- Email: [email protected]
Mercury capture by manganese modified copper oxide
ARTICLE IN PRESS
JID: JTICE
[m5G;February 16, 2018;15:22]
Journal of the Taiwan Institute of Chemical Engineers 0 0 0 (2018) 1–6
Contents lists available at ScienceDirect
Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice
Mercury capture by manganese modified copper oxide Ping He∗, Zhong-zhi Zhang, Xiao-long Peng, Jiang Wu, Nai-chao Chen School Energy and Mechanical Engineering, Shanghai University of Electric Power, Shanghai 200090, China
a r t i c l e
i n f o
Article history: Received 2 December 2017 Revised 27 January 2018 Accepted 27 January 2018 Available online xxx Keywords: Manganese Copper oxide CuMnO2 Mercury
a b s t r a c t To increase elemental mercury removal efficiency of CuO without HCl gas, the Mn-modified CuO samples were proposed. In this work, four doping ratios of Mn to CuO were fabricated by a hydrothermal synthesis method. The surface morphologies, crystalline structures and chemical states of as-prepared sorbents were characterized by scanning electron microscope (SEM), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS), respectively. The results showed that all the Mn-modified samples contained the CuMnO2 phase with different contents. The Hg0 removal experimental results indicated that the Mn-modified CuO sorbents had the higher elemental mercury removal efficiency than pure CuO. The pseudo-first-order kinetic model fitted well to the adsorption experimental data. The CuMnO2 content in sorbents had a positive effect on mercury capture. We speculated that CuMnO2 acted as a dominant factor for the mercury removal capacity of Mn-modified CuO sorbents. The high valance Mn oxidized the elemental mercury to oxidized mercury. Moreover, CuO can re-oxidize the low valance Mn to the high valance, resulting in the further increase in Hg0 removal efficiency. © 2018 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
1. Introduction Mercury control technique has considerably attracted international attention due to its toxicity and persistent bioaccumulation [1,2]. Coal-fired power plant is classified as the greatest anthropogenic source of mercury emission [3]. Mercury exists three forms in coal-fired flue gas: elemental (Hg0 ), oxidized (Hg2+ ), and particle-bound (Hgp ). Among these mercury species, Hg2+ and Hgp are relatively easy to remove from flue gas by using typical air pollution control devices (APCDs), such as ESPs (Electrostatic Precipitators) and wet-FGD (Flue Gas Desulfurization). Elemental mercury (Hg0 ), however, is difficult to be captured, because it is insoluble in water and chemical inertness [4]. Copper, a transition metal, has special catalytic effects for mercury oxidation due to its ability to store/release oxygen via the redox shift between Cu2+ and Cu+ . Previous studies [5–8] showed that CuO or CuCl2 has very strong mercury capture capacity in HCl gas, because Cu2+ can make HCl generate much active Cl which can oxidize mercury. Hence, CuO is regarded as a potential catalyst for Hg0 removal under high-concentration HCl environment. However, flue gas from coal-fired power plant in China only has a low concentration of HCl gas. Hence, fabricating an effective catalyst to oxide mercury in few HCl or no HCl gas becomes very important. Lattice oxygen has been previously suggested as a versatile oxidant
∗
Corresponding author. E-mail address: [email protected] (P. He).
for mercury in a Mars–Maessen mechanism [9], and oxides such as manganese oxide were shown to have good capacities for mercury in various gas streams [10–12]. Many pre-treatments for sorbents were performed to avoid the effect of natural flaw of sorbent in order to improve the adsorption capacity [13,14]. The catalyst contained manganese-based oxides also has well performance for mercury removal [15]. Al2 O3 [16] and TiO2 [17] were employed as catalyst supports to disperse MnOx for guaranteeing the high utilization of Mn active sites [11]. Co-MnOx [18], Ce-MnOx [10], graphene oxide [19], and some other elements are used to modify MnOx . Furthermore, manganese oxide also has an enhancing effect on Hg0 oxidization in the presence of NOx [11], some special crystal structure provides a higher catalytic oxidation performance. In this work, we modified copper oxide by manganese via a hydrothermal synthesis method under different ratios of Mn to Cu elements. The effect of Mn on mercury capture of CuO was evaluated by adsorption experiments. 2. Experimental 2.1. Synthesizing Mn-modified CuO nanoparticles Typical hydrothermal synthesis method was used to synthesize Mn-modified CuO nanoparticles in this work. For comparison, the pure CuO nanoparticles were also fabricated. Firstly, CuO·3H2 O and MnSO4 ·H2 O powders were placed in 30 ml deionized water with magnetic stirring, then 10 ml 2.5 mol/l NaOH was dropped into this solution. After 20 min continuous stirring, the solution
https://doi.org/10.1016/j.jtice.2018.01.045 1876-1070/© 2018 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE 2
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
Fig. 1. Schematic diagram of the experimental setup of mercury adsorption.
was warmed at 130 °C for 15 h in an electric oven. Thereafter, the nanoparticles were collected and then washed by deionized water and ethanol for three times. Finally, these nanoparticles were dried at 80 °C for 12 h. Four weigh ratios of MnSO4 ·H2 O to CuO·3H2 O were designed to 0 wt%, 20 wt%, 40 wt%, and 60 wt%, and the total mass of each sample was set as 500 mg. All the drugs were purchased from Guoyao Chemical Reagent Co. Ltd. All chemicals were analytical grade. In order to describe clearly and easily, all the samples were denoted as Samples I–IV, respectively. The as-fabricated samples were characterized by scanning electron microscope (SEM) (Phillips XL30 FEG/NEW), X-ray diffraction (XRD) (Bruker D8 Advance, Germany), and X-ray photoelectron spectroscopy (XPS) (PHI 50 0 0C ESCA System). 2.2. Adsorption tests Fig. 1 shows the schematic of the adsorption setup. The high pure nitrogen gas with a flow of 200 ml/min passed through mercury generating device (PSA 10.536 SIR GALAHAD II, UK) to carry the Hg0 vapor and then mixed the other N2 gas with the flow of 600 ml/min in the mixture. The Hg0 inlet concentration was controlled around 50 μg/m3 . Then, the mixed gas flowed through the fixed-bed reactor where the sample was placed in an adsorption quartz tube with 6 mm inner diameter. The temperature was set at 120 °C, which was equal to the temperature of flue gas emitted from coal-fired power plant. The Hg0 vapor concentration was measured by a mercury analyzer (Lumix 915 M). The remaining gas was absorbed by KMnO4 solution. Finally, it should point out that, in a real flue gas containing SO2 , SO3 , NO, and NO2 , there is the possibility of sorbent poisoning through formation of manganese sulfates or nitrates [9,20], and this will be examined in future work. Here, Hg0 removal efficiency (η) was used to evaluate the mercury capture capacity of sample, which was calculated by the following formula:
η=
Hg0in − Hg0out Hg0in
× 100%
(1)
Where Hg0 in and Hg0 out represent the mercury concentrations at the inlet and outlet of the fixed-bed reactor, respectively. 3. Results and discussion The surface morphologies of the samples, as shown in Fig. 2, were characterized by SEM. Fig. 2(a) indicated that Sample I had a nanosheet structure and the average thickness was 50–60 nm.
For the Mn-modified CuO, the nanosheet particles became smaller with increase of Mn weight ratio, as shown in Fig. 2(b)–(d). Additionally, Fig. 2(d) suggested that when the MnSO4 ·H2 O weight ratio increases to 60%, the original nanosheet changed to the cubic nanoparticles. Fig. 3 showed the XRD pattern of each sample. For Sample I, the XRD curve was readily indexed to the phase of tenorite (JCPDS card No. 72-0629). As compared to Sample I, Sample II generated two small peaks of crednerite (JCPDS card No. 75-1010), indicating that manganese mixed into CuO structure. Thus the new component CuMnO2 was formed and coexisted with CuO. For Sample III, all peaks were consistent with the phase of crednerite. This suggested that a relative highly pure CuMnO2 was fabricated. Sample IV had many phases and became a mixed material, including CuO, MnO2 , CuMnO2 , and Cu1.5 Mn1.5 O4 . MnO2 growth indicated that elemental Mn concentration reached to a saturated level to modify copper oxide. Furthermore, the phase of copper oxide formed again, although its XRD peak was very small. Thus, high Mn concentration was not beneficial for fabricating pure CuMnO2 . The XRD patterns showed that the CuMnO2 content of Sample IV was considerably larger than that of Sample II, indicating that the Mn concentration plays a curial role in fabricating CuMnO2 . Fig. 4 showed the Hg0 removal efficiency of each examined sample. The measured data showed that all the Mn-modified CuO samples had the higher Hg0 removal efficiencies than pure CuO. The average Hg0 removal efficiencies of Samples I–IV were 14.86%, 20.05%, 72.46%, and 60.15% after 3 h adsorption time, indicating that manganese is favorable for the mercury removal due to its strong oxidizing ability. However, the manganese content considerably affected the mercury capture. The highest Hg0 removal efficiency was Sample III, rather than Sample IV that had the highest manganese content. Hence, some components maybe determine the enhanced mercury capture. As mentioned above, the XRD patterns indicated that Sample III had the highest pure CuMnO2 . Meanwhile, Samples II and IV also contained the relatively low content of CuMnO2 component. The results indicated that the higher content of CuMnO2 the catalysts had, the higher mercury removal efficiency they had. Hence, we speculated that the CuMnO2 would play a vital role on the improvement of the Hg0 removal efficiency for Mn-modified CuO sorbent. The Lagergren pseudo-first-order [21] and Lagergren pseudosecond-order [22] models were wildly used in estimation of adsorption kinetics. The pseudo-first-order model is described as:
dC = −K1 (Cm − Cs ) dt
(2)
Where K1 is the constant of speed rate of first-order model (1 h−1 ), Cm (mg/g) is the amount of mercury adsorption at equi-
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
3
Fig. 2. SEM images of (a) Sample I, (b) Sample II, (c) Sample III, and (d) Sample IV.
100
Sample III 60
Sample IV
40
Sample II
20
Sample I
0
Hg removal efficiency(%)
80
0 0
30
60
90
120
150
180
Time (min) Fig. 4. Mercury removal efficiencies of Samples I–IV. Fig. 3. XRD patterns of each sample.
