Capture of Gas-Phase Arsenic by Ferrospheres ... - ACS Publications

Sep 6, 2016 - The wet magnetic separation is found to be an effective way to extract ferrospheres with a higher Fe content from fly ashes; it can sign...
0 downloads 0 Views 6MB Size
Article pubs.acs.org/EF

Capture of Gas-Phase Arsenic by Ferrospheres Separated from Fly Ashes Kaihua Zhang,*,†,‡ Dongxue Zhang,† Kai Zhang,†,‡ and Yan Cao§ †

Department of Energy Power & Mechanical Engineering and ‡Beijing Key Laboratory of Emission Surveillance and Control for Thermal Power Generation, North China Electric Power University, Beijing 102206, People’s Republic of China § Institute for Combustion Science and Environmental Technology, Chemistry Department, Western Kentucky University, Bowling Green, Kentucky 42101, United States ABSTRACT: Coal combustion may release toxic arsenic into the environment, which seriously threatens human health and ecological safety. Fly ash shows its potentials as an arsenic sorbent in its low cost and no adverse impacts on reutilization but limits in its adsorption capacity for arsenic. On the basis of the iron component in fly ash possessing an obvious effect on arsenic removal, this study tries to isolate ferrospheres with an enriched iron (Fe) content directly from fly ashes for an improved arsenic adsorption. The wet magnetic separation is found to be an effective way to extract ferrospheres with a higher Fe content from fly ashes; it can significantly enrich the iron content to 50% (by weight) in ferrospheres. The comparison of arsenic capture by ferrospheres and fly ashes was made experimentally in a simulated flue gas at 873 K. Results show the Fe-enriched ferrospheres can achieve a higher removal efficiency of 42.75% in this study, which is almost twice that of the raw ashes. With the temperature increasing from 573 to 1073 K, the arsenic capture by ferrospheres was enhanced initially but then decreased. It is likely that the condensation/physical adsorption predominates in temperatures ranging between 573 and 723 K, while the chemical adsorption predominates at further higher temperatures. It is found that the optimal temperature to maintain a high arsenic capture efficiency using ferrospheres is around 873 K. Further investigation based on X-ray diffraction and X-ray photoelectron spectroscopy spectra of the ferrospheres before and after the adsorption reaction confirms that the major contributor to arsenic capture by ferrospheres is the Fe active component in them rather than Si, Al, and Ca compositions. mercury in the flue gas using the Ontario Hydro method at a 300 MW coal-fired power plant in China. The results indicated that As concentrations in the flue gas were 153.27 and 41.13 μg/m3 before and after ESP, while Hg concentrations were about 5.49 and 5.21 μg/m3, respectively. Although ESP units had an obvious synergