{(\mathrm{n}-1)+}{-\mathrm{O}-\mathrm{NO}}_2{-\mathrm{NH}}_3]),\end{array}$$

(37)

where k_NO, k_SCR-ER, k_SCR-LH and k_NSCR are the rate of NO reduction, the reaction rate constant of the SCR reaction (i.e., N₂ formation) through the ER mechanism, the reaction rate constant of the SCR reaction through the LH mechanism, and the reaction rate constant of the NSCR reaction (i.e., N₂O formation), respectively.

During the SCR reaction, the above kinetic constants can be obtained through the steady–state kinetic study involved in differential and integral calculus. It is a developing trend of micro–kinetics that is beneficial for defining the contribution of ER and LH mechanism to the whole SCR reaction.

5. New Insight from DFT Calculation

Computer-simulation based on quantum chemistry theory has become a visual tool for the chemical material research. In recent years, the application of quantum chemical calculation using density functional theory (DFT) has been developed rapidly in the field of SCR for the mechanisms of reaction and poisoning. A typical application of quantum chemistry includes three aspects as below.

5.1. Analysis of Material Properties

Li et al. [187] calculated orbital energy and surface charge of MnO₂ crystal with different planes and found that the activity of O₂ adsorption was in the sequence of (110) > (111) > (001), which was closely related to the Highest Occupied Molecular Orbital (HOMO) energy and surface residual electrons. Yu and Liu et al. [188,189] further found the adsorption capacity on beta-MnO₂ (110) was in the sequence of hydroxyl > water > hydroxyl radical > oxygen. Peng et al. [190] prepared CeO₂ catalysts with different metal doped and studied NH₃ adsorption on the catalyst in experimental and computational methods simultaneously of. CeO₂ doped with four metals (Fe, Mn, La and Y) were prepared by co-precipitation, and the corresponding slab model were constructed by DFT calculations to investigate the influences of dopants on the adsorption of NH₃ molecules. A higher reducibility, a larger quantity of labile surface oxygen and a greater extent of surface distortion are responsible for the high NH₃ activation of the Fe–Ce and Mn–Ce, which is higher than the La–Ce and Y–Ce samples. Moreover, Fe–Ce and Mn–Ce can provide more NH₃ adsorption sites on the Lewis acid sites while the Mn dopant can also provide more Brønsted acid sites, which could react with NH⁴⁺. The Mulliken charges indicated that the lone pair electrons of nitrogen can transfer to Fe or Mn cations, whereas this process is less favorable for the La and Y cations. Fe and Mn dopant atoms can significantly change the catalytic surfaces of CeO₂ (bond lengths and angles), forming more oxygen defects. These distortions of structure resulted in the enhancement of the catalysts reducibility.

Tang’s team [36,38] calculated the reaction between NO and NH₃ on the surface, as well as the impact of H₂O on the constructed oxygen-rich and oxygen-depleted plane (110) of alpha-MnO₂ based on topology. Duan et al. [191] calculated and analyzed the effect of Ni and Co that doped on MnO₂ through DFT calculation in multiple perspectives. These studies provide clues for a further understanding of the reaction mechanism.

5.2. Visualization of Reaction and Poisoning Mechanisms

The SCR reaction mechanism over Mn–CeO₂ catalysts has been investigated by DFT calculations [192,193]. Song et al. [193] investigated the NH₃–SCR of NO mechanism over a mode of Mn cations doped into the CeO₂ (111) surface by DFT + U calculations. They found that NH₃ was referentially adsorbed on the Lewis acid Mn sites and further dissociated one of its N–H bonds to form the key NH₂ intermediate. NO adsorption on this NH2 intermediate resulted in the form of nitrosamine (NH2NO) that could then undergo further N–H cleavage reactions to form OH groups. They proposed that the redox mechanism involved doped Mn, as Lewis acid sites for ammonia adsorption and O vacancies in the ceria surface makes N₂O decompose into the desired N₂ product.

Peng et al. [194,195] studied the alkali metals and As₂O₅ poisoning resistant capacity of V₂O₅-WO₃-TiO₂ by employing DFT theoretical approaches. They found that the alkali atom mainly influences the active site V species rather than W oxides, and the decrease of acidity after poisoned might directly reduce the catalytic activity calculated by the density of states (DOS) and projected DOS (PDOS) [194]. Phil et al. [196] inferred that the catalyst of high resistance to SO₂ poisoning should be combined with sulfate weakly using quantum chemical calculation, which means that the bonding strength of metal atom and oxygen atom is weaker. In this theory, they selected the transition metal Sb and prepared Sb–V₂O₅/TiO₂ catalyst of high resistance to H₂O and SO₂ poisoning.

