Combined fast selective reduction using Mn-based catalysts and nonthermal plasma for NOx removal

Abstract

In this study, the concept of fast SCR for NO reduction with NH₃ as reducing agent is realized via the combination of nonthermal plasma (NTP) with Mn-based catalyst. Experimental results indicate that 10% wt. Mn-Ce-Ni/TiO₂ possesses better physical and chemical properties of surface, resulting in higher NO removal efficiency if compared with 10% wt. Mn-Ce/TiO₂ and 10% wt. Mn-Ce-Cu/TiO₂. Mn-Ce-Ni/TiO₂ of 10% wt. achieves 100% NO_x conversion at 150 °C, while 10% wt. Mn-Ce/TiO₂ and 10% wt. Mn-Ce-Cu/TiO₂ need to be operated at a temperature above 200 °C for 100% NO_x conversion. However, NO conversion achieved with 10% wt. Mn-Ce-Ni/TiO₂ is significantly reduced as H₂O_(g) and SO₂ are introduced into the SCR system simultaneously. Further, two-stage system (SCR with DBD) is compared with the catalyst-alone for NO_x conversion and N₂ selectivity. The results indicate that 100% NO_x conversion can be achieved with two-stage system at 100 °C, while N₂ selectivity reaches 80%. Importantly, NO_x conversion achieved with two-stage system could maintain >95% in the presence of C₂H₄, CO, SO₂, and H₂O_(g), indicating that two-stage system has better tolerance for complicated gas composition. Overall, this study demonstrates that combining NTP with Mn-based catalyst is effective in reducing NO_x emission at a low temperature (≤200 °C) and has good potential for industrial application.

Introduction

Continuous emission of large quantity of NO_x into atmosphere has caused various environmental problems, such as acid rain, photochemical smog, and visibility degradation. Also, it is harmful to human health. Several post-combustion methods including selective catalyst reduction (SCR), selective non-catalyst reduction (SNCR), and non-selective catalyst reduction (NSCR) are available to control NO_x emissions. Among them, selective catalyst reduction (SCR) is of the highest NO_x removal efficiency and NH₃ is commonly applied as a reducing agent to reduce NO_x to form N₂ in the presence of O₂ (Jun et al. 2015). Generally, NO_x removal efficiency achieved with SCR process could reach 90% if appropriate catalyst is applied (Wu et al. 2007). The main reaction of SCR is described as rReaction (1).

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

(1)

V₂O₅-WO₃/TiO₂ is typically used in SCR process; however, it needs to be operated within the temperature range of 250–350 °C for high efficiency. The narrow reaction temperature window is considered as one of the main disadvantages of SCR. In addition, the catalyst deactivation always takes place, because SCR catalyst is generally placed before particle removal equipment and flue gas desulfurization (FGD) (Phil et al. 2008). Also, SCR process shows good efficiency for NO_x removal with urea as a reducing agent for the treatment of flue gas from diesel engine (Choi and Woo 2015; Cimino et al. 2016). Urea has several advantages such as easier transportation, higher safety of storage, and higher stability if compared with NH₃ (Feng and Lü 2015). However, the temperature of low-loading diesel engine exhaust is relatively low (100–200 °C). As mentioned previously, the most disadvantage of applying SCR in controlling diesel emission is that the catalysts have to be operated at a temperature higher than 200 °C. As a result, conventional SCR could not be applied for removing NO_x from diesel engine exhaust (Chen et al. 2016; Yoshida et al. 2012). Therefore, how to develop a low-temperature SCR has become an emerging issue. Recently, relevant study indicated that SCR catalyst has good potential to convert the NO_x to N₂ at lower temperatures; further, it can be placed downstream of the particle control device such as bag filter and the desulfurization system to avoid the deactivation of catalyst (Huang et al. 2016).

Manganese-based (Mn-based) catalysts such as MnO_x, MnO_x/TiO₂ (Jin et al. 2014), Mn-Ce/TiO₂ (Cao et al. 2015), and Mn-Ce-Fe/γ-Al₂O₃ (Thirupathi and Smirniotis 2011a) have attracted much attention due to their high capability in converting NO_x at low temperatures (≦200 °C). Previous studies indicated that applying different metals as promoters can enhance the activity of Mn-based catalyst, such as Ce, Cu, Fe, and Ni. Especially Ce is most applied for developing low-temperature catalysts (Liu et al. 2017). Kwon et al. (2015) reported that Ce has good activity toward NO_x reduction due to its redox properties. Additionally, Ce could store and release oxygen via the transformation between Ce⁴⁺ and Ce³⁺ to promote SCR reaction (Kwon et al. 2015). Zhou et al. (2016) investigated Ce-based catalysts with different Ce valences (Zhou et al. 2016) and indicated that catalyst has a higher sulfur dioxide resistance as the ratio of Ce⁴⁺/(Ce⁴⁺+Ce³⁺) is increased. Thirupathi et al. (2011) indicated that activity of Mn-based catalyst can be promoted by adding Ni for SCR process, because Ni can also enhance redox capacity of catalyst (Thirupathi and Smirniotis 2011b).

However, Mn-based catalysts still face serious drawbacks although it has potential to be operated at a lower temperature for SCR process (Yang et al. 2016). For example, N₂O is the inevitable product in a low-temperature SCR process, resulting in a lower N₂ selectivity. Niu et al. (2016) indicated that N₂O may be generated from direct oxidation of NH₃ in the excess O₂ via reaction (2) at a low temperature (≤150 °C) or NO would react with adsorbed NH₃ on catalyst surface via reaction (3) (Niu et al. 2016).

$$ 2{NH}_3+2{\mathrm{O}}_2\to {\mathrm{N}}_2\mathrm{O}+3{\mathrm{H}}_2\mathrm{O} $$

(2)

$$ 4\mathrm{NO}+4{NH}_{3\left(\mathrm{ads}\right)}+3{\mathrm{O}}_2\to 4{\mathrm{N}}_2\mathrm{O}+6{\mathrm{H}}_2\mathrm{O} $$

(3)

Similarly, Mn-based catalysts have poor durability and resistance in complicated gas composition (especially in the presence of SO₂ and H₂O_(g)); Mn-based catalysts could be seriously poisoned by formation of ammonia bisulfate (NH₄HSO₄ and (NH₄)₂SO₄) on the catalyst. On the other hand, SCR performance could be significantly enhanced as fast-SCR is applied. Especially, fast-SCR is effective to reduce NO to N₂ at a low temperature. Generally, fast-SCR has a higher reaction rate and wider reaction temperature window if compared with conventional SCR. Fast-SCR was first investigated by Koebel et al. (2001, 2002) and Madia et al. (2002) and can be described as reaction (4). Relevant study indicated that redox mechanism in fast-SCR is similar to the standard SCR reaction (Grossale et al. 2008). For fast-SCR process, NO₂ plays a crucial role which serves as a more efficient oxidizing agent than O₂ in the redox process of the SCR reaction. In fast-SCR, NH₄NO₃ is formed via the reaction between NH₃ and NO₂, and subsequently, NH₄NO₃ will be decomposed by NO described in reactions (5) and (6). Relevant study indicated that reactions (5) and (6) are closely related to the fast-SCR chemistry (Grossale et al. 2008). The vital process is the redox reactions as indicated in reactions (5) and (6), which dominate the reaction rate of fast-SCR. It has been proved that existence of NO₂ in the exhaust gas is essential and the content of NO₂ approximately equals to that of NO favors fast-SCR reaction (Iwasaki and Shinjoh 2010).

$$ \mathrm{NO}+{\mathrm{N}\mathrm{O}}_2+2{NH}_3\to 2{\mathrm{N}}_2+3{\mathrm{H}}_2\mathrm{O} $$

(4)

$$ 2{NH}_3+2{\mathrm{N}\mathrm{O}}_2\to {\mathrm{N}}_2+{NH}_4{\mathrm{N}\mathrm{O}}_3+{\mathrm{H}}_2\mathrm{O} $$

(5)

$$ {NH}_4{\mathrm{N}\mathrm{O}}_3+\mathrm{NO}\to {\mathrm{N}}_2+{\mathrm{N}\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O} $$

(6)

However, NO_x in the typical flue gas of combustion system is composed of 95% NO and 5% NO₂; therefore, fast-SCR process is difficult to be applied. How to effectively convert NO into NO₂ is an important step for fast-SCR process (Jõgi et al. 2016; Kang et al. 2006a). Nonthermal plasma (NTP) has been demonstrated as an effective technology to oxidize NO into NO₂ or NO₃ ⁻ via the collision with highly active radicals generated, i.e., O and OH (Patil et al. 2016). Combining catalyst with NTP to form plasma catalysis is a novel reaction system, which is regarded as feasible technology for reducing the emission of gaseous pollutants. Generally, plasma catalysis system can be distinguished into two configurations, i.e., in-plasma catalysis (IPC) and post-plasma catalysis (PPC). The former is similar to a packed-bed reactor, namely, catalyst is directly packed into the discharge zone. The latter implies that catalyst is located downstream NTP reactor, and catalysis system and plasma reactor are operated separately.

Bröer and Hammer (2000) reported that NO_x conversion can reach 90% at 160 °C by combining dielectric barrier discharges (DBD) with V₂O₅-WO₃/TiO₂ catalyst to form two-stage system (Bröer and Hammer 2000). Tran et al. (2004) applied DBD and In-doped γ-alumina catalyst to form a two-stage system for reducing of NO_x and indicated that 100% NO_x conversion could be achieved at 350 °C (Tran et al. 2004). Cho et al. (2012) also applied plasma to assist the HC-SCR and indicated that NO_x conversion reached 100% with the operating temperature of 300 °C and applied voltage of 16 kV (Cho et al. 2012). Hence, PPC has good potential to reduce the emission of NO_x.

As mentioned previously, Mn-based catalysts have potential to be applied for NO_x reduction at a temperature below 200 °C. The aim of this study is to develop low-temperature SCR catalysts through various Mn-modified catalysts, which are Mn-Ce/TiO₂, Mn-Ce-Ni/TiO₂, and Mn-Ce-Cu/TiO₂, respectively. Additionally, a lab-scale two-stage system consisting of NTP and low-temperature SCR catalysts is developed to reduce NO from simulated gas streams. In this two-stage system, NTP can be regarded as an oxidation system to achieve NO₂/NO ≈ 1. It is expected that combining Mn-based catalyst with plasma would enhance NO_x conversion and reduce the formation of N₂O. Furthermore, the effects of operating parameters on NO_x removal efficiency are extensively evaluated via a lab-scale experimental setup, and gas streams with complicated gas compositions are applied to evaluate the practical application of this two-stage system for NO removal.

Experimental

Catalyst preparation

A series of Mn-based catalysts including 10% wt. Mn-Ce/TiO₂, 10% wt. Mn-Ce-Ni/TiO₂, and 10% wt. Mn-Ce-Cu/TiO₂ were prepared by wet impregnation method. First of all, corresponding metal nitrates (including Mn(NO₃)₂·4H₂O (purity: 98%, Showa), Ce(NO₃)₃·6H₂O (purity: 99%, ACROS), Ni(NO₃)₂·6H₂O (purity: 99%, Showa), and Cu(NO₃)·3H₂O, (purity: 100%, Showa) as precursors were dissolved in deionized water to prepare aqueous solution with appropriate stoichiometry. All ratios of active phases of catalysts (i.e., Mn-Ce, Mn-Ce-Ni, and Mn-Ce-Cu) were controlled at 1. Subsequently, corresponding P25-type TiO₂ (purity: 100%, Degussa) used as support was added into mixture solution and mixed completely at 80 °C until water was evaporated. Secondly, the residual solid precursor was placed into an oven to dry overnight, and then calcined in air at 400 °C for 4 h with the heating rate of 5 °C/min. Finally, all catalysts were sieved to 30–70 mesh for catalytic test.

Catalyst characterization

The crystal structures of catalysts were characterized by powder X-ray diffraction (XRD). The XRD (D8AXRD BRUKER, Germany) profiles were obtained using an X-ray diffractometer, operated at 40 kV and 40 mA using Cu-Kα radiation with a nickel filter, diffraction with 2θ were recorded over the range of 10–80°. The specific surface areas, total pore volume, and average pore diameter of the catalysts were measured by N₂ adsorption at −196 °C with a Micromeritics ASAP 2010 (ASAP2010 Micromeritics, USA). Field-emission scanning electron microscopy (FE-SEM) equipped with an energy-dispersive X-ray spectroscopy (EDS) was used to examine the micro-structure of catalyst surface and the elemental composition of catalysts by S80 JEOL (SEM, S80 JEOL, Japan). X-ray photoelectron spectroscopy (XPS, Sigma Probe VG, UK) was used to analyze the chemical state of elements of the catalysts, and XPS spectra were recorded with monochromatic Al anode X-ray which was equipped with a concentric hemispherical analyzer. Al Kα (1486.6 eV) X-ray source was used for excitation. The binding energies were referenced to the C1s line at 284.5 eV. Thermogravimetric analysis (TGA, Pyris 1 TGA PERKIN ELMER, USA) was carried out to investigate the thermal stability of sulfate species in a static N₂ atmosphere, with Hitachi STA7300. The TGA experiments were conducted with 10–20-mg catalysts, and it was analyzed with the temperature ranging from room temperature to 900 °C at a rate of 10 °C/min.

Activity test

Low-temperature reduction of NO with ammonia was first performed with SCR process at a fixed-bed quartz reactor (I.D.= 20 mm) containing 6 g catalyst in excess oxygen. The SCR process was carried out with the temperature ranging from 50 to 200 °C and the total gas flow rate was fixed at 1500 sccm, corresponding to a gas hourly space velocity (GHSV) of 20,000 h⁻¹. Typical gas mixture consisted of NO = 300 ppm, NH₃ = 300 ppm, O₂ = 10%, H₂O_(g) = 10% (when used), SO₂ = 100 ppm (when used), and N₂ as balance. Various gases were provided by gas cylinders and a set of mass flow controllers (MFC) were used for regulating the gas flow rates, while H₂O_(g) was introduced into the system by a peristaltic pump. For the analysis of exhaust, a condensation system was used to capture H₂O_(g) in order to avoid the damage to instruments. An NO_x analyzer (Testo 350, Germany) was used to measure NO and NO₂ concentrations while N₂O was analyzed by Fourier transform infrared spectrophotometer (FTIR, Nicolet 6700 Thermo Scientific, USA).

Further, the two-stage system was constituted by combining DBD-type (dielectric barrier discharge) NTP with Mn-based catalyst, and it was evaluated for the effectiveness in removing NO_x. As shown in Fig. 1, the catalysis system (SCR process) was placed downstream of the NTP (similar to PPC system). Likewise, the total gas flow rate of two-stage system was controlled at 1500 sccm. In two-stage system, all operating conditions were similar to SCR process. For NTP system, the DBD-type reactor was constituted of a quartz tube (I.D. = 10 mm) filled with glass beads, a stainless steel rod, and a stainless steel wire mesh, where stainless steel rod and a stainless steel wire mesh were used as the inner and outer electrodes, respectively. The operating temperature and pressure of NTP were 25 °C and atmospheric pressure, respectively. The DC pulse power was applied to generate plasma, and the applied voltage and frequency were controlled at 12–20 kV and 8–10 kHz, respectively, for test. The discharge power was measured by a digital oscilloscope equipped with a current probe and a high-voltage probe. The NO_x conversion and N₂ selectivity are calculated via Eqs. (7) and (8), respectively:

$$ \mathrm{NO}\kern0.6em \mathrm{conversion}\left(\%\right)=\frac{{\left[\mathrm{NO}\right]}_{\mathrm{in}}-{\left[\mathrm{NO}\right]}_{\mathrm{out}}}{{\left[\mathrm{NO}\right]}_{\mathrm{in}}}\times 100\left(\%\right) $$

(7)

$$ S{N}_2=\frac{\left[\mathrm{N}\mathrm{Oin}-\mathrm{NOout}\right]-\left[\mathrm{N}{\mathrm{O}}_2\right]-2\left[{\mathrm{N}}_2\mathrm{O}\right]}{\left[\mathrm{N}\mathrm{Oin}-\mathrm{NOout}\right]}\times 100\% $$

(8)

Fig. 1

Schematic of the two-stage system for NO_x reduction

Results and discussion

Catalyst characterizations—XRD, BET, and SEM

The XRD patterns of Mn-Ce/TiO₂, Mn-Ce-Ni/TiO₂, and Mn-Ce-Cu/TiO₂ are presented in Fig. 2, and the result indicates that all catalysts prepared present strong diffraction peaks of TiO₂, being mainly contributed by anatase crystalline. It is observed that three Mn-based catalysts have basically amorphous peaks of MnO_x on the catalyst surface. Previous studies reveal that amorphous MnO₂ is of higher activity if compared with crystalline MnO₂ for NO reduction (Boningari et al. 2015; Tang et al. 2007). The presence of amorphous manganese oxide in Mn-based catalysts may be one of the key factors for their excellent activities in removing NO_x (Kang et al. 2006b; Peña et al. 2004). In addition, Cu and Ni are not observed in XRD profiles of Mn-Ce-Cu/TiO₂ and Mn-Ce-Ni/TiO₂, respectively. It is speculated that they may be incorporated into Mn lattice to form solid solution. The diffraction peaks of CeO₂ are not present in Mn-Ce-Cu/TiO₂ due to less Ce content. As shown in EDS results (see Table 1), Mn-Ce-Cu/TiO₂ has lower contents of Cu, Mn, and Ce, and it is speculated that active phase (Mn-Ce-Cu) is not well supported on TiO₂ due to the addition of Cu.

Fig. 2

XRD patterns of Mn-based catalysts. a Mn-Ce-Ni/TiO₂. b Mn-Ce/TiO₂. c Mn-Ce-Cu/TiO₂

Table 1 Elemental compositions of the catalysts prepared

The specific surface areas (S_BET), average pore diameter, and pore volume of catalysts and TiO₂ are summarized in Table 2. The results indicate that BET surface areas of Mn-Ce/TiO₂, Mn-Ce-Ni/TiO₂, and Mn-Ce-Cu/TiO₂ are 48, 50, and 30 m²/g, respectively. Obviously, BET surface areas of catalysts prepared decrease as TiO₂ is impregnated with various metal elements. Possibly, parts of TiO₂ pores are covered. Also, it is found that pore volume of TiO₂ decreases to some extent after impregnation. Among Mn-based catalysts, Mn-Ce-Ni/TiO₂ shows the largest BET (~50 m²/g) and average pore diameter (8.6 nm), and it may correspond to the highest activity due to smaller diffusion limitation. As shown in Table 1, EDS analysis indicates that the elemental composition of the catalysts prepared deviates a little from the designated formula of catalysts. In addition, it is observed that Mn-Ce-Ni/TiO₂ has the highest oxygen content among three catalysts prepared. According to SCR mechanism, NH₃ is first adsorbed on reactive oxygen of catalyst. Hence, it is speculated that higher oxygen atom content may result in higher NO conversion. Figure 3 displays the SEM images of Mn-Ce-Ni/TiO₂, Mn-Ce/TiO₂, and Mn-Ce-Cu/TiO₂, respectively. The results reveal that three Mn-based catalysts are mainly constituted by pseudo-spherical particles; they are composed of irregular small particles corresponding to the prepared method. The average particle sizes of 3 Mn-based catalysts were calculated by Sherr’s equation, and their particle sizes are Mn-Ce-Ni/TiO₂ = 25 nm, Mn-Ce/TiO₂ = 28 nm, and Mn-Ce-Cu/TiO₂ = 31 nm, respectively. The particle sizes of catalysts rank in the order of Mn-Ce-Cu/TiO₂ > Mn-Ce/TiO₂ > Mn-Ce-Ni/TiO₂. Especially, Mn-Ce-Cu/TiO₂ reveals severe sintering after calcination (see Fig. 3), resulting in lower BET (30 m²/g). On the other hand, sintering of Mn-Ce-Ni/TiO₂ is not significant if compared with Mn-Ce/TiO₂ and Mn-Ce-Cu/TiO₂. Previous study indicated that addition of Ni to Mn-based catalysts can prevent its sintering, and higher catalytic activity is obtained (Thirupathi et al. 2011). Overall, Mn-Ce-Ni/TiO₂ presents better surface characteristics if compared with Mn-Ce/TiO₂ and Mn-Ce-Cu/TiO₂.

Table 2 BET, pore diameter, and pore volume of TiO₂ and three Mn-based catalysts

Fig. 3

SEM images of Mn-based catalysts. a Mn-Ce-Ni/TiO₂. b Mn-Ce/TiO₂. c Mn-Ce-Cu/TiO₂

SCR performance

Low-temperature SCR performance

Catalytic activities of three Mn-based catalysts prepared are first evaluated for NH₃-SCR reaction at various temperatures. As shown in Fig. 4, the NO_x conversions achieved increase with increasing temperature for three Mn-based catalysts. At 50 °C, NO_x conversion achieved with Mn-Ce-Ni/TiO₂, Mn-Ce/TiO₂, and Mn-Ce-Cu/TiO₂ are 58, 31, and 30%, respectively. As temperature is increased to 150 °C, NO_x conversions achieved with Mn-Ce-Ni/TiO₂ and Mn-Ce/TiO₂ reach 100% while Mn-Ce-Cu/TiO₂ needs to be operated at 200 °C for reaching 95% NO_x conversion. Activities of three Mn-based catalysts for NO reduction rank in the order of Mn-Ce-Ni/TiO₂ > Mn-Ce/TiO₂ > Mn-Ce-Cu/TiO₂. For the effect of Ni on activity of Mn-based catalyst, previous study indicated that Mn/TiO₂ has positive interaction if doped with Ni (Thirupathi et al. 2011). This interaction may lead to the enhancement of the oxygen mobility, resulting in higher activity for SCR process.

Fig. 4

The performance of three Mn-based catalysts (NO: 300 ppm, NH₃: 300 ppm, O₂: 10%, balance: N₂ and GHSV = 20,000 h⁻¹)

Additionally, the kinetics is investigated by the power rate law model. The activation energies are calculated on the basis of the conversion smaller than 20% by increasing GHSV. The kinetics study is conducted by assuming steady-state before and after reaction, and a mathematical model of a differential-type reactor is used for determining the reaction rates (−r _i ) at different temperatures (T) and inlet NO concentrations (C _i ). Subsequently, regression straight lines of 1/r _i versus 1/C _i are plotted to obtain the values of rate constant (k) with different temperatures. Then, a regression line is obtained by plotting ln k versus 1/T, and activation energy (E _a ) is calculated by Arrhenius equation (see Fig. 5). The result indicates that activation energy with Mn-Ce-Ni/TiO₂ as catalyst for SCR process is 23.1 kJ/mol.

Fig. 5

Arrhenius plot for rate constant of NO conversion with Mn-Ce-Ni/TiO₂ as catalyst

XPS analysis of catalysts

XPS spectra are used to understand the chemical bonding states on the catalyst surface for analysis of mechanism. XPS Mn 2p spectra of three fresh catalysts are presented in Fig. 6a. For all three catalysts, two main peaks corresponding to Mn 2p_3/2 and Mn 2p_1/2 are located at 642 and 653 eV, respectively. Among them, Mn 2p_3/2 can be separated into two peaks, corresponding to Mn⁴⁺ (642.0–642.5 eV) and Mn³⁺ (641.0–641.3 eV), respectively (Chang et al. 2013; Shaju et al. 2002).

Fig. 6

XPS analysis of three catalysts. a Mn 2p. b Ce 3d. c O 1s (①MnCeNi/TiO₂, ② MnCe/TiO₂, ③ MnCeCu/TiO₂)

Figure 6b shows the valence state of Ce 3d of three catalysts. The Ce 3d peaks were fitted into the sub-bands, and those marked as U₁ and U₂ represent Ce⁺³ species. The sub-bands marked as V₁, V₂, V₃, V₄, and V₅ are attributed to Ce⁴⁺ species. In addition, XPS spectra of the O 1s of three Mn-based catalysts are presented in Fig. 6c. The results indicate that the two main peaks appeared at 529~530 and 531~532 eV, corresponding to lattice oxygen (O _β ) and adsorbed surface oxygen (O _α ), respectively (Peng et al. 2013).

Fang et al. (2015) indicated that Mn⁴⁺ has a higher activity than Mn³⁺ and Mn²⁺ toward NO_x removal (Fang et al. 2015). Wang et al. (2016) mentioned that Mn⁴⁺ species could enhance the oxidation of NO to NO₂, further promote SCR performance at a low temperature (Wang et al. 2016). Wan et al. (2014) indicated that the surface Mn⁴⁺ species is the predominant valence state of the Mn-based catalysts and the activity of Mn-based catalysts increases with increasing Mn⁴⁺. To summarize above results, Mn⁴⁺ has better electron mobility and thus higher production of oxygen defect (Wan et al. 2014). Regarding the effect of Ce, previous study indicated that the activity of SCR catalyst increases with increasing Ce⁴⁺/(Ce⁴⁺+Ce³⁺) ratio. As shown in Fig. 6c, O 1s of three catalysts have two peaks, which are O_β and O_α, respectively. Relevant study indicated that the catalyst activity is related to ratio of O _α /(O _α  + O _β ) in SCR process (Shu et al. 2014). Generally, O _α has a higher mobility than O_β, leading to the enhancement of SCR catalytic activity (Ma et al. 2015). Indeed, the results of three Mn-based catalyst are similar to that reported in literature, and the ratios of Mn⁴⁺/(Mn⁴⁺+Mn³⁺), Ce⁴⁺/(Ce⁴⁺+Ce³⁺), and O_α/(O_α + O_β) of three catalysts are summarized at Table 3. Mn-Ce-Ni/TiO₂ reveals the best SCR performance of three catalysts, and it also has the highest Mn⁴⁺/(Mn⁴⁺+Mn³⁺), Ce⁴⁺/(Ce⁴⁺+Ce³⁺), and O_α/(O_α + O_β) ratios. Previous study indicated that addition of Ni to Mn-based catalyst improves surface properties of Mn-based catalyst, resulting in better catalysis, because Ni has good interaction with Mn and Ti (Thirupathi et al. 2011). Conversely, the performance of Mn-Ce-Cu/TiO₂ decreases due to weak or no interaction of Cu with Ti (Boccuzzi et al. 1997). The results indicate that Ni is a good promoter, and it could increase the formation of Mn⁴⁺, resulting in higher SCR performance. With the addition of nickel into Mn-Ce/TiO₂ catalyst, the ratio of Mn⁴⁺/(Mn⁴⁺+Mn³⁺) increases from 72 to 79%. Obviously, the activity of Mn-Ce/TiO₂ is significantly improved by modification with Ni. Experimental results indicate that Mn-Ce-Ni/TiO₂ has the highest activity on NO_x reduction. Therefore, Mn-Ce-Ni/TiO₂ is mainly applied for further tests.

Table 3 Percentages of different valence states for Mn and Ce of three catalysts

Effect of SO₂ and H₂O_(g) on low temperature SCR

As reported previously, Mn-Ce-Ni/TiO₂ catalyst is demonstrated with the highest activity in SCR process. The long-term activity is further investigated for NO_x removal, and the results indicate that NO_x conversion maintains above 99% in the time course over 80 h at 150 °C, indicating high stability of Mn-Ce-Ni/TiO₂ for NO reduction.

It is important to evaluate the effects of H₂O_(g) and SO₂ on the catalytic activity(Shen et al. 2014). Figure 7 illustrates the effects of 10% H₂O_(g) and 100 ppm SO₂ on SCR performance at 150 °C. The NO_x conversion decreases from 100 to 80% as 100 ppm SO₂ and 10% H₂O_(g) are introduced into the gas stream simultaneously. As SO₂ and H₂O_(g) feedings are turned off, the NO_x conversion gradually increases to 90%. These results indicate that the activity of Mn-Ce-Ni/TiO₂ could be inhibited as H₂O_(g) and SO₂ are simultaneously present in flue gas. Possibly, SO₂ are oxidized into SO₃ or sulfate species on MnO_x during SCR process, then SO₃ or sulfate species move to Ce or Ni to form bulk-sulfate species. Yu et al. (2010) mentioned that the NO conversion can be recovered to 90%, being attributed to decomposition of adsorbed ammonium sulfate when H₂O_(g) and SO₂ are turned off (Yu et al. 2010).

Fig. 7

The durability test and tolerance of SO₂ and H₂O_(g) for SCR reaction

To confirm SCR behaviors in the presence of SO₂ and H₂O_(g), used Mn-Ce-Ni/TiO₂ catalysts were analyzed by thermogravimetric analysis. The sulfate species and H₂O_(g) adsorbed on the surface of catalysts would lead to weight increase. As shown in Fig. 8, the TGA curve presents three major weight losses for Mn-Ce-Ni/TiO₂. The first one appears at 100 °C, which can be attributed to the evaporation of water. The second and third losses appear at approximately 400 and 800 °C, respectively, which are originated from the decomposition of sulfate species, such as NH₄HSO₄ and (NH₄)₂SO₄ (Putluru et al. 2015).

Fig. 8

TGA curve of used Mn-Ce-Ni/TiO₂ catalyst subjected to H₂O_(g) and SO₂

NO_x reduction via two-stage system

The system combining NTP and Mn-Ce-Ni/TiO₂ catalyst is evaluated for the effectiveness in reducing NO_x. In order to elucidate the effects of NTP on catalytic performance, the gas compositions are kept the same as that listed in section “SCR performance.” The DBD experiment is conducted with the applied voltage of 15.5 kV and a frequency of 10,000 Hz to generate plasma. The NO and NO₂ concentrations after plasma are measured as 165 and 153 ppm, respectively, and the ratio of NO₂/NO is about 0.93. Figure 9 shows the SCR performances of the Mn-Ce-Ni/TiO₂ catalyst without and with DBD, respectively. At 50 °C, the NO_x conversion achieved with two-stage system reaches 80%, indicating that the NO_x conversion achieved with t two-stage system is higher than that achieved with the catalyst alone by 20%. When temperature is increased to 100 °C, two-stage system reaches 100%. The experimental results indicate that combining Mn-Ce-Ni/TiO₂ with NTP is more effective in reducing NO_x compared with the catalysis alone at a low temperature. Hence, it can be applied for controlling NO_x emission from low-loading diesel engine in which the temperature of exhaust gas ranges from 100 to 200 °C (Xie et al. 2012).

Fig. 9

Comparison of MnCeNi/TiO₂ alone and MnCeNi/TiO₂ + plasma , a NO_x conversion, b N₂O concentration, and c N₂ selectivity (300 ppm NO, 300 ppm NH₃, 10% O₂, applied voltage: 15.5 kV, frequency: 10 kHz and packed with glass beads)

As NO reacts with NH₃, N₂O is an important by-product. Thus, N₂O is also measured in this study and the results indicate that the two-stage system produced a lower N₂O concentration if compared with catalysis alone. At various temperatures tested, the N₂O concentration generated by two-stage system are all below 30 ppm, while the catalysis-alone generates 90 ppm of N₂O at 150 °C. The N₂O concentration of catalyst alone is significantly higher than that generated by the two-stage system. Niu et al. (2016) indicated that N₂O may be generated from the direct oxidation of NH₃ in the excess O₂ via reaction (2) or NO would react with adsorbed NH₃ on surface of catalyst via reaction (3) at a low temperature (≤150 °C). Also, the N₂O formation pathway in SCR reaction is the decomposition of intermediate NH₄NO₃ via reaction (9) (Grossale et al. 2008).

$$ {NH}_4{\mathrm{N}\mathrm{O}}_3\to {\mathrm{N}}_2\mathrm{O}+2{\mathrm{H}}_2\mathrm{O} $$

(9)

In two-stage system, the plasma can assist the reaction of SCR to induce fast-SCR reaction, achieving a higher NO conversion and a lower N₂O generation at a temperature ranging from 50 to 200 °C. The two-stage system can also make SCR reaction more complete than catalyst alone. Moreover, NH₃ would be effectively consumed by reaction with NO₂ to form NH₄NO₃. Subsequently, NH₄NO₃ is reduced with NO via reaction (10) to inhibit N₂O formation. In addition, N₂ selectivity of the two-stage system is higher than that of catalyst alone at a temperature range of 50–200 °C because two-stage system generates a lower N₂O concentration if compared with the catalyst-alone.

$$ \mathrm{NO}+{NH}_4{\mathrm{N}\mathrm{O}}_3\to {\mathrm{N}\mathrm{O}}_2+{\mathrm{N}}_2+2{\mathrm{H}}_2\mathrm{O} $$

(10)

Effect of gas composition on two-stage system

In order to understand the effect of gas composition on the ratio of NO₂/NO, Fig. 10a–c shows the variation of NO₂/NO ratio with different gas compositions. The complicated gas compositions are added (including 10% CO₂, 0.5% CO, and 300 ppm C₂H₄) individually, for test. The concentrations of different compositions are chosen based on typical exhaust of diesel engine. As shown in Fig. 10, the NO₂/NO ratio still can reach 1 for different gas compositions. With the existence of 300 ppm C₂H₄, 0.5% CO, and 10% CO₂, the applied voltage needed to achieve NO₂/NO = 1 increases to 16.5, 16.5, and 17.5 kV, respectively, if compared with the NO + O₂ only. The gas composition with CO₂ has the most effect on the NO₂/NO ratio because the concentration of CO₂ is the highest than others. In addition, CO₂ belongs to electronegative characteristic gas which further forms negative ions such as CO₂ ⁻ to inhibit gas discharge. Generally, existence of electronegative characteristic gas may consume energetic electron in NTP system, resulting in the reduction of discharge performance and a lower ratio of NO₂/NO. Figure 10d shows that the variation of NO₂/NO ratio with applied voltage when C₂H₄, CO, CO₂, SO₂, and H₂O are present in the flue gas simultaneously. Obviously, the applied voltage has to be increased raise from 15.5 to 18 kV to achieve the fast-SCR condition (NO₂/NO ≈ 1).

