Support effect of Mn-based catalysts for gaseous elemental mercury oxidation and adsorption

Abstract

Mn-Based catalysts with a Mn loading of 4 wt% were prepared using an impregnation method. Their Hg⁰ removal efficiencies were in the following order: 4% Mn/MK10 (montmorillonite K 10) > 4% Mn/SiO₂ > 4% Mn/TiO₂ > 4% Mn/Al₂O₃. Results from various characterization techniques, such as BET, XRD, STEM, H₂-TPR, Hg⁰-TPD and XPS, suggested that the species, morphologies, and distributions of MnO_x led to the different performances in Hg⁰ removal. The support effect on Hg⁰ adsorption and oxidation was studied. For 4% Mn/Al₂O₃ and 4% Mn/SiO₂, physical adsorption dominates Hg⁰ removal at low temperature (150 °C), while chemical adsorption is more important at high temperature (350 °C). At both low and high temperatures, chemical adsorption and oxidation play leading roles in 4% Mn/TiO₂ and 4% Mn/MK10, respectively. The effect of MnO_x species and their morphologies on Hg⁰ oxidation was investigated. Amorphous MnO₂ benefits Hg⁰ oxidation at both low and high temperatures, while amorphous Mn₂O₃ only facilitates Hg⁰ oxidation at high temperature. Hg⁰ oxidation on Mn-based catalysts follows a Mars–Maessen mechanism where the lattice oxygen of MnO_x reacts with the absorbed Hg⁰.

---

1. Introduction

Mercury emission from coal-fired flue gas is a hazardous atmospheric pollutant. Elemental mercury (Hg⁰), the primary mercury component in the gaseous phase, is especially hard to remove due to its high vapor pressure and low water solubility.^1,2 Numerous efforts have been made to remove gaseous Hg⁰, with adsorption and catalytic oxidation methods being considered the most effective ones.^3,4 Many previous studies focused more on Hg⁰ removal than Hg⁰ oxidation, principally because most of the used materials were adsorbents rather than catalysts.^5–10 Nevertheless, some shortcomings, such as the limited adsorption capacity and high energy consumption for regeneration, limit the development of adsorbents. In contrast, catalysts are much more selective and recyclable than adsorbents, and therefore, catalytic oxidation is a more promising method to remove Hg⁰.

There have been some reports on Hg⁰ oxidation in recent years.^11–18 Among the most promising materials, transition metal oxides have been extensively studied.^13–18 Manganese oxide (MnO_x), which has multiple oxidation states and can participate in various redox reactions, is particularly conducive to Hg⁰ oxidation.^16–18 Yang et al.¹⁹ reported that the Mn⁴⁺ active sites played an important role in Hg⁰ oxidation over Fe–Mn spinel catalysts. Li et al.²⁰ found that the dominance of Mn⁴⁺ on the Mn–Ce/Ti catalyst was mainly responsible for its excellent catalytic performance. Moreover, several studies have shown that it is the amorphous or poorly crystalline Mn⁴⁺ that is beneficial to Hg⁰ oxidation.^21–23 That is, the species and morphologies of MnO_x can significantly influence the Hg⁰ oxidation performance.

According to Cimino's study,²⁴ Mn/TiO₂ showed a much better Hg⁰ capture efficiency than Mn/Al₂O₃, because of the higher proportion of Mn⁴⁺ sites on TiO₂. This means that the valence and content of active components could be strongly affected by the supports. Results from Zhang et al.²⁵ and Ding et al.²⁶ also suggest the potential influence of supports on the active components, which eventually affects the Hg⁰ adsorption or oxidation. Therefore, a suitable support is vital for effectively controlling the species and morphologies of MnO_x, and further promoting Hg⁰ adsorption and oxidation.

Montmorillonite, a common Si–Al-based mineral that is widely used in adsorption and catalysis fields (such as the removal of VOCs,^27,28 CO₂,^29,30 and SO₂ (ref. 31)), has barely been used in Hg⁰ removal. Our previous work has proved the Hg⁰ removal ability of Mn/MK10 (montmorillonite K 10) in flue gas.³² The characteristics of MK10, such as layered structure, porosity, high surface acidity, and excellent thermal stability, all facilitate Hg⁰ adsorption and oxidation.³³ In the present work, we further demonstrate that Mn/MK10 has much better Hg⁰ removal performance than Mn/TiO₂, Mn/Al₂O₃, and Mn/SiO₂. However, both SiO₂ and Al₂O₃ are also the primary components of MK10, and TiO₂ is often employed as an efficacious support for Hg⁰ removal. Therefore, here we focus on the effects of various supports on Mn-based catalysts for Hg⁰ adsorption and oxidation, and provide new theories for developing Hg⁰ removal materials.

In this work, Mn-based catalysts were prepared using an impregnation method. The Hg⁰ adsorption and oxidation performances over Mn/MK10 in the temperature range of 100–400 °C were compared with those over Mn/TiO₂, Mn/Al₂O₃, and Mn/SiO₂. The species and morphologies of MnO_x over various Mn-based catalysts were analyzed by the Brunauer–Emmett–Teller (BET), X-ray diffraction (XRD), high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), hydrogen temperature-programmed reduction (H₂-TPR), Hg temperature-programmed desorption (Hg-TPD), and X-ray photoelectron spectroscopy (XPS) methods in order to clarify the support effect and the Hg⁰ oxidation mechanism.

2. Experimental

2.1. Catalyst preparation

Mn-Based catalysts with a Mn loading of 4 wt% were synthesized using an impregnation method. Briefly, requisite quantities of Mn(NO₃)₂ precursor and supports (TiO₂, Al₂O₃, SiO₂, or MK10) were homogeneously mixed in deionized water and stirred at 25 °C for 4 h. Then, excess water in the mixtures was evaporated at 55 °C using a rotary evaporator, and the obtained solid was dried at 110 °C for 12 h and then calcined at 425 °C for 4 h. The obtained samples were crushed and sieved through 60–80 mesh (180–250 μm) screens to test the catalytic activity.

2.2. Catalyst characterization

The specific surface areas and pore volumes were measured by N₂ adsorption in an automatic porosity analyzer (Autosorb-iQ, Quantachrome). Samples were subjected to XRD using a Rigaku D/Max-RA powder diffractometer with Cu Kα radiation to obtain XRD patterns. HAADF-STEM images were collected on an FEI Tecnai G₂ F20 field emission electron microscope. Before the H₂-TPR test, the catalyst was degassed in Ar at 300 °C for 1 h. After cooling to room temperature (RT), the catalyst was heated from RT to 600 °C at a ramp of 10 °C min⁻¹ in a flow of 10% H₂/Ar. XPS measurement was performed on an ESCALAB 250Xi system using an Al Kα source.

Prior to the Hg-TPD tests, 0.05 g of the catalyst was first exposed to the simulated gas containing 120.0 μg m⁻³ Hg⁰ at either 150 °C or 350 °C for 2 h. Then, the catalyst was purged with N₂ at a flow rate of 500 mL min⁻¹ at 100 °C. When the Hg⁰ concentration at the outlet was reduced to zero, the catalyst was heated from 100 °C to 700 °C in pure N₂ at a heating rate of 10 °C min⁻¹.

2.3. Catalyst activity test

Catalyst activity tests were carried out with 0.05 g of sample in a fixed-bed quartz tube reactor (Fig. 1) that is 4 mm in inner diameter. An Hg⁰ permeation tube was used to generate a gas containing 120.0 μg m⁻³ Hg⁰ with a balance of N₂. The simulated flue gas (Hg⁰ + 5% O₂ + N₂: the conditions from Fig. 3 to 5) was introduced into the reactor at the steady state. To prevent Hg⁰ deposition on the inner surface, all lines through which Hg⁰ passed were heated to 90 °C. Lumex RA915M Zeeman mercury analyzers were used to continuously monitor Hg⁰ concentration, and a mercury concentration mass balance was conducted before each test. The total gas flow rate, space velocity, and steady reaction time were 500 mL min⁻¹, 478 000 h⁻¹, and 10 h, respectively. As shown in Fig. 1, the red dashed box includes two tubular furnaces and two mercury analyzers, the corresponding details of which are shown in Fig. 2. The gaseous Hg⁰ concentration after passing through the catalyst was detected using mercury analyzer I. The sum of the gaseous Hg⁰ and Hg²⁺ concentrations, detected using mercury analyzer II, was obtained by heating the outlet mixed gas to 800 °C in tubular furnace II after passing through mercury analyzer I. Therefore, the total Hg⁰ removal, oxidation, and adsorption efficiencies (η_cap, η_oxi, and η_ads, respectively) are calculated as follows:

η_cap = [(Hg⁰_inlet − Hg⁰_outlet1)/Hg⁰_inlet] × 100% (1)

η_oxi = [(Hg⁰_outlet2 − Hg⁰_outlet1)/Hg⁰_inlet] × 100% (2)

η_ads = η_cap − η_oxi (3)

where Hg⁰_inlet (μg m⁻³) is the concentration of Hg⁰ measured at the inlet of tubular furnace I and Hg⁰_outlet1 (μg m⁻³) and Hg⁰_outlet2 (μg m⁻³) are the concentrations of Hg⁰ measured at the outlet of tubular furnaces I and II, respectively.

Fig. 1 Diagrammatic sketch of the experimental setup.

Fig. 2 Diagrammatic sketch of the tubular furnaces and mercury analyzers.

3. Results and discussion

3.1. Activity tests

A series of activity tests were carried out to examine the ratio of removed Hg⁰ to inlet Hg⁰ over 600 min at different temperatures over various Mn-based catalysts. Fig. 3(a) shows the ratio at various temperatures over pure supports (TiO₂, Al₂O₃, SiO₂, and MK10), with all the values below 10%. In particular, no removed Hg⁰ was observed over the SiO₂ support owing to its inert characteristics. Adding 4 wt% of Mn significantly enhanced the Hg⁰ removal performance of all the supports. As shown in Fig. 3(b), increasing the temperature from 100 °C to 300 °C increased the Hg⁰ removal ability of Mn/TiO₂, Mn/Al₂O₃, and Mn/SiO₂ catalysts. However, further increasing the temperature to 400 °C reduced the ratio. In the case of Mn/MK10, the Hg⁰ removal performance is relatively high and stable across the whole temperature range. Moreover, the ratio of the four Mn-based catalysts can be ordered as follows: 4% Mn/MK10 > 4% Mn/SiO₂ > 4% Mn/TiO₂ > 4% Mn/Al₂O₃, demonstrating that the interactions between the active component and different supports are all different, even with the same loading of Mn.

Fig. 3 The ratio of removed Hg⁰ to inlet Hg⁰ over 600 min at different temperatures over (a) various pure supports and (b) various Mn-based catalysts.

