Oxidation and adsorption of gas-phase Hg⁰ over a V₂O₅/AC catalyst

Abstract

Oxidation and adsorption of gas-phase Hg⁰ over a V₂O₅/AC catalyst was studied in N₂ and N₂ + O₂ atmospheres. The results showed that the V₂O₅/AC catalyst had a high Hg⁰ capture capability in N₂ and N₂ + O₂, which can be attributed to the critical role of V₂O₅ oxidation activity for oxidizing Hg⁰ to Hg²⁺ (form HgO). O₂ promoted Hg⁰ oxidation and its role was mainly to resume the oxidation activity of V₂O₅ by replenishing O to the used V₂O₅ sites. It was found that only a portion of the V₂O₅ sites on the V₂O₅/AC surface were effective for Hg⁰ oxidation. The high Hg⁰ capture capability of V₂O₅/AC was due to the combined effects of oxidation and adsorption, and the effect extent of each was different at different temperatures.

---

1. Introduction

Mercury (Hg) is a toxic persistent bioaccumulative pollutant and its emission in flue gas from coal combustion has become a worldwide environmental concern and must be removed before emission.^1,2 Coal combustion is considered to be the primary anthropogenic source of Hg emission, accounting for approximately 30% of the entire emission amount around the world.³ Mercury in coal combustion flue gas is often classified into three main forms: elemental mercury (Hg⁰), oxidized mercury (Hg²⁺) and particle-bound mercury (Hg^p). Hg²⁺ and Hg^p are relatively easy to remove from flue gas using existing air pollution control devices. Hg²⁺ is water-soluble and therefore it can be effectively removed by wet flue gas desulfurization (WFGD) equipment.⁴ Hg^p can be captured in electrostatic precipitators (ESPs) or a fabric filter (FF) along with fly ash particles.⁵ Hg⁰, however, is difficult to capture because of its high volatility and insolubility in water. Therefore, Hg⁰ is the dominant mercury form released to the atmosphere and becomes the key to Hg emission control.

Activated carbon injection, which can remove both Hg⁰ and Hg²⁺, is currently the best commercially available technology for Hg removal from flue gas. However, the major drawbacks of activated carbon are low utilization efficiency, poor capacity, narrow temperature range, high cost and incapable regeneration.^6,7 Obviously, it is required to develop cost-effective technologies for Hg removal from coal combustion flue gas.

Catalytic oxidation Hg⁰ to Hg²⁺ is considered to be a promising method for Hg⁰ removal. In recent years, metal oxides based catalysts have been widely studied for Hg⁰ catalytic oxidation and showed high Hg⁰ oxidation activities, including V₂O₅, MnO₂, CeO₂, Fe₂O₃, CuO, Mo₃O₄, etc.^8–16 It was found that V₂O₅/AC catalyst showed excellent Hg⁰ capture capability at stack temperature in our previous research.¹⁷ The forms of Hg-captured over V₂O₅/AC were HgO and HgSO₄ in the presence of O₂, SO₂ and H₂O. However, the Hg⁰ capture mechanism over V₂O₅/AC was not clear since the Hg⁰ capture process was found involving a number of steps in the multi-components atmosphere. In order to better understand Hg⁰ oxidation and adsorption over V₂O₅/AC, Hg⁰ capture was studied in simple atmosphere N₂ and N₂ + O₂ in this work.

2. Experimental

2.1 Catalyst preparation

The V₂O₅/AC catalysts were prepared by pore volume impregnation of activated coke (AC, 30–60 mesh, BET area = 960 m² g⁻¹, micropore volume = 0.24 cm³ g⁻¹) with an aqueous solution of ammonium metavanadate and oxalic acid. The detailed preparation procedure of V₂O₅/AC catalysts was described in our previous research.¹⁷ All reagents used in this work were of analytical grade and were purchased from Sinopharm Chemical Reagent Co. Ltd. Several catalysts with different V₂O₅ loading (wt%) were prepared and named according to the weight percentage of V₂O₅ in V₂O₅/AC. V1.0/AC referred to V₂O₅/AC catalyst containing 1.0 wt% V₂O₅, for example. The BET area and micropore volume of V1.0/AC were 770 m² g⁻¹ and 0.21 cm³ g⁻¹.