Fig. 10

Variation of NO₂/NO with different gas composition a NO + O₂ + CO, b NO + O₂ + C₂H₄, c NO + O₂ + CO₂, d NO + O₂ + CO+ CO₂ + C₂H₄ + SO₂ + H₂O

To evaluate the performance of two-stage system to reduce NO_x to N₂ with complicated gas compositions, the reaction condition is as follows: 300 ppm NO, 300 ppm NH₃, 300 ppm C₂H₄, 0.5% CO, 10% CO₂, 2% H₂O_(g), and 100 ppm SO₂. Results indicate that 300 ppm NO in complicated gas compositions could be converted into 156 ppm NO and 152 ppm NO₂ with the applied voltage of 18 kV and frequency of 10 kHz (as presented in Fig. 10d). As shown in Fig. 11, the NO_x reduction performance reveals an increased trend with increasing temperature and NO_x conversion reaches 90% at 100 °C. The two-stage system of this study shows better performance for SCR process if compared with literature (as summarize at Table 4).

Fig. 11

Dependence of NO_x reduction activity on temperature of two-stage system

Table 4 Comparison of performance of the NTP enhanced SCR with literature

In addition, it is found that SO₂ concentration is reduced from 100 to 63 ppm as the gases pass through the NTP reactor, and it may be oxidized into SO₃ or H₂SO₄ by O, OH, or O₃. Relevant study indicated that the removal process of SO₂ involves oxidation of SO₂ by O, OH, and O₃, forming SO₃ and H₂SO₄ in the NTP system (Obradović et al. 2011). The results indicate that NO_x conversion reaches 95% at 150 °C by two-stage system and it has a good tolerance for H₂O_(g) and SO₂ if compared with catalyst-alone. In the two-stage system, NTP system plays an essential role which can effectively oxidize NO to NO₂ for fast-SCR process. Goo et al. (2007) use V₂O₅-WO₃-MnO₂/TiO₂ with NH₃ as a reducing agent to conduct conventional SCR and fast-SCR for NO removal with the temperature ranging from 200 to 400 °C. The results indicate that fast-SCR has better tolerance for H₂O_(g) if compared with conventional SCR. They suggest that the effect of water presence is marginal due to high reaction rate of fast SCR (Goo et al. 2007). In addition, they point out that >90% conversion could be achieved with fast-SCR (with V₂O₅-WO₃-MnO₂/TiO₂ as catalyst) in the presence of 8% H₂O_(g) and 100 ppm SO₂ at 200 °C. Based on literature and the results obtained, it is speculated that fast-SCR has better NO conversion in the presence of H₂O_(g) and SO ₂ due to high reaction rate. Also, addition of another reducing agent (C₂H₄) in simulated gas streams may help increase NO conversion. These results indicate that the two-stage system has a high NO_x conversion at a low temperature, and it can operate not only at simple gas composition but also complicated gas condition. The two-stage plasma catalysis system has good potential to improve the activity of NO reduction at a low temperature.

Conclusions

Conventional SCR process still faces some challenges, for example, it needs to be operated within the temperature range of 250–350 °C for high efficiency. SCR catalyst suffers from deactivation, because conventional SCR process is typically installed before dust removal equipment and flue gas desulfurization (FGD) due to temperature limit. In addition, conventional SCR process cannot be applied for effective removal of NO_x from diesel engine exhaust due to its low temperature (100–200 °C). In this study, NTP and Mn-based catalyst are combined to form a two-stage system for NO reduction. First, three Mn-based catalysts including Mn-Ce/TiO₂, Mn-Ce-Ni/TiO₂, and Mn-Ce-Cu/TiO₂ are evaluated for SCR performance. Experimental results indicate that Mn-Ce-Ni/TiO₂ owns the best activity for SCR reaction among three Mn-based catalysts prepared. However, Mn-Ce-Ni/TiO₂ still needs to be operated at 150 °C for 100% NO_x conversion, and it also generates significant N₂O. Further, NTP is placed upstream of SCR system to transform part of NO into NO₂ for achieving fast-SCR. Results indicate that the NO_x conversion and N₂ selectivity achieved can reach 100 and 80%, respectively, and the N₂O concentration measured is lower than that of catalysis alone at 100 °C. The condition of NO₂/NO = 1 can be achieved for fast SCR by applying appropriate voltage with plasma system. This study demonstrates that combining Mn-based catalysts with NTP can successfully achieve fast SCR reaction to reduce NO_x to N₂, with a high efficiency and N₂ selectivity. This study indicates that plasma-assisted system can improve the performance of SCR catalyst, and it has high performance at a low temperature (≦200 °C).

Nomenclature

Item

10% wt. Mn-Ce/TiO2 = Mn-Ce/TiO2

10% wt. Mn-Ce-Ni/TiO2 = Mn-Ce-Ni/TiO2

10% wt. Mn-Ce-Cu/TiO2 = Mn-Ce-Cu/TiO2

Dielectric barrier discharge = DBD

Nonthermal plasma = NTP

Selective catalytic reduction = SCR

SCR system + NTP = two-stage system