To further illustrate the differences among these Mn-based catalysts, the Hg⁰ oxidation and adsorption performances at various temperatures were examined. In particular, at the representative temperature of 350 °C, two obvious lines representing Hg⁰ breakthrough curves were detected for all four catalysts, where the blue and red lines represent Hg⁰_outlet1 and Hg⁰_outlet2, respectively (Fig. 4). Correspondingly, the areas marked with vertical lines, blanks, and oblique lines represent the adsorbed mercury on the catalyst, gaseous Hg²⁺, and gaseous Hg⁰, respectively, in order to estimate the proportions of Hg⁰ adsorption and Hg⁰ oxidation. The results are summarized in Fig. 5. In the proportions of Hg⁰ adsorption (slash column) and Hg⁰ oxidation (blank column) from 100 °C to 400 °C in Fig. 5, the difference details of these catalysts are much clearer. On the one hand, Hg⁰ removal on 4% Mn/Al₂O₃ mainly depends on adsorption. For 4% Mn/TiO₂ and 4% Mn/SiO₂, although there is a small degree of Hg⁰ oxidation at high temperature, adsorption dominates the overall Hg⁰ removal. In the case of 4% Mn/MK10, however, Hg⁰ oxidation plays the main role, as >60% of the Hg⁰ is oxidized when reacting at above 150 °C. On the other hand, the efficiencies of both Hg⁰ adsorption and Hg⁰ oxidation are improved when the temperature is increased from 100 °C to 300 °C (low-temperature window), while these efficiencies decrease with further increase in the temperature from 300 °C to 400 °C (high-temperature window) for Mn/TiO₂, Mn/Al₂O₃, and Mn/SiO₂. Therefore, in the following experiments, 150 °C and 350 °C were selected to represent the low- and high-temperature windows, respectively. A series of characterization methods, including BET, XRD, STEM, and H₂-TPR, were used to analyze the species, morphologies, and distributions of MnO_x over various Mn-based catalysts.

Fig. 4 Breakthrough curves of Hg⁰ on various Mn-based catalysts (reacted at 350 °C.)

Fig. 5 Hg⁰ adsorption and oxidation efficiencies on various Mn-based catalysts.

3.2. Species and morphologies of MnO_x over Mn-based catalysts

3.2.1. BET analysis. The textural properties, including specific surface area and pore volume, of pure supports and the corresponding Mn-based catalysts are summarized in Table 1. The order of both surface area and pore volume for the Mn-based catalysts are as follows: 4% Mn/SiO₂ > 4% Mn/Al₂O₃ > 4% Mn/MK10 > 4% Mn/TiO₂, in the same order as that for the pure supports. This means that loading 4 wt% Mn onto these supports had little effect on the physical structure.³⁴ Besides, Hg⁰ removal performances over these catalysts in Fig. 5 were found to have little to do with their corresponding textural properties, suggesting that physical adsorption is not the key factor in the different Hg⁰ oxidation and adsorption efficiencies. Nevertheless, Hg⁰ removal performances of 4% Mn/SiO₂ and 4% Mn/Al₂O₃ might be partly influenced by physical adsorption, owing to their relatively large surface area and strong adsorption.^35,36

Table 1 Textural properties of various samples before and after Mn impregnation

Sample	TiO2	Al2O3	SiO2	MK10
BET (m^2 g^−1)	71.7	167.4	372.5	64.2
V pore (cm^3 g^−1)	0.23	0.56	0.74	0.11

---

Sample	4% Mn/TiO2	4% Mn/Al2O3	4% Mn/SiO2	4% Mn/MK10
BET (m^2 g^−1)	67.4	169.9	333.4	88.4
V pore (cm^3 g^−1)	0.22	0.53	1.36	0.14

---

3.2.2. XRD analysis. To identify the structure of Mn species on various samples, XRD measurements were carried out for both pure supports and their Mn-based catalysts. As shown in Fig. 6, the XRD patterns of the pure supports (blue line) exhibit their corresponding characteristic peaks, whereas there are some discrepancies among the characteristic peaks of MnO_x in the corresponding Mn-based catalysts (red line). For instance, the characteristic peaks of MnO₂ (2θ = 28.7°, 37.4°, and 56.8°) and Mn₂O₃ (2θ = 33.0° and 55.2°) were clearly observed for 4% Mn/Al₂O₃, while only the MnO₂ characteristic peaks were present for 4% Mn/TiO₂ and 4% Mn/SiO₂. In contrast, there was no apparent characteristic peak of MnO_x for 4% Mn/MK10. Therefore, MnO_x on the MK10 support is highly dispersed or is present as an amorphous phase, efficiently promoting the catalytic oxidation activity according to other literature studies.^37,38 From the Hg⁰ adsorption and oxidation efficiencies over various Mn-based catalysts in Fig. 5, it is clear that the same Mn loading on different supports can form different MnO_x species, which eventually affect the Hg⁰ adsorption and oxidation.²⁴ In short, the interaction between the active components and supports is the main reason for the different Hg⁰ removal performance on various Mn-based catalysts.^20,39

Fig. 6 XRD patterns of various samples before and after Mn impregnation.