librium, Cs is the amount of mercury adsorption at different times (mg/g) and t is the adsorption time (h). Integration for Eq. (4) applies boundary conditions, Cs = 0 to Cs = Cs at t = 0 to t = t. The pseudo-second-order model is given as:
t 1 t = + 2 Cs Cm K2 × Cm
(3)
Where K2 is the constant of speed rate of second-order model (g/mg/h). Table 1 listed the amount of mercury adsorption at equilibrium measured from experiments and predicted values from models. The linear correlation coefficients (R2 ) of these two models were also shown in the Table 1. The correlation coefficients of these two methods were in the range of from 0.997 to 1 for mercury adsorption by all samples.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE 4
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
Fig. 5. XPS spectra of Cu 2p3 of (a) fresh and (b) spent Sample I, Mn 2p3 of (c) fresh and (d) spent Sample III, Cu 2p3 of (e) fresh and (f) spent Sample III. Table 1 Correlation coefficients of Mercury adsorption obtained by experiment, pseudo-first-order and pseudo-second-order kinetic models. Sample
I II III IV
Experiment
mg/g 0.0192 0.0567 0.0877 0.0794
Pseudo-first-order model
Pseudo-first-order model
K1
Cm
R2
K2
Cm
R2
0.00225 0.00787 0.0 020 0 0.0 040 0
mg/g 0.05767 0.07262 0.29163 0.15476
0.99957 0.99711 0.99991 0.99985
0.01172 0.0478 0.00201 0.00873
mg/g 0.10548 0.11257 0.53981 0.26836
0.99961 0.99837 0.99992 0.99993
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
It was found that the calculated amounts of mercury adsorption at the equilibrium of pseudo-first-order were quite close to those from the experimental values. We suggested that the adsorption of mercury by all the samples fitted well to pseudo-first-order kinetic model. The XPS analysis was conducted to determine the chemical state and the relative proportion of the elements on the surface of the Mn-modified CuO. Fig. 5(a) showed the Cu 2p3 spectra of Sample I. A peak was observed at 933.85 eV, which was close to the XPS spectra of CuO at 933.80 eV [23]. There are two satellite peaks at approximately 934.1 eV and 932.8 eV, which are assigned to Cu2+ and Cu+ , respectively [24,25]. In order to distinguish the effect of Cu2+ and Cu+ on mercury capture, the area ratio of the satellite XPS curve to the total XPS curve was calculated to characterize the proportion in total mass. After Hg adsorption test, the Cu2+ proportion of Sample I (Fig. 5(b)) decreased from 72.7% to 56.14%, indicating some Cu2+ cations transferred to Cu+ cations during mercury removal process. Cu2+ cation exhibited some activity towards elemental mercury, while the 14.86% mercury removal efficiency indicated that pure CuO, without HCl gas, cannot have a satisfied mercury capture capacity. The Mn 2p3 peak of fresh Sample III (Fig. 5(c)) was also analyzed. Three satellite peaks were found at 634.4, 642.3, and 641.0 eV, corresponding to Mn4+ , Mn3+ and Mn2+ [26–28] with area proportions of 70.3%, 22.4% and 7.2%, respectively. High valance Mn (Mn4+ ) was the main state of elemental Mn for the Mn-modified CuO sorbent, indicating that such sorbent had the potential oxidized capacity for mercury. Fig. 5(d) shows the Mn 2p3 spectra of spent Sample III. The Mn4+ , Mn3+ and Mn2+ area proportions were 47.9%, 33.5%, and 18.46%, respectively. This suggested some Mn4+ became Mn3+ or Mn2+ after adsorption test, implying that Hg0 would be oxidized to Hg2+ . In addition, the Cu 2p3 spectra of Sample III (Fig. 5(e)) showed that the area proportions of Cu2+ and Cu+ were 39.7% and 60.2%, respectively. Sample III had the lower Cu2+ cations proportion than sample I. It was speculated that Cu2+ cation would interact with Mn3+ cation and become to Cu+ cation: Cu2+ + Mn3+ →Mn4+ + Cu+ , thus the content of Cu+ cation increased considerably. During the adsorption processing, some Cu2+ cations of Sample III, like Sample I, also oxidized the elemental mercury to oxidized mercury and thus transferred to Cu+ cations, as shown in Fig. 5(f), resulting in the further decrease in the content of Cu2+ cation. Based on the above analysis, Hg0 removal over CuMnO2 catalyst attributes to the catalyst oxidation process. The gaseous mercury was absorbed on the surface of CuMnO2 , which belonged to the physical-adsorption process.
Hg0 (g ) → Hg0 (ads )
(4)
Then, the adsorbed Hg0 was catalysis oxidized to Hg2+ by the transfer of Mn from high valance (Mn4+ ) to low valance (Mn3+ ). In this process, CuO re-oxidized the increased Mn3+ to replenish high valance of Mn (Mn4+ ) which was favorable for the high mercury removal capacity.
Hg0 (ads ) + Mn4+ → Hg2+ (ads ) + Mn3+
(5)
Thereafter, the oxidized mercury reacted with the surface oxygen to form HgO on the CuMnO2 surface. Moreover, CuO has an outstanding oxygen storage and release performance [5], enhancing the oxidized mercury capture. 4. Conclusions In this work, Mn-modified CuO catalysts were fabricated by the hydrothermal synthesis method to explore the effect of Mn content on mercury capture capacity of CuO without HCl gas. The morphologies, crystalline structures and chemical states of the deposited samples were comprehensively investigated. All the Mn-