2.2 Hg⁰ oxidation and adsorption experiment

Hg⁰ oxidation and adsorption experiments were carried out in a fixed-bed quartz reactor as shown in Fig. 1. 0.2 g V₂O₅/AC catalyst was used for each experiment in the temperature range of 120 to 180 °C in N₂ and N₂ + O₂ (6.3%) atmospheres. The total flow rate of N₂ or N₂ + O₂ was 105 mL min⁻¹, corresponding to a space velocity of about 15 000 h⁻¹. An Hg⁰ permeation tube (VICI Metronics) was used as a vapor-phase Hg⁰ source and an Hg⁰ concentration of 420.4 μg m⁻³ was generated. All of the Teflon tubes for Hg⁰-containing gas delivery were kept at 120 °C to avoid Hg⁰ adsorption. Hg⁰ oxidation and adsorption experiments lasted from 2 to 24 h. For safety purpose, the tail gas from the reactor was treated before release to the atmosphere. The samples after Hg⁰ capture were sealed in vials at room temperature and analyzed soon.

Fig. 1 Schematic diagram of Hg⁰ oxidation and adsorption experiment. (1) Rotameter; (2) mass flow controller; (3) water bath; (4) gas mixing chamber; (5) heating tape; (6) Hg⁰ permeation tube; (7) quartz reactor; (8) furnace; (9) quartz wool; (10) catalyst; (11) temperature controller; (12) tail gas cleaner.

To identify the roles of V₂O₅ and O₂ in Hg⁰ capture over V₂O₅/AC, O₂-response experiments were also performed in the fixed-bed reactor using V0.1/AC with a much lower V₂O₅ loading. The experimental conditions were the same as those mentioned above except for the switch of atmospheres between N₂ and N₂ + O₂. The Hg⁰ concentrations in the influent and effluent gas were measured online by an atomic fluorescence spectrometer (AFS). To measure the total Hg in the effluent gas, a KBH₄ solution was used to reduce Hg²⁺ to Hg⁰ before the AFS.

2.3 Hg desorption experiment

Temperature programmed desorption (TPD) coupled with an AFS (referred as TPD-AFS) was used to measure Hg release online from the Hg⁰-captured catalyst upon heating to 1000 °C in Ar. About 5 mg Hg⁰-captured catalyst was loaded in a quartz tube reactor, purged for 30 min and then heated up to 1000 °C at a heating rate of 10 °C min in an Ar flow of 100 mL min⁻¹. The Hg⁰ exiting from the reactor was analyzed directly by the AFS while the total Hg was measured by bubbling the effluent gas through a KBH₄ solution to reduce Hg²⁺ to Hg⁰ before the AFS.

2.4 Hg content analysis

The amount of Hg captured by V₂O₅/AC and AC was defined as the difference of Hg content before and after Hg⁰ capture and determined by following the Chinese national standard, GB/T 16659-1996, in which the solid samples were digested by 2H₂SO₄ + 5HNO₃ with V₂O₅ in a sealed vial at 120 °C for 24 h and the dissolved Hg²⁺ was analyzed by an atomic fluorescence spectrometer with KBH₄ as a reducing agent. All the measurements were duplicated and the relative deviation was less than 5%.

2.5 XPS and SEM-EDX analyses

X-ray photoelectron spectroscopy (XPS) analysis was employed to identify the speciation of Hg adsorbed over V₂O₅/AC using an ESCALAB250 spectrometer (Thermo-VG Scientific) with Al Kα source at 10 kV. The binding energies of C 1s and Hg 4f were scanned and calibrated by the C 1s peak at 284.6 eV. The Hg⁰-captured V₂O₅/AC catalysts were also characterized by scanning electron microscope-energy dispersive X-ray (SEM-EDX) technique on a LEO435vp instrument.

3. Results and discussion

3.1 Roles of V₂O₅ and O₂ in Hg⁰ capture over V₂O₅/AC

Fig. 2 compares Hg⁰ capture capabilities of AC, V0.5/AC and V1.0/AC in N₂ and N₂ + O₂ atmospheres at 120 °C. A high Hg⁰ concentration (420.4 μg m⁻³) was used to speed up the experiments.

Fig. 2 Effect of V₂O₅ and O₂ on Hg⁰ capture at 120 °C with an inlet Hg⁰ concentration of 420.4 μg m⁻³.

It can be seen that AC showed a low Hg⁰ capture capability and had reached saturated adsorption capacity in 24 h, about 15 μg g⁻¹-cat in both atmospheres, and O₂ showed a little effect on Hg⁰ capture. All these can be due to the mainly physisorption of Hg⁰ on AC.¹⁸ In contrast to AC, V1.0/AC had a much higher Hg⁰ capture capability and saturation was not reached in 24 h, indicating the important effect of V₂O₅. Furthermore, O₂ obviously promoted Hg⁰ capture over V1.0/AC, from 83 μg g⁻¹-cat in N₂ to 125 μg g⁻¹-cat in N₂ + O₂. These effects were similar to those for SO₂ removal over the same V₂O₅/AC catalysts¹⁹ and can mainly be attributed to the catalytic oxidation activity of V₂O₅.