3.2.3. STEM analysis. To directly observe the distribution of MnO_x species over the various catalysts, STEM measurements were conducted, and the results are shown in Fig. 7. It was found that MnO_x (green color) agglomerated to form small particles on the TiO₂ support in Fig. 7(a), and a large number of these small crystal clusters were distributed uniformly on TiO₂. Considering the XRD characterization results, these small particles could be clearly assigned to crystalline MnO₂. Similarly, numerous small particle clusters with homogeneous distribution can be observed on the Al₂O₃ surface (Fig. 7(b)). However, the particles on the Al₂O₃ support contained not only crystalline MnO₂ but also crystalline Mn₂O₃ on the basis of the XRD patterns. Fig. 7(c) shows the distribution of MnO_x on the SiO₂ support. In contrast to Fig. 7(a) and (b), MnO_x in Fig. 7(c) agglomerated into relatively large particles, and their distribution on SiO₂ was not very uniform. This reveals that a large amount of crystalline MnO₂ aggregated together, and the agglomerated particles are hardly attached to SiO₂. In other words, there is no strong interaction between MnO₂ and the SiO₂ support, mainly due to the inert characteristics of SiO₂.⁴⁰ According to Fig. 7(d), MnO_x is evenly distributed on the surface of MK10. Unlike the other three Mn-based catalysts, only a very small amount of particle clusters appears on MK10, while most MnO_x is uniformly dispersed on the support surface in an amorphous state. This is not only in accordance with the XRD results, but also confirms that amorphous MnO_x is more conducive to Hg⁰ oxidation.^21–23,37

Fig. 7 STEM patterns of the various Mn-based catalysts.

3.2.4. H₂-TPR analysis. To further investigate the reducibility of MnO_x over the various Mn-based catalysts, H₂-TPR tests were performed for both pure supports and the corresponding Mn-based catalysts. Firstly, the results of the pure supports are shown in Fig. 8 as blue lines. There were no obvious reduction peaks for pure Al₂O₃ and SiO₂ in the temperature range of 25–600 °C, highlighting their inert nature as supports.⁴¹ In contrast, the reduced characteristic peaks at 550 °C and 500 °C were observed in pure TiO₂ and MK10, respectively. This suggests their relatively active role as supports compared to Al₂O₃ and SiO₂.³⁴

Fig. 8 H₂-TPR patterns of the various samples before and after Mn impregnation.

Fig. 8 also shows the H₂-TPR results of various Mn-based catalysts as red lines. Two distinct reduction characteristic peaks were observed in Mn/TiO₂, Mn/SiO₂, and Mn/MK10. Basically, these peaks at low and high temperatures were attributed to the reduction from Mn⁴⁺ to Mn³⁺ and Mn³⁺ to Mn²⁺, respectively.^24,42 It is worth noting that there are some differences in the peak positions for the Mn-based catalysts. Firstly, three reduction characteristic peaks appeared on 4% Mn/Al₂O₃. The intermediate peak might result from the conversion from Mn³⁺ (Mn₂O₃) to Mn^8/3+ (Mn₃O₄),⁴³ as the XRD results showed the existence of crystalline Mn³⁺. Therefore, these three reduction peaks on 4% Mn/Al₂O₃ correspond to the reduction from MnO₂ to Mn₂O₃, Mn₂O₃ to Mn₃O₄, and Mn₃O₄ to MnO. Secondly, the characteristic peaks in 4% Mn/SiO₂ were apparently lower in temperature than those in other Mn-based catalysts, but were similar to pure MnO_x in other literature studies.^44,45 This is primarily ascribed to the poor interaction between MnO_x and SiO₂. Consequently, the peaks at lower temperatures are caused by the existence of independent MnO_x. In other words, the H₂-TPR result of 4% Mn/SiO₂ shows the reduction peak of MnO_x, and is hardly related to the SiO₂ support.

It is known that calculating the TOF (turnover frequency) of the catalyst is a useful method to find out the best catalytic system among various catalysts. In this study, the TOFs of the four Mn-based catalysts at different temperatures (150 °C or 350 °C) were calculated. The H₂ consumption and Mn dispersion of each Mn-based catalyst can be easily determined by comparing the H₂-TPR areas of the catalysts and the quantitative reference material (pure CuO). Therefore, the TOFs of these catalysts were calculated and are summarized in Table 2. As shown in Table 2, the TOFs (s⁻¹) of Mn/MK10 are tens or even hundreds of times higher than those of the other three catalysts, indicating that Mn/MK10 is the best catalytic system among these catalysts. Additionally, the orders of TOFs at both 150 °C and 350 °C are the same as the orders of Hg⁰ removal efficiencies over these catalysts, which illustrates that the order of the TOFs is meaningful data to determine the ability of Hg⁰ oxidation among various catalysts.

Table 2 The TOFs (s⁻¹) of the various Mn-based catalysts at different temperatures

	4% Mn/TiO2	4% Mn/Al2O3	4% Mn/SiO2	4% Mn/MK10
150 °C	4.25 × 10^−14	3.34 × 10^−14	1.25 × 10^−13	2.47 × 10^−12
350 °C	6.98 × 10^−13	2.92 × 10^−14	5.02 × 10^−13	2.73 × 10^−12

---

3.3. Support effect on Hg⁰ adsorption and oxidation

To comprehensively investigate the support effect on Hg⁰ oxidation and adsorption on the various Mn-based catalysts, Hg-TPD tests were performed after reaction at different temperatures, and the results are shown in Fig. 9 (blue lines: reacted at 150 °C, red lines: reacted at 350 °C).