modified CuO samples had higher elemental mercury removal efficiency than pure CuO. The mercury removal efficiency can improve from 14.86% to 72.46%. The pseudo-first-order kinetic model fitted well to the adsorption experimental data. The phase of CuMnO2 in the Mn-modified CuO catalysts may play a vital role on the enhanced Hg0 removal capacity. The high valance of Mn oxidized the elemental mercury to oxidized mercury. In addition, CuO was benefit for transfer the low valance Mn to the high valance and thus strength the oxidation properties of Mn-modified CuO sorbent. Hence, CuMnO2 , as compared to pure CuO, can be regarded as the good candidate to remove the elemental mercury without HCl gas. Acknowledgments This work is supported by the National Natural Science Foundation of China under Grant no. 51606115, and Natural Science Foundation of Shanghai under Grant no. 16ZR1413500. Declaration of interest None. References [1] Gustin MS, Lindberg SE, Weisberg PJ. An update on the natural sources and sinks of atmospheric mercury. Appl Geochem 2008;23:482–93. [2] Li J, Chen B. An overview of mercury emissions by global fuel combustion: the impact of international trade. Renew Sustain Energy Rev 2016;65:345–55. [3] Pudasainee D, Kim JH, Yoon YS, Seo YC. Oxidation, reemission and mass distribution of mercury in bituminous coal-fired power plants with SCR, CS-ESP and wet FGD. Fuel 2012;93:312–18. [4] Hao Y, Wu S, Pan Y, Li Q, Zhou J, Xu Y, Qian G. Characterization and leaching toxicities of mercury in flue gas desulfurization gypsum from coal-fired power plants in China. Fuel 2016;177:157–63. [5] Du W, Yin L, Zhuo Y, Xu Q, Zhang L, Chen C. Performance of CuOx – neutral Al2 O3 sorbents on mercury removal from simulated coal combustion flue gas. Fuel Process Technol 2015;131:403–8. [6] Xu W, Wang H, Zhou X, Zhu T. CuO/TiO2 catalysts for gas-phase Hg0 catalytic oxidation. Chem Eng J 2014;243:380–5. [7] Yamaguchi A, Akiho H, Ito S. Mercury oxidation by copper oxides in combustion flue gases. Powder Technol 2008;180:222–6. [8] Mushtaq M, Bhatti HN, Iqbal M, Noreen S. Eriobotrya japonica seed biocomposite efficiency for copper adsorption: isotherms, kinetics, thermodynamic and desorption studies. J Environ Manag 2016;176:21–33. [9] Granite EJ, Pennline HW, Hargis RA. Novel sorbents for mercury removal from flue gas. Ind Eng Chem Res 20 0 0;39:1020–9. [10] Ma Y, Mu B, Yuan D, Zhang H, Xu H. Design of MnO2 /CeO2 –MnO2 hierarchical binary oxides for elemental mercury removal from coal-fired flue gas. J Hazard Mater 2017;333:186–93. [11] Wang P, Su S, Xiang J, Cao F, Sun L, Hu S, Lei S. Catalytic oxidation of Hg0 by CuO–MnO2 –Fe2 O3 /γ -Al2 O3 catalyst. Chem Eng J 2013;225:68–75. [12] Xu H, Qu Z, Zhao S, Mei J, Quan F, Yan N. Different crystal-forms of one-dimensional MnO2 nanomaterials for the catalytic oxidation and adsorption of elemental mercury. J Hazard Mater 2015;299:86–93. [13] Ehsan A, Bhatti HN, Iqbal M, Noreen S. Native, acidic pre-treated and composite clay efficiency for the adsorption of dicationic dye in aqueous medium. Water Sci Technol 2017;75:753–64. [14] Rashid A, Bhatti HN, Iqbal M, Noreen S. Fungal biomass composite with bentonite efficiency for nickel and zinc adsorption: a mechanistic study. Ecol Eng 2016;91:459–71. [15] Li J, Yan N, Qu Z, Qiao S, Yang S, Guo Y, Liu P, Jia J. Catalytic oxidation of elemental mercury over the modified catalyst Mn/α -Al2 O3 at lower temperatures. Environ Sci Technol 2010;44:426–31. [16] Qiao SH, Chen J, Li JF, Qu Z, Liu P, Yan NQ, Jia JQ. Adsorption and catalytic oxidation of gaseous elemental mercury in flue gas over MnOx /alumina. Ind Eng Chem Res 2009;48:3317–22. [17] Ji L, Sreekanth PM, Smirniotis PG, Thiel SW, Pinto NG. Manganese oxide/titania materials for removal of NOx and elemental mercury from flue gas. Energy Fuels 2008;22:2299–306. [18] Zhang A, Zheng W, Song J, Hu S, Liu Z, Xiang J. Cobalt manganese oxides modified titania catalysts for oxidation of elemental mercury at low flue gas temperature. Chem Eng J 2014;236:29–38. [19] Sreeprasad TS, Maliyekkal SM, Lisha KP, Pradeep T. Reduced graphene oxide-metal/metal oxide composites: facile synthesis and application in water purification. J Hazard Mater 2011;186:921–31. [20] Presto AA, Granite EJ. Impact of sulfur oxides on mercury capture by activated carbon. Environ Sci Technol 2007;41:6579–84.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
5
JID: JTICE 6
ARTICLE IN PRESS
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
[21] Bernal MP, Lopez-Real JM. Natural zeolites and sepiolite as ammonium and ammonia adsorbent materials. Bioresour Technol 1993;43:27–33. [22] Ho YS, McKay G. Pseudo-second order model for sorption processes. Process Biochem 1999;34:451–65. [23] Strohmeier BR, Levden DE, Field RS, Hercules DM. Surface spectroscopic characterization of CuAl2 O3 catalysts. J Catal 1985;94:514–30. [24] Nefedov VI, Firsov MN, Shaplygin IS. Electronic structures of MRhO2 , MRh2 O4 , RhMO4 and Rh2 MO6 on the basis of X-ray spectroscopy and ESCA data. J Electron Spectrosc 1982;26:65–78.