To confirm the catalytic oxidation activity of V₂O₅ in Hg⁰ capture, XPS analysis was used to identify the forms of Hg captured over V₂O₅/AC and AC. Fig. 3 shows detailed scans around energies corresponding to the Hg 4f spectral region. It can be seen that both of the fresh AC (a) and V1.0/AC (d) showed one peak at around 102 eV, which can be ascribed to Si of SiO₂ in the AC support (SiO₂ content = 7.59%).^20,21 However, V1.0/AC-N₂ (e, Hg⁰ capture in N₂) and V1.0/AC-N₂ + O₂ (f, Hg⁰ capture in N₂ + O₂), in addition to the peak at 102 eV, showed a shoulder peak at around 104.8 eV, which can be attributed to Hg²⁺ of HgO.²² This suggested that V₂O₅ can oxidize Hg⁰ to Hg²⁺ without or with O₂ over V₂O₅/AC, which was consistent with the findings of Hg⁰ oxidation on commercial V₂O₅-based SCR catalysts.^23,24 Furthermore, the higher intensity of the shoulder peak in spectrum (f) than that in spectrum (e) indicated that more HgO was formed in the presence of O₂, which explained the higher Hg⁰ capture amount in N₂ + O₂ in Fig. 2. It should be noted that there was no visible peak around 104.8 eV for Hg⁰-captured AC samples, indicating little Hg²⁺ was formed. Unfortunately, it was unsuccessful to identify the existence of Hg⁰ because of the interferences of Si.

Fig. 3 XPS analyses of Hg for fresh and Hg⁰-captured AC and V1.0/AC.

To further understand the roles of V₂O₅ and O₂ in Hg⁰ capture over V₂O₅/AC, O₂-response experiments were performed using V0.1/AC and the results are shown in Fig. 4. It can be seen that no Hg was detected in the effluence gas for about 10 minutes and then Hg breakthrough started, followed by an Hg concentration increase with time. Hg breakthrough reached about 40% in 83 minutes at which O₂ was added to the feed. Obviously, Hg breakthrough dropped quickly and decreased to 17% within 7 minutes (in 90 minutes), then maintained at this level until O₂ was cut off (in 105 minutes). Subsequently, Hg breakthrough rose up to the level of that at O₂ addition in a short time (in 115 minutes). It is clear that O₂ was important for Hg⁰ capture over V₂O₅/AC. Since no Hg⁰ oxidation by O₂ was measured in the gas phase, the effect of O₂ should be on V₂O₅. In other words, the role of V₂O₅ was to oxidize Hg⁰ to form Hg²⁺, in which V⁵⁺ in V₂O₅ was reduced to V⁴⁺ and lost its oxidation activity, while the role of O₂ was to resume the oxidation activity of V₂O₅ by replenishing O to the used V₂O₅ sites. This was consistent with SO₂ oxidation over the same V₂O₅/AC catalysts¹⁹ and similar to Hg⁰ oxidation on metal oxide catalysts.^23–25

Fig. 4 O₂-response experiment in Hg⁰ capture over V0.1/AC at 120 °C with an inlet Hg⁰ concentration of 420.4 μg m⁻³.

3.2 Usability of V₂O₅ in V₂O₅/AC for Hg⁰ capture

In order to evaluate the usability of V₂O₅ in V₂O₅/AC for Hg⁰ capture, Hg⁰ saturated adsorption (100% Hg⁰ breakthrough) experiments were carried out using five catalysts with low V₂O₅ loading and the results are shown in Fig. 5. ΔHg was the amount of Hg⁰ captured by V₂O₅ sites, i.e. the difference of Hg⁰ saturated adsorption capacity between V₂O₅/AC and AC. It can be seen that ΔHg/V₂O₅ mole ratio was about 5% and 10% for Hg⁰-captured V₂O₅/AC samples in N₂ and N₂ + O₂, respectively. Obviously, V₂O₅ was not fully utilized, indicating that not all of the V₂O₅ sites were effective for Hg⁰ capture. Furthermore, ΔHg/V₂O₅ mole ratio in N₂ + O₂ was approximately twice than that in N₂. Combined with the O₂-response experiments results in Fig. 4, it was likely that the effective V₂O₅ sites can capture Hg⁰ in N₂ and lost their oxidation activity while these used V₂O₅ sites can be resumed by O₂ and capture Hg⁰ once again.