Fig. 9 Hg-TPD patterns of the various Mn-based catalysts after reaction at 150 °C and 350 °C.

To qualitatively analyze Hg⁰ oxidation and adsorption and to further prove the production of gaseous Hg²⁺ on the various Mn-based catalysts, Table 3 summarizes the contents of Hg⁰ removed and desorbed before and after Hg-TPD tests. Before the tests, the catalysts were first exposed to simulated gas with 120.0 μg m⁻³ Hg⁰ for 2 h. Thus, the inlet content of Hg⁰ is 240.0 μg h m⁻³. The content of removed Hg⁰ (A) is the difference between the inlet and the remaining contents of Hg⁰ during the Hg⁰ removal process. The adsorbed Hg⁰ on the catalyst is gradually desorbed with the temperature increasing from 100 °C to 700 °C during the Hg-TPD tests. Thus, the content of desorbed Hg⁰ (B) is the integral of the desorbed Hg⁰ concentration during the Hg-TPD tests. Therefore, the difference between A and B can be regarded as the content of gas-phase Hg²⁺ (C) produced during the Hg⁰ removal process. In this case, qualitative rather than quantitative analysis of the table data is appropriate.

Table 3 Contents of Hg⁰ adsorption and desorption over Mn-based catalysts

	After reaction at 150 °C			After reaction at 350 °C
(μg h m^−3)	A	B	C	A	B	C
4% Mn/TiO2	186.4	186.4	0	198.7	157.5	41.2
4% Mn/Al2O3	16.6	16.6	0	166.4	166.4	0
4% Mn/SiO2	118.2	118.2	0	209.3	145.8	33.5
4% Mn/MK10	231.5	24.7	206.8	234.6	4.7	229.9

---

As shown in Fig. 9(a), there are two desorption peaks at 350 °C and 436 °C on 4% Mn/TiO₂ after reaction at 150 °C (low temperature), while only one characteristic peak at 437 °C appears after reaction at 350 °C (high temperature). All three peaks are ascribed to adsorbed Hg²⁺, as the BET surface area of 4% Mn/TiO₂ is only 67.4 m² g⁻¹ in Table 1. The desorption peak of physical adsorption on Mn-based catalysts usually appears at around 200 °C,^10,46 meaning that chemical adsorption dominates Hg⁰ removal at both low and high temperatures on 4% Mn/TiO₂. Nevertheless, a small amount of gaseous Hg²⁺ was formed at 350 °C (Table 3), which means that there is a small degree of Hg⁰ oxidation at high temperature (consistent with the results in Fig. 5).

As shown in Fig. 9(b), the desorption peak only appears at 210 °C on 4% Mn/Al₂O₃ after reaction at 150 °C, implying that physical adsorption plays a key role in Hg⁰ removal. As the reaction temperature increases to 350 °C, the characteristic peaks move to 337 °C and 436 °C. Both peaks are assigned to adsorbed Hg²⁺, revealing the leading role of chemical adsorption at high temperature over 4% Mn/Al₂O₃. That is, increasing the reaction temperature from 150 °C to 350 °C can lead to the transformation from physical adsorption into chemical adsorption. Furthermore, there is no formation of gaseous Hg²⁺ according to both the Hg-TPD results in Table 3 and the efficiency results in Fig. 5. Thus, there is no Hg⁰ oxidation on 4% Mn/Al₂O₃.

As shown in Fig. 9(c), 4% Mn/SiO₂ displayed two strong desorption peaks at 225–255 °C and 475–485 °C after reaction at different temperatures, which are ascribed to physically adsorbed Hg⁰ and chemically adsorbed Hg²⁺, respectively. The changes in the peak intensities and areas are quite obvious when the reaction temperature increases from 150 °C to 350 °C. Consequently, the Hg⁰ removal mechanisms over 4% Mn/SiO₂ are different at different temperatures. Apparently, the reaction at 150 °C is dominated by physical adsorption, since the intensity and area of the lower characteristic peak are much larger than those of the higher characteristic peak.

However, the peak area at 476 °C is much larger when 4% Mn/SiO₂ is reacted at 350 °C, so chemical adsorption plays the leading role at high temperature. Besides, a small part of gaseous Hg²⁺ is formed from comparing the contents of Hg⁰ removed and desorbed (Table 3), illustrating a small extent of Hg⁰ oxidation at high temperature, just as the results in Fig. 5.

As shown in Fig. 9(d), the desorption peak at 242 °C after 4% Mn/MK10 was reacted at 150 °C should have been assigned to physical adsorption. In this case, however, it is attributed to the adsorbed Hg²⁺ instead, because of the excellent Hg⁰ oxidation ability of 4% Mn/MK10 at 150 °C (Fig. 5) and the small BET surface area (Table 1). With the reaction temperature increasing to 350 °C, the characteristic peak moved to 427 °C while the peak intensity and area decreased. This phenomenon implies that more gaseous Hg²⁺ is produced at 350 °C, further demonstrating the strong oxidation of Hg⁰ over 4% Mn/MK10. Combining these results with the data in Table 3, we conclude that catalytic oxidation dominates Hg⁰ removal on 4% Mn/MK10.

Based on the above discussion, it is obvious that various supports can lead to obviously different Hg⁰ removal mechanisms. Therefore, the support effect is one of the key factors that can significantly influence Hg⁰ oxidation and adsorption. More specifically, physical adsorption dominates Hg⁰ removal at low temperature, while chemical adsorption is more important at high temperature over 4% Mn/Al₂O₃ and 4% Mn/SiO₂. At both temperatures, chemical adsorption and oxidation play the main roles for 4% Mn/TiO₂ and 4% Mn/MK10, respectively.