[25] Ghijsen J, Tjeng H, van Elp L, Eskes J, Westerink H, Sawatzky J, A, GCzyzyk M. Electronic structure of Cu2 O and CuO. Phys Rev B Condens Matter 1988;38:11322–30. [26] Di Castro V, Polzonetti G. XPS study of MnO oxidation. J Electron Spectrosc 1989;48:117–23. [27] Umezawa Y, N Reilley C. Effect of argon ion bombardment on metal complexes and oxides studied by X-ray photoelectron spectroscopy. Anal Chem 1978;50:1290–5. [28] Veprˇek S, Cocke DL, Kehl S, Oswald HR. Mechanism of the deactivation of Hopcalite catalysts studied by XPS, ISS, and other techniques. J Catal 1986;100:250–63.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
JID: JTICE
[m5G;February 16, 2018;15:22]
Journal of the Taiwan Institute of Chemical Engineers 0 0 0 (2018) 1–6
Contents lists available at ScienceDirect
Journal of the Taiwan Institute of Chemical Engineers journal homepage: www.elsevier.com/locate/jtice
Mercury capture by manganese modified copper oxide Ping He∗, Zhong-zhi Zhang, Xiao-long Peng, Jiang Wu, Nai-chao Chen School Energy and Mechanical Engineering, Shanghai University of Electric Power, Shanghai 200090, China
a r t i c l e
i n f o
Article history: Received 2 December 2017 Revised 27 January 2018 Accepted 27 January 2018 Available online xxx Keywords: Manganese Copper oxide CuMnO2 Mercury
a b s t r a c t To increase elemental mercury removal efficiency of CuO without HCl gas, the Mn-modified CuO samples were proposed. In this work, four doping ratios of Mn to CuO were fabricated by a hydrothermal synthesis method. The surface morphologies, crystalline structures and chemical states of as-prepared sorbents were characterized by scanning electron microscope (SEM), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS), respectively. The results showed that all the Mn-modified samples contained the CuMnO2 phase with different contents. The Hg0 removal experimental results indicated that the Mn-modified CuO sorbents had the higher elemental mercury removal efficiency than pure CuO. The pseudo-first-order kinetic model fitted well to the adsorption experimental data. The CuMnO2 content in sorbents had a positive effect on mercury capture. We speculated that CuMnO2 acted as a dominant factor for the mercury removal capacity of Mn-modified CuO sorbents. The high valance Mn oxidized the elemental mercury to oxidized mercury. Moreover, CuO can re-oxidize the low valance Mn to the high valance, resulting in the further increase in Hg0 removal efficiency. © 2018 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
1. Introduction Mercury control technique has considerably attracted international attention due to its toxicity and persistent bioaccumulation [1,2]. Coal-fired power plant is classified as the greatest anthropogenic source of mercury emission [3]. Mercury exists three forms in coal-fired flue gas: elemental (Hg0 ), oxidized (Hg2+ ), and particle-bound (Hgp ). Among these mercury species, Hg2+ and Hgp are relatively easy to remove from flue gas by using typical air pollution control devices (APCDs), such as ESPs (Electrostatic Precipitators) and wet-FGD (Flue Gas Desulfurization). Elemental mercury (Hg0 ), however, is difficult to be captured, because it is insoluble in water and chemical inertness [4]. Copper, a transition metal, has special catalytic effects for mercury oxidation due to its ability to store/release oxygen via the redox shift between Cu2+ and Cu+ . Previous studies [5–8] showed that CuO or CuCl2 has very strong mercury capture capacity in HCl gas, because Cu2+ can make HCl generate much active Cl which can oxidize mercury. Hence, CuO is regarded as a potential catalyst for Hg0 removal under high-concentration HCl environment. However, flue gas from coal-fired power plant in China only has a low concentration of HCl gas. Hence, fabricating an effective catalyst to oxide mercury in few HCl or no HCl gas becomes very important. Lattice oxygen has been previously suggested as a versatile oxidant
∗
Corresponding author. E-mail address: [email protected] (P. He).
for mercury in a Mars–Maessen mechanism [9], and oxides such as manganese oxide were shown to have good capacities for mercury in various gas streams [10–12]. Many pre-treatments for sorbents were performed to avoid the effect of natural flaw of sorbent in order to improve the adsorption capacity [13,14]. The catalyst contained manganese-based oxides also has well performance for mercury removal [15]. Al2 O3 [16] and TiO2 [17] were employed as catalyst supports to disperse MnOx for guaranteeing the high utilization of Mn active sites [11]. Co-MnOx [18], Ce-MnOx [10], graphene oxide [19], and some other elements are used to modify MnOx . Furthermore, manganese oxide also has an enhancing effect on Hg0 oxidization in the presence of NOx [11], some special crystal structure provides a higher catalytic oxidation performance. In this work, we modified copper oxide by manganese via a hydrothermal synthesis method under different ratios of Mn to Cu elements. The effect of Mn on mercury capture of CuO was evaluated by adsorption experiments. 2. Experimental 2.1. Synthesizing Mn-modified CuO nanoparticles Typical hydrothermal synthesis method was used to synthesize Mn-modified CuO nanoparticles in this work. For comparison, the pure CuO nanoparticles were also fabricated. Firstly, CuO·3H2 O and MnSO4 ·H2 O powders were placed in 30 ml deionized water with magnetic stirring, then 10 ml 2.5 mol/l NaOH was dropped into this solution. After 20 min continuous stirring, the solution
https://doi.org/10.1016/j.jtice.2018.01.045 1876-1070/© 2018 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE 2
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
Fig. 1. Schematic diagram of the experimental setup of mercury adsorption.