Fig. 5 Mole ratio of ΔHg/V₂O₅ vs. V₂O₅ loading for Hg⁰-captured V₂O₅/AC in N₂ and N₂ + O₂ at 120 °C.

Fig. 6 shows SEM images of Hg⁰-captured V1.0/AC in N₂ (a) and N₂ + O₂ (b). It can be seen that two kinds of positions (marked as 1 and 2) can be distinguished on both of the two samples surface. EDX analysis indicated that positions 1 contained more inorganic salts while positions 2 contained more carbon, in which the corresponding V₂O₅ contents were 0.18% and 0.34%, respectively. For Hg⁰-captured V1.0/AC in N₂, Hg contents in positions 1 and 2 were 0.05% and 0.01%, measured by EDX. These suggested that neither V₂O₅ nor Hg distributed on the sample surface homogeneously and Hg⁰ trended to bond with the V₂O₅ in the position containing more inorganic salts. This trend became more evident when Hg and V₂O₅ contents were measured in tens of samples. As for Hg⁰-captured V1.0/AC in N₂ + O₂, Hg contents in positions 1 and 2, which showed the similar trend to that in the Hg⁰-captured sample in N₂, were 0.13% and 0.02%, respectively. Meanwhile, the Hg contents in the two positions were higher than those in Fig. 6(a), which was consistent with the results in Fig. 2. Furthermore, all these above provided an information that not all of the V₂O₅ sites were useful for Hg⁰ capture.

Fig. 6 SEM images of Hg⁰-captured V1.0/AC in N₂ (a) and N₂ + O₂ (b).

3.3 Role of adsorption in Hg⁰ capture over V₂O₅/AC

To further understand Hg⁰ adsorption and oxidation over V₂O₅/AC, Hg⁰ capture experiments were carried out at various temperatures in the stack temperature range. Fig. 7 compares Hg⁰ capture capabilities of V1.0/AC at 120, 150 and 180 °C. It can be seen that Hg⁰ capture capability of V1.0/AC decreased as an increase of temperature in N₂ atmosphere, indicating that a lower temperature favored Hg⁰ capture. However, it showed a different trend in the presence of O₂ and the highest was at 150 °C, with an Hg⁰ capture amount of 139 μg g⁻¹-cat in 24 h, which was consistent with the findings of Hg⁰ capture in the presence of O₂, SO₂ and H₂O in our previous study.¹⁷ Since Hg⁰ capture over V₂O₅/AC was the combination effects of adsorption and oxidation,¹⁷ the difference of Hg⁰ capture with or without O₂ was mainly due to the different effect extent of adsorption and oxidation at different temperature. In other words, adsorption was a more important effect factor than oxidation for Hg⁰ capture in N₂, while it was opposite in N₂ + O₂. Moreover, it is worth to note that the amounts of Hg captured by V1.0/AC at 180 °C were still much higher than those by AC at 120 °C in Fig. 2, indicating the crucial role of V₂O₅ oxidation activity in Hg⁰ capture.

Fig. 7 Hg⁰ capture over V1.0/AC in N₂ and N₂ + O₂ at 120, 150 and 180 °C with an inlet Hg⁰ concentration of 420.4 μg m⁻³.

To clear the effect of adsorption, TPD-AFS was used to measure the release behavior of Hg captured by V1.0/AC upon heating to 1000 °C in Ar. The results are shown in Fig. 8. It can be seen that Hg release started at about 150 °C and appeared a major peak at about 270 °C for both of the two samples. The starting release temperature 150 °C was very important and it explained the decrease of Hg⁰ capture capability as the temperature increased from 150 °C to 180 °C in Fig. 7, i.e. desorption. Furthermore, the Hg release peak at about 270 °C can be attributed to reduction of HgO by AC.^17,26

Fig. 8 Hg release of fresh and Hg⁰-captured V1.0/AC in N₂ and N₂ + O₂.

4. Conclusions

The high Hg⁰ capture capability of V₂O₅/AC was mainly due to the combination effects of oxidation and adsorption. The catalytic oxidation activity of V₂O₅ played a crucial role in Hg⁰ capture and oxidized Hg⁰ to form HgO, in which it was reduced to V₂O₄ and lost its oxidation activity. O₂ had a promotion effect on Hg⁰ capture over V₂O₅/AC by replenishing oxygen to the used V₂O₅ to resume its oxidation activity. The resumed V₂O₅ can oxidize and capture Hg⁰ once again. Not all of the V₂O₅ sites over V₂O₅/AC were useful for Hg⁰ capture, and those dispersed with more inorganic salts on V₂O₅/AC surface are more effective to capture Hg⁰. The increasing temperature had a negative influence on Hg⁰ capture in N₂ whereas there was an optimal temperature near 150 °C in the presence of O₂, above which Hg captured over V₂O₅/AC started to release.