3.4. Effect of species and morphologies of MnO_x on Hg⁰ oxidation

To clarify the change of valence state and the relative proportion of the Mn element on the surface of Mn-based catalysts, XPS was adopted to investigate these catalysts before and after Hg⁰ removal at 150 °C and 350 °C. The Mn2p results over different catalysts are shown in Fig. 10 and Table 4. Just like other Mn-containing catalysts, the Mn2p region of the catalysts in the present study included Mn2p_1/2 with a binding energy of 654.5 eV and Mn2p_3/2 with a binding energy of 640–644 eV,^47,48 in which the latter can be further fitted to the three peaks at around 640.7, 642.0, and 643.5 eV, corresponding to Mn²⁺, Mn³⁺, and Mn⁴⁺, respectively.⁴⁹

Fig. 10 XPS patterns of various catalysts before and after reaction at 150 °C and 350 °C.

Table 4 MnO_x species on various catalysts before and after reaction at 150 or 350 °C

	Fresh		150 °C		350 °C
	Mn^4+	Mn^3+	Mn^4+	Mn^3+	Mn^4+	Mn^3+	Mn^2+
4% Mn/TiO2	54.4	45.6	47.1	52.9	55.2	35.4	9.4
4% Mn/Al2O3	47.9	52.1	46.1	53.9	42.4	57.7	—
4% Mn/SiO2	55.7	44.3	53.5	46.5	56.2	35.0	8.8
4% Mn/MK10	88.5	11.5	58.9	41.1	57.1	42.9	—

---

As shown in Fig. 10(a) and Table 4, the main Mn species in a fresh 4% Mn/TiO₂ catalyst are crystalline Mn⁴⁺ (54.4%) and amorphous Mn³⁺ (45.6%). This is similar to the values in 4% Mn/Al₂O₃ and 4% Mn/SiO₂, which contain 47.9% and 55.7% crystalline Mn⁴⁺ (Fig. 10(b) and (c)), respectively. However, 88.5% of the Mn was amorphous Mn⁴⁺ in 4% Mn/MK10 as seen in Fig. 10(d), being much higher than those for the other three Mn-based catalysts.

After reaction at 150 °C, about 6.9% of the crystalline Mn⁴⁺ was converted to Mn³⁺ on 4% Mn/TiO₂, indicating the strong chemical adsorption. In other words, crystalline Mn⁴⁺ facilitates the chemical adsorption of Hg⁰ at low temperature. Meanwhile, there is hardly any conversion of crystalline Mn⁴⁺ after reaction at 150 °C on both 4% Mn/Al₂O₃ and 4% Mn/SiO₂, due to the strong physical adsorption at low temperature over these catalysts according to the Hg-TPD results. Furthermore, around 29.6% of the amorphous Mn⁴⁺ was reduced to Mn³⁺ after reaction at 150 °C, revealing the strong Hg⁰ oxidation performance of 4% Mn/MK10. That is, the high Hg⁰ oxidation efficiency of 4% Mn/MK10 is ascribed to the large amount of amorphous Mn⁴⁺.^37,50

After reaction at 350 °C, 31.4% of the amorphous Mn⁴⁺ was reduced to Mn³⁺ in 4% Mn/MK10, also demonstrating the strong Hg⁰ oxidation ability of amorphous Mn⁴⁺ at high temperature. The percentage of amorphous Mn³⁺ decreased, while that of Mn⁴⁺ hardly changed in both 4% Mn/TiO₂ and 4% Mn/SiO₂. Nevertheless, 9.4% and 8.8% Mn²⁺ was produced in these two catalysts, respectively, implying the transformation from amorphous Mn³⁺ into Mn²⁺. In combination with the Hg⁰ removal efficiencies and the previous discussions, we conclude that amorphous Mn³⁺ is good for Hg⁰ oxidation at high temperature. Besides, the chemical adsorption over 4% Mn/Al₂O₃ at high temperature is in good agreement with the results in Fig. 10(b) and Table 4 (i.e., about 3.7% of the crystalline Mn⁴⁺ was transformed into Mn³⁺ after reaction at 350 °C).

Therefore, the species and morphologies of MnO_x play important roles in Hg⁰ oxidation. Clearly, amorphous MnO₂ is beneficial to Hg⁰ oxidation and amorphous Mn₂O₃ is also good for Hg⁰ oxidation at high temperature, whereas chemical adsorption is mainly promoted by crystalline MnO₂.