was warmed at 130 °C for 15 h in an electric oven. Thereafter, the nanoparticles were collected and then washed by deionized water and ethanol for three times. Finally, these nanoparticles were dried at 80 °C for 12 h. Four weigh ratios of MnSO4 ·H2 O to CuO·3H2 O were designed to 0 wt%, 20 wt%, 40 wt%, and 60 wt%, and the total mass of each sample was set as 500 mg. All the drugs were purchased from Guoyao Chemical Reagent Co. Ltd. All chemicals were analytical grade. In order to describe clearly and easily, all the samples were denoted as Samples I–IV, respectively. The as-fabricated samples were characterized by scanning electron microscope (SEM) (Phillips XL30 FEG/NEW), X-ray diffraction (XRD) (Bruker D8 Advance, Germany), and X-ray photoelectron spectroscopy (XPS) (PHI 50 0 0C ESCA System). 2.2. Adsorption tests Fig. 1 shows the schematic of the adsorption setup. The high pure nitrogen gas with a flow of 200 ml/min passed through mercury generating device (PSA 10.536 SIR GALAHAD II, UK) to carry the Hg0 vapor and then mixed the other N2 gas with the flow of 600 ml/min in the mixture. The Hg0 inlet concentration was controlled around 50 μg/m3 . Then, the mixed gas flowed through the fixed-bed reactor where the sample was placed in an adsorption quartz tube with 6 mm inner diameter. The temperature was set at 120 °C, which was equal to the temperature of flue gas emitted from coal-fired power plant. The Hg0 vapor concentration was measured by a mercury analyzer (Lumix 915 M). The remaining gas was absorbed by KMnO4 solution. Finally, it should point out that, in a real flue gas containing SO2 , SO3 , NO, and NO2 , there is the possibility of sorbent poisoning through formation of manganese sulfates or nitrates [9,20], and this will be examined in future work. Here, Hg0 removal efficiency (η) was used to evaluate the mercury capture capacity of sample, which was calculated by the following formula:
η=
Hg0in − Hg0out Hg0in
× 100%
(1)
Where Hg0 in and Hg0 out represent the mercury concentrations at the inlet and outlet of the fixed-bed reactor, respectively. 3. Results and discussion The surface morphologies of the samples, as shown in Fig. 2, were characterized by SEM. Fig. 2(a) indicated that Sample I had a nanosheet structure and the average thickness was 50–60 nm.
For the Mn-modified CuO, the nanosheet particles became smaller with increase of Mn weight ratio, as shown in Fig. 2(b)–(d). Additionally, Fig. 2(d) suggested that when the MnSO4 ·H2 O weight ratio increases to 60%, the original nanosheet changed to the cubic nanoparticles. Fig. 3 showed the XRD pattern of each sample. For Sample I, the XRD curve was readily indexed to the phase of tenorite (JCPDS card No. 72-0629). As compared to Sample I, Sample II generated two small peaks of crednerite (JCPDS card No. 75-1010), indicating that manganese mixed into CuO structure. Thus the new component CuMnO2 was formed and coexisted with CuO. For Sample III, all peaks were consistent with the phase of crednerite. This suggested that a relative highly pure CuMnO2 was fabricated. Sample IV had many phases and became a mixed material, including CuO, MnO2 , CuMnO2 , and Cu1.5 Mn1.5 O4 . MnO2 growth indicated that elemental Mn concentration reached to a saturated level to modify copper oxide. Furthermore, the phase of copper oxide formed again, although its XRD peak was very small. Thus, high Mn concentration was not beneficial for fabricating pure CuMnO2 . The XRD patterns showed that the CuMnO2 content of Sample IV was considerably larger than that of Sample II, indicating that the Mn concentration plays a curial role in fabricating CuMnO2 . Fig. 4 showed the Hg0 removal efficiency of each examined sample. The measured data showed that all the Mn-modified CuO samples had the higher Hg0 removal efficiencies than pure CuO. The average Hg0 removal efficiencies of Samples I–IV were 14.86%, 20.05%, 72.46%, and 60.15% after 3 h adsorption time, indicating that manganese is favorable for the mercury removal due to its strong oxidizing ability. However, the manganese content considerably affected the mercury capture. The highest Hg0 removal efficiency was Sample III, rather than Sample IV that had the highest manganese content. Hence, some components maybe determine the enhanced mercury capture. As mentioned above, the XRD patterns indicated that Sample III had the highest pure CuMnO2 . Meanwhile, Samples II and IV also contained the relatively low content of CuMnO2 component. The results indicated that the higher content of CuMnO2 the catalysts had, the higher mercury removal efficiency they had. Hence, we speculated that the CuMnO2 would play a vital role on the improvement of the Hg0 removal efficiency for Mn-modified CuO sorbent. The Lagergren pseudo-first-order [21] and Lagergren pseudosecond-order [22] models were wildly used in estimation of adsorption kinetics. The pseudo-first-order model is described as:
dC = −K1 (Cm − Cs ) dt
(2)
Where K1 is the constant of speed rate of first-order model (1 h−1 ), Cm (mg/g) is the amount of mercury adsorption at equi-
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
3
Fig. 2. SEM images of (a) Sample I, (b) Sample II, (c) Sample III, and (d) Sample IV.
100
Sample III 60
Sample IV
40
Sample II
20
Sample I
0
Hg removal efficiency(%)
80
0 0
30
60
90
120
150
180
Time (min) Fig. 4. Mercury removal efficiencies of Samples I–IV. Fig. 3. XRD patterns of each sample.