Maitarad et al. [192] constructed the Mn–CeO₂ (110) model on the basis of a combination of experiment and calculation. DFT calculations demonstrated that the catalyst has excellent catalytic activity and NO and NH₃ can be adsorbed on the catalyst surface. Lu et al. [197] proposed that the formation of bidentate sulfates were mainly generated on the plane (110) of the CeO₂ surface during the SCR process by means of density of states, electron localization functions and charge density difference. Liu et al. [198] investigated the SO₂ poisoning mechanism of Mn_xCe_1−xO₂ catalyst used CeO₂ as the model by the method of calculation. By calculating the electronic structure, density of states and the energy barrier that SO₂ needs to overcome in the formation of sulfates of CeO₂ and Mn_xCe_1−xO₂ being obtained, it was noted that the CeO₂ deactivation was due to the formation of CeSO₄ rather than Ce₂(SO₄)₃, which was acidized by SO₂. For the Mn_xCe_1−xO₂ catalyst, it was easier to be poisoned since the introduction of Mn enhanced the thermal stability of the surface sulfate.

5.3. Design of Functional Materials

Cheng et al. [199], using DFT, found that surface oxygen vacancy was beneficial for the catalytic activity due to the decrease of exposed Mn atom valence and reduction of reaction energy barrier. Peng et al. [194] proposed that more oxygen vacancies will generate more empty orbits for lone pair electrons, resulting in more NH₃ adsorption and higher redox property of the catalyst. Therefore, it is a novel way to improve the catalytic activity through the oxygen defects. Phil et al. [196] believed that the catalyst with weak bond energy of Metal-SO₂ or sulfate will be highly resistant to SO₂. Thus, they selected the transition metal Sb using quantum chemical calculation and prepared Sb-V₂O₅/TiO₂ catalyst with high resistance to H₂O and SO₂.

Plainly, the quantum chemical calculation contributes a lot to the further research of catalytic mechanisms and phase characteristics of catalyst surface. It needs more in-depth theoretical investigation to help us understand the reaction mechanism of NH₃–SCR catalysts and SO₂ poisoning mechanism at the molecule level, which will benefit preparation of the catalysts with higher activity and resistance.

6. Conclusions

There is strong interest in developing excellent efficiency, strong stability, and high resistance catalysts for SCR in the low–temperature range (<250 °C) for the removal of NO_x from stationary sources such as coal-fired power plants, steel, cement, glass and other industries. Because of the various valence states and excellent redox performance at low temperatures, Mn-based catalysts have aroused widespread concern regarding the low temperature NH₃–SCR, while the poisoning effects by H₂O or/and SO₂ are some of the weaknesses. Many types of research have been done to study the SCR mechanisms, reaction kinetics and poisoning effects. It is accepted that the phenomena of LH and ER mechanisms or the coexistence of both in the SCR process are the major theories. The inhibition mechanisms of H₂O and SO₂ on activity are mainly due to the competitive adsorption with NH₃ and NO, reduction of interface reaction, deposition of ammonium sulfate and sulfurization of active components. Many works still remain to be done to study the inhibition mechanism of H₂O and SO₂ to the catalyst, especially for the impacts on the catalytic reaction paths and bulk phase structure of catalysts. Quantum chemical calculation using density functional theory has been introduced at the molecule level to analyze the catalyst’ properties, explicit reaction and poisoning mechanisms, and directs the design of a functional catalyst.

In recent years, many Mn-based catalysts have been widely investigated to improve the H₂O or/and SO₂ resistance, such as modified MnO_x–CeO₂, multi-metal oxides with the structure of special crystal or/and shape (such as perovskite-type, spinel, nanocages, hollow nanofibers, hexagonal microsheets, MOFs and core-shell, etc.), modified TiO₂ supporter and novel carbon supporter (ACF, CNTs, GE), etc. These catalysts show a relatively good resistance because of their pre-sulfating of dopants for protecting MnO_x, which are expected to break the resistance to H₂O or/and SO₂. Despite these great efforts that have been achieved, it is still a challenge to develop low temperature de-NO_x catalysts with a high tolerance and stability in the long-term and at a large-scale. We believe that the intensive study of H₂O and SO₂ inhibition effects through the combination analysis of in situ DRIFTS and DFT and the amplification of tolerance mechanism will be helpful to guide the design of excellent SCR catalysts with special crystal or/and shape structures.

Because of the stringent regulations on NO_x emissions and disadvantages of commercial V-based catalysts, the catalysts (techniques) have already received significant attention with the promising De-NO_x performance at low temperatures (<250 °C), including the non V-based catalysts for low temperature SCR, novel design of catalysts for the direct decomposition without any agent, quantitative catalytic oxidation for pre-oxidation combined absorption and development of novel NTP-assisted catalysis technology. Mn-based catalysts have attracted considerable attention for utilization in low-temperature SCR of NO_x in the presence of excess of oxygen and sulfur dioxide with a broad temperature window, without sacrificing too much of the efficiency.