3.5. Mechanism of Hg⁰ oxidation

From the above studies, the catalytic mechanism of Hg⁰ oxidation at different temperature windows can be explained by the Mars–Maessen mechanism. In this mechanism, adsorbed Hg⁰ would react with the active oxygen atom produced from MnO_x to form HgO. Based on the characterization results and discussion, not only the species and the morphologies of MnO_x but also the temperature could influence Hg⁰ oxidation. The Hg⁰ oxidation mechanism can be explained by the following reactions:

Hg⁰(g) → Hg⁰(ads) (4)

when both MnO₂ and Mn₂O₃ are amorphous, and MnO₂ is dominant in MnO_x, the oxidation of Hg⁰ at both low- and high-temperature windows will mainly result from the reduction of MnO₂:

Hg⁰(ads) + 2MnO₂ → Hg⁰(ads) + Mn₂O₃ (5)

HgO(ads) → HgO(g) (6)

Mn₂O₃ + 1/2O₂(g) → 2MnO₂ (7)

when MnO₂ is crystalline while Mn₂O₃ is amorphous, there will be a small degree of Hg⁰ oxidation occurring at the high-temperature window because of the reduction of Mn₂O₃:

Hg⁰(ads) + Mn₂O₃ → HgO(ads) + 2MnO (8)

HgO(ads) → HgO(g) (9)

2MnO + 1/2O₂(g) → Mn₂O₃ (10)

Therefore, the overall reaction pathway of Hg⁰ oxidation can be described as follows: Hg⁰ is firstly adsorbed on the active sites of Mn-based catalysts to form Hg⁰ (ads), and then one active oxygen atom is lost from amorphous MnO_x to form HgO. The missing lattice oxygen in amorphous MnO_x would be replenished by oxygen from the flue gas. Thus, the catalytic oxidation reaction can be carried out continuously.

From all of the above results and discussion, the support effect and Hg⁰ removal mechanisms over various Mn-based catalysts at different temperatures can be summarized in Fig. 11.

Fig. 11 Hg⁰ oxidation mechanisms of Mn-based catalysts.

4. Conclusions

The order of Hg⁰ removal efficiency for various Mn-based catalysts is found to be: 4% Mn/MK10 > 4% Mn/SiO₂ > 4% Mn/TiO₂ > 4% Mn/Al₂O₃. The species, morphologies and distributions of MnO_x are responsible for the differences in Hg⁰ removal. The support effects on Hg⁰ adsorption and oxidation were studied. Physical adsorption dominates Hg⁰ removal at low temperature, while chemical adsorption is more important at high temperature on 4% Mn/Al₂O₃ and 4% Mn/SiO₂. At both temperatures, chemical adsorption and oxidation play the leading roles in 4% Mn/TiO₂ and 4% Mn/MK10, respectively. Furthermore, the effects of the species and morphologies of MnO_x on Hg⁰ oxidation are as follows. Amorphous MnO₂ is beneficial to Hg⁰ oxidation, while amorphous Mn₂O₃ is only good for Hg⁰ oxidation at high temperature. A likely reaction pathway for Hg⁰ oxidation over Mn-based catalysts is the Mars–Maessen mechanism where the lattice oxygen of MnO_x reacts with the absorbed Hg⁰. As a novel Mn-based catalyst with high Hg⁰ oxidation efficiency, the Mn/MK10 catalyst will be further examined in order to find the origin of its large amount of amorphous MnO₂.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Key Research and Development Plan (No. 2017YFC0210501) and the Youth Innovation Promotion Association CAS.