librium, Cs is the amount of mercury adsorption at different times (mg/g) and t is the adsorption time (h). Integration for Eq. (4) applies boundary conditions, Cs = 0 to Cs = Cs at t = 0 to t = t. The pseudo-second-order model is given as:
t 1 t = + 2 Cs Cm K2 × Cm
(3)
Where K2 is the constant of speed rate of second-order model (g/mg/h). Table 1 listed the amount of mercury adsorption at equilibrium measured from experiments and predicted values from models. The linear correlation coefficients (R2 ) of these two models were also shown in the Table 1. The correlation coefficients of these two methods were in the range of from 0.997 to 1 for mercury adsorption by all samples.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE 4
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
Fig. 5. XPS spectra of Cu 2p3 of (a) fresh and (b) spent Sample I, Mn 2p3 of (c) fresh and (d) spent Sample III, Cu 2p3 of (e) fresh and (f) spent Sample III. Table 1 Correlation coefficients of Mercury adsorption obtained by experiment, pseudo-first-order and pseudo-second-order kinetic models. Sample
I II III IV
Experiment
mg/g 0.0192 0.0567 0.0877 0.0794
Pseudo-first-order model
Pseudo-first-order model
K1
Cm
R2
K2
Cm
R2
0.00225 0.00787 0.0 020 0 0.0 040 0
mg/g 0.05767 0.07262 0.29163 0.15476
0.99957 0.99711 0.99991 0.99985
0.01172 0.0478 0.00201 0.00873
mg/g 0.10548 0.11257 0.53981 0.26836
0.99961 0.99837 0.99992 0.99993
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
ARTICLE IN PRESS
JID: JTICE
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
It was found that the calculated amounts of mercury adsorption at the equilibrium of pseudo-first-order were quite close to those from the experimental values. We suggested that the adsorption of mercury by all the samples fitted well to pseudo-first-order kinetic model. The XPS analysis was conducted to determine the chemical state and the relative proportion of the elements on the surface of the Mn-modified CuO. Fig. 5(a) showed the Cu 2p3 spectra of Sample I. A peak was observed at 933.85 eV, which was close to the XPS spectra of CuO at 933.80 eV [23]. There are two satellite peaks at approximately 934.1 eV and 932.8 eV, which are assigned to Cu2+ and Cu+ , respectively [24,25]. In order to distinguish the effect of Cu2+ and Cu+ on mercury capture, the area ratio of the satellite XPS curve to the total XPS curve was calculated to characterize the proportion in total mass. After Hg adsorption test, the Cu2+ proportion of Sample I (Fig. 5(b)) decreased from 72.7% to 56.14%, indicating some Cu2+ cations transferred to Cu+ cations during mercury removal process. Cu2+ cation exhibited some activity towards elemental mercury, while the 14.86% mercury removal efficiency indicated that pure CuO, without HCl gas, cannot have a satisfied mercury capture capacity. The Mn 2p3 peak of fresh Sample III (Fig. 5(c)) was also analyzed. Three satellite peaks were found at 634.4, 642.3, and 641.0 eV, corresponding to Mn4+ , Mn3+ and Mn2+ [26–28] with area proportions of 70.3%, 22.4% and 7.2%, respectively. High valance Mn (Mn4+ ) was the main state of elemental Mn for the Mn-modified CuO sorbent, indicating that such sorbent had the potential oxidized capacity for mercury. Fig. 5(d) shows the Mn 2p3 spectra of spent Sample III. The Mn4+ , Mn3+ and Mn2+ area proportions were 47.9%, 33.5%, and 18.46%, respectively. This suggested some Mn4+ became Mn3+ or Mn2+ after adsorption test, implying that Hg0 would be oxidized to Hg2+ . In addition, the Cu 2p3 spectra of Sample III (Fig. 5(e)) showed that the area proportions of Cu2+ and Cu+ were 39.7% and 60.2%, respectively. Sample III had the lower Cu2+ cations proportion than sample I. It was speculated that Cu2+ cation would interact with Mn3+ cation and become to Cu+ cation: Cu2+ + Mn3+ →Mn4+ + Cu+ , thus the content of Cu+ cation increased considerably. During the adsorption processing, some Cu2+ cations of Sample III, like Sample I, also oxidized the elemental mercury to oxidized mercury and thus transferred to Cu+ cations, as shown in Fig. 5(f), resulting in the further decrease in the content of Cu2+ cation. Based on the above analysis, Hg0 removal over CuMnO2 catalyst attributes to the catalyst oxidation process. The gaseous mercury was absorbed on the surface of CuMnO2 , which belonged to the physical-adsorption process.
Hg0 (g ) → Hg0 (ads )
(4)
Then, the adsorbed Hg0 was catalysis oxidized to Hg2+ by the transfer of Mn from high valance (Mn4+ ) to low valance (Mn3+ ). In this process, CuO re-oxidized the increased Mn3+ to replenish high valance of Mn (Mn4+ ) which was favorable for the high mercury removal capacity.
Hg0 (ads ) + Mn4+ → Hg2+ (ads ) + Mn3+
(5)
Thereafter, the oxidized mercury reacted with the surface oxygen to form HgO on the CuMnO2 surface. Moreover, CuO has an outstanding oxygen storage and release performance [5], enhancing the oxidized mercury capture. 4. Conclusions In this work, Mn-modified CuO catalysts were fabricated by the hydrothermal synthesis method to explore the effect of Mn content on mercury capture capacity of CuO without HCl gas. The morphologies, crystalline structures and chemical states of the deposited samples were comprehensively investigated. All the Mn-