Acknowledgements

The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (21203003, 51404014), Anhui Provincial Natural Science Foundation (1308085QB44), Natural Science Research Key Project of Anhui Provincial Department of Education (KJ2016A860) and Natural Science Research Project of Anhui Provincial Department of Education (AQKJ2015B003). Here, we also express our sincere gratitude to professor Zhenyu Liu and professor Jianli Yang for their help in the written of this paper.

References

1. G. R. Golding, C. A. Kelly, R. Sparling, P. C. Loewen and T. Barkay, Environ. Sci. Technol., 2007, 41, 5685–5692.
2. J. L. Kirk and V. L. Louis, Environ. Sci. Technol., 2009, 43, 2254–2261.
3. E. G. Pacyna, J. M. Pacyna, F. Steenhuisen and S. Wilson, Atmos. Environ., 2006, 40, 4048–4063.
4. M. Díaz-Somoano, S. Unterberger and K. R. Hein, Fuel Process. Technol., 2007, 88, 259–263.
5. J. Wang, W. Wang, W. Xu, X. Wang and S. Zhao, J. Environ. Sci., 2011, 23, 1839–1844.
6. A. P. Jones, J. W. Hoffmann, D. N. Smith, T. J. Feeley and J. T. Murphy, Environ. Sci. Technol., 2007, 41, 1365–1371.
7. S. Sjostrom, M. Durham, C. J. Bustard and C. Martin, Fuel, 2010, 89, 1320–1322.
8. H. M. Xu, Z. Qu, C. X. Zong, F. Q. Quan, J. Mei and N. Q. Yan, Appl. Catal., B, 2015, 186, 30–40.
9. F. Scala and S. Cimino, Chem. Eng. J., 2015, 278, 134–139.
10. W. Xu, H. Wang, X. Zhou and T. Zhu, Chem. Eng. J., 2014, 243, 380–385.
11. C. He, B. X. Shen, J. H. Chen and J. Cai, Environ. Sci. Technol., 2014, 48, 7891–7898.
12. B. Zhao, H. H. Yi, X. L. Tang, Q. Li, D. D. Liu and F. Y. Gao, Chem. Eng. J., 2016, 286, 585–593.
13. Q. Wan, L. Duan, K. He and J. Li, Chem. Eng. J., 2011, 170, 512–517.
14. N. Fernández-Miranda, M. A. Lopez-Anton, M. Díaz-Somoano and M. R. Martínez-Tarazona, Chem. Eng. J., 2016, 285, 77–82.
15. S. H. Qiao, J. Chen, J. F. Li, Z. Qu, P. Liu, N. Q. Yan and J. P. Jia, Ind. Eng. Chem. Res., 2009, 48, 3317–3322.
16. B. Zhao, X. W. Liu, Z. J. Zhou, H. Z. Shao and M. H. Xu, Chem. Eng. J., 2016, 284, 1233–1241.
17. J. W. Wang, J. L. Yang and Z. Y. Liu, Fuel Process. Technol., 2010, 91, 676–680.
18. E. S. Olson, S. J. Miller, R. K. Sharma, G. E. Dunham and S. A. Benson, J. Hazard. Mater., 2000, 74, 61–79.
19. Y. Xiao, Q. Y. Liu, Z. Y. Liu and Z. G. Huang, Appl. Catal., B, 2008, 82, 114–119.
20. C. He, B. X. Shen and F. K. Li, J. Hazard. Mater., 2016, 304, 10–17.
21. Y. M. Sabri, S. J. Ippolito, J. Tardio, P. D. Morrison and S. K. Bhargava, Fuel, 2015, 159, 917–924.
22. P. Humbert, Solid State Commun., 1986, 60, 21–24.
23. S. Straube, T. Hahn and H. Koeser, Appl. Catal., B, 2008, 79, 286–295.
24. Y. Li, P. D. Murphy, C. Wu, K. W. Powers and J. J. Bonzongo, Environ. Sci. Technol., 2008, 42, 5304–5309.
25. H. Li, C. Wu, Y. Li, L. Li, Y. Zhao and J. Zhang, J. Hazard. Mater., 2012, 243, 117–123.
26. S. Keith, Chem. Phys. Lett., 2004, 386, 65–69.

---