References

1. N. Pirrone, P. Costa, J. M. Pacyna and R. Ferrara, Atmos. Environ., 2001, 35, 2997–3006.
2. L. Zhao, C. Li, X. Zhang, G. Zeng, J. Zhang and Y. Xie, Catal. Sci. Technol., 2015, 5, 3459–3472.
3. S. Straube, T. Hahn and H. Koeser, Appl. Catal., B, 2008, 79, 286–295.
4. S. Zhao, Y. Ma, Z. Qu, N. Yan, Z. Li, J. Xie and W. Chen, Catal. Sci. Technol., 2014, 4, 4036–4044.
5. J. He, G. K. Reddy, S. W. Thiel, P. G. Smirniotis and N. G. Pinto, Energy Fuels, 2013, 27, 4832–4839.
6. M. M. Wiatros-Motyka, C. G. Sun, L. A. Stevens and C. E. Snape, Fuel, 2013, 109, 559–562.
7. T. M. Bisson, Z. Xu, R. Gupta, Y. Maham, Y. Liu, H. Yang, I. Clark and M. Pate, Fuel, 2013, 108, 54–59.
8. X. Fan, C. Li, G. Zeng, X. Zhang, S. Tao, P. Lu, S. Li and Y. Zhao, Fuel Process. Technol., 2012, 104, 325–331.
9. D. Ballestero, C. Gómez-Giménez, E. García-Díezb, R. Juanb, B. Rubiob and M. T. Izquierdob, J. Hazard. Mater., 2013, 260, 247–254.
10. F. Scala, C. Anacleria and S. Cimino, Fuel, 2013, 108, 13–18.
11. N. Fernández-Miranda, M. A. Lopez-Anton, M. Díaz-Somoano and M. R. Martínez-Tarazona, Chem. Eng. J., 2016, 285, 77–82.
12. S. Zhao, Z. Qu, N. Yan, Z. Li, W. Zhu, J. Pan, J. Xu and M. Li, RSC Adv., 2015, 5, 30841–30850.
13. H. Li, S. Wu, L. Li, J. Wang, W. Ma and K. Shih, Catal. Sci. Technol., 2015, 5, 5129–5138.
14. C. H. Chiu, H. C. His and C. C. Lin, Fuel Process. Technol., 2014, 126, 138–144.
15. R. Stolle, H. Koeser and H. Gutberlet, Appl. Catal., B, 2014, 144, 486–497.
16. D. Jampaiah, S. J. Ippolito, Y. M. Sabri, J. Tardio, P. R. Selvakannan, A. Nafady, B. M. Reddy and S. K. Bhargava, Catal. Sci. Technol., 2016, 6, 1792–1803.
17. B. Zhao, X. Liu, Z. Zhou, H. Shao and M. Xu, Chem. Eng. J., 2016, 284, 1233–1241.
18. D. Jampaiah, S. J. Ippolito, Y. M. Sabri, B. M. Reddy and S. K. Bhargava, Catal. Sci. Technol., 2015, 5, 2913–2924.
19. S. Yang, Y. Guo, N. Yan, D. Wu, H. He, Z. Qu, C. Yang, Q. Zhou and J. Jia, ACS Appl. Mater. Interfaces, 2011, 3, 209–217.
20. H. Li, C. Y. Wu, Y. Li and J. Zhang, Appl. Catal., B, 2012, 111–112, 381–388.
21. A. Zhang, Z. Zhang, H. Lu, Z. Liu, J. Xiang, C. Zhou, W. Xing and L. Sun, Ind. Eng. Chem. Res., 2015, 54, 2930–2939.
22. C. H. Chiu, H. C. His, H. P. Lin and T. C. Chang, J. Hazard. Mater., 2015, 291, 1–8.
23. Y. Liu, Q. Zhong and X. Liu, J. Hazard. Mater., 2015, 283, 252–259.
24. S. Cimino and F. Scala, Ind. Eng. Chem. Res., 2016, 55, 5133–5138.
25. L. Zhang, Y. Zhuo, W. Du, Y. Tao, C. Chen and X. Xu, Ind. Eng. Chem. Res., 2012, 51, 5192–5298.
26. F. Ding, Y. Chun, L. Mi, H. Li, Y. Li and J. Zhang, Ind. Eng. Chem. Res., 2012, 51, 3039–3047.
27. I. Fatimah and T. Huda, Appl. Clay Sci., 2013, 74, 115–120.
28. D. Chen, M. Huang, S. He, S. He, L. Ding, Q. Wang, S. Yu and S. Miao, Appl. Clay Sci., 2016, 119, 109–115.
29. M. Tahir and N. S. Amin, Appl. Catal., B, 2013, 142–143, 512–522.
30. L. Stevens, K. Williams, W. Y. Han, T. Drage, J. Wood and J. Wang, Chem. Eng. J., 2013, 215–216, 699–708.
31. D. Olszewska, Fuel Process. Technol., 2011, 92, 2412–2419.
32. Y. Wu, W. Xu, Y. Yang, M. Shao, T. Zhu and L. Tong, RSC Adv., 2016, 6, 104294–104302.
33. B. S. Kumar, A. Dhakshinamoorthy and K. Pitchumani, Catal. Sci. Technol., 2014, 4, 2378–2396.
34. J. Xie, N. Yan, S. Yang, Z. Qu, W. Chen, W. Zhang, K. Li, P. Liu and J. Jia, Res. Chem. Intermed., 2012, 38, 2511–2522.
35. P. Wang, S. Su, J. Xiang, F. Cao, L. Sun, S. Hu and S. Lei, Chem. Eng. J., 2015, 225, 68–75.
36. Z. Tan, J. Xiang, S. Su, H. Zeng, C. Zhou, L. Sun, S. Hu and J. Qiu, J. Hazard. Mater., 2012, 239–240, 160–166.
37. Z. Wu, B. Jiang and Y. Liu, Appl. Catal., B, 2008, 79, 347–355.
38. Z. Liu, X. Li, J. Y. Lee and T. B. Bolin, Chem. Eng. J., 2015, 275, 1–7.
39. H. Li, C. Y. Wu, Y. Li and J. Zhang, Environ. Sci. Technol., 2011, 45, 7394–7400.
40. A. K. Llunga and R. Meijboom, Appl. Catal., B, 2017, 203, 505–514.
41. W. Du, L. Yin, Y. Zhuo, Q. Xu, L. Zhang and C. Chen, Fuel Process. Technol., 2015, 131, 403–408.
42. A. Zhang, W. Zheng, J. Song, S. Hu, Z. Liu and J. Xiang, Chem. Eng. J., 2014, 236, 29–38.
43. G. K. Reddy, J. He, S. W. Thiel, N. G. Pinto and P. G. Smirniotis, J. Phys. Chem. C, 2015, 119, 8634–8644.
44. J. Xie, H. Xu, Z. Qu, W. Huang, W. Chen, Y. Ma, S. Zhao, P. Liu and N. Yan, J. Colloid Interface Sci., 2014, 428, 121–127.
45. D. Jampaiah, K. M. Tur, P. Venkataswamy, S. J. Ippolito, Y. M. Sabri, J. Tardio, S. K. Bhargava and B. M. Reddy, RSC Adv., 2015, 5, 30331–30341.
46. H. Xu, Z. Qu, C. Zong, W. Huang, F. Quan and N. Yan, Environ. Sci. Technol., 2015, 49, 6823–6830.
47. Y. Xie, C. Li, L. Zhao, G. Zeng, X. Zhang, W. Zhang and S. Tao, Appl. Surf. Sci., 2015, 333, 59–67.
48. S. Yang, Y. Guo, N. Yan, D. Wu, H. He, J. Xie, Z. Qu and J. Jia, Appl. Catal., B, 2010, 101, 698–708.
49. C. He, B. Shen, J. Chen and J. Cai, Environ. Sci. Technol., 2014, 48, 7891–7898.
50. J. Yang, Y. Zhao, J. Zhang and C. Zheng, Environ. Sci. Technol., 2014, 48, 14837–14843.

---