modified CuO samples had higher elemental mercury removal efficiency than pure CuO. The mercury removal efficiency can improve from 14.86% to 72.46%. The pseudo-first-order kinetic model fitted well to the adsorption experimental data. The phase of CuMnO2 in the Mn-modified CuO catalysts may play a vital role on the enhanced Hg0 removal capacity. The high valance of Mn oxidized the elemental mercury to oxidized mercury. In addition, CuO was benefit for transfer the low valance Mn to the high valance and thus strength the oxidation properties of Mn-modified CuO sorbent. Hence, CuMnO2 , as compared to pure CuO, can be regarded as the good candidate to remove the elemental mercury without HCl gas. Acknowledgments This work is supported by the National Natural Science Foundation of China under Grant no. 51606115, and Natural Science Foundation of Shanghai under Grant no. 16ZR1413500. Declaration of interest None. References [1] Gustin MS, Lindberg SE, Weisberg PJ. An update on the natural sources and sinks of atmospheric mercury. Appl Geochem 2008;23:482–93. [2] Li J, Chen B. An overview of mercury emissions by global fuel combustion: the impact of international trade. Renew Sustain Energy Rev 2016;65:345–55. [3] Pudasainee D, Kim JH, Yoon YS, Seo YC. Oxidation, reemission and mass distribution of mercury in bituminous coal-fired power plants with SCR, CS-ESP and wet FGD. Fuel 2012;93:312–18. [4] Hao Y, Wu S, Pan Y, Li Q, Zhou J, Xu Y, Qian G. Characterization and leaching toxicities of mercury in flue gas desulfurization gypsum from coal-fired power plants in China. Fuel 2016;177:157–63. [5] Du W, Yin L, Zhuo Y, Xu Q, Zhang L, Chen C. Performance of CuOx – neutral Al2 O3 sorbents on mercury removal from simulated coal combustion flue gas. Fuel Process Technol 2015;131:403–8. [6] Xu W, Wang H, Zhou X, Zhu T. CuO/TiO2 catalysts for gas-phase Hg0 catalytic oxidation. Chem Eng J 2014;243:380–5. [7] Yamaguchi A, Akiho H, Ito S. Mercury oxidation by copper oxides in combustion flue gases. Powder Technol 2008;180:222–6. [8] Mushtaq M, Bhatti HN, Iqbal M, Noreen S. Eriobotrya japonica seed biocomposite efficiency for copper adsorption: isotherms, kinetics, thermodynamic and desorption studies. J Environ Manag 2016;176:21–33. [9] Granite EJ, Pennline HW, Hargis RA. Novel sorbents for mercury removal from flue gas. Ind Eng Chem Res 20 0 0;39:1020–9. [10] Ma Y, Mu B, Yuan D, Zhang H, Xu H. Design of MnO2 /CeO2 –MnO2 hierarchical binary oxides for elemental mercury removal from coal-fired flue gas. J Hazard Mater 2017;333:186–93. [11] Wang P, Su S, Xiang J, Cao F, Sun L, Hu S, Lei S. Catalytic oxidation of Hg0 by CuO–MnO2 –Fe2 O3 /γ -Al2 O3 catalyst. Chem Eng J 2013;225:68–75. [12] Xu H, Qu Z, Zhao S, Mei J, Quan F, Yan N. Different crystal-forms of one-dimensional MnO2 nanomaterials for the catalytic oxidation and adsorption of elemental mercury. J Hazard Mater 2015;299:86–93. [13] Ehsan A, Bhatti HN, Iqbal M, Noreen S. Native, acidic pre-treated and composite clay efficiency for the adsorption of dicationic dye in aqueous medium. Water Sci Technol 2017;75:753–64. [14] Rashid A, Bhatti HN, Iqbal M, Noreen S. Fungal biomass composite with bentonite efficiency for nickel and zinc adsorption: a mechanistic study. Ecol Eng 2016;91:459–71. [15] Li J, Yan N, Qu Z, Qiao S, Yang S, Guo Y, Liu P, Jia J. Catalytic oxidation of elemental mercury over the modified catalyst Mn/α -Al2 O3 at lower temperatures. Environ Sci Technol 2010;44:426–31. [16] Qiao SH, Chen J, Li JF, Qu Z, Liu P, Yan NQ, Jia JQ. Adsorption and catalytic oxidation of gaseous elemental mercury in flue gas over MnOx /alumina. Ind Eng Chem Res 2009;48:3317–22. [17] Ji L, Sreekanth PM, Smirniotis PG, Thiel SW, Pinto NG. Manganese oxide/titania materials for removal of NOx and elemental mercury from flue gas. Energy Fuels 2008;22:2299–306. [18] Zhang A, Zheng W, Song J, Hu S, Liu Z, Xiang J. Cobalt manganese oxides modified titania catalysts for oxidation of elemental mercury at low flue gas temperature. Chem Eng J 2014;236:29–38. [19] Sreeprasad TS, Maliyekkal SM, Lisha KP, Pradeep T. Reduced graphene oxide-metal/metal oxide composites: facile synthesis and application in water purification. J Hazard Mater 2011;186:921–31. [20] Presto AA, Granite EJ. Impact of sulfur oxides on mercury capture by activated carbon. Environ Sci Technol 2007;41:6579–84.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045
5
JID: JTICE 6
ARTICLE IN PRESS
[m5G;February 16, 2018;15:22]
P. He et al. / Journal of the Taiwan Institute of Chemical Engineers 000 (2018) 1–6
[21] Bernal MP, Lopez-Real JM. Natural zeolites and sepiolite as ammonium and ammonia adsorbent materials. Bioresour Technol 1993;43:27–33. [22] Ho YS, McKay G. Pseudo-second order model for sorption processes. Process Biochem 1999;34:451–65. [23] Strohmeier BR, Levden DE, Field RS, Hercules DM. Surface spectroscopic characterization of CuAl2 O3 catalysts. J Catal 1985;94:514–30. [24] Nefedov VI, Firsov MN, Shaplygin IS. Electronic structures of MRhO2 , MRh2 O4 , RhMO4 and Rh2 MO6 on the basis of X-ray spectroscopy and ESCA data. J Electron Spectrosc 1982;26:65–78.
[25] Ghijsen J, Tjeng H, van Elp L, Eskes J, Westerink H, Sawatzky J, A, GCzyzyk M. Electronic structure of Cu2 O and CuO. Phys Rev B Condens Matter 1988;38:11322–30. [26] Di Castro V, Polzonetti G. XPS study of MnO oxidation. J Electron Spectrosc 1989;48:117–23. [27] Umezawa Y, N Reilley C. Effect of argon ion bombardment on metal complexes and oxides studied by X-ray photoelectron spectroscopy. Anal Chem 1978;50:1290–5. [28] Veprˇek S, Cocke DL, Kehl S, Oswald HR. Mechanism of the deactivation of Hopcalite catalysts studied by XPS, ISS, and other techniques. J Catal 1986;100:250–63.
Please cite this article as: P. He et al., Mercury capture by manganese modified copper oxide, Journal of the Taiwan Institute of Chemical Engineers (2018), https://doi.org/10.1016/j.jtice.2018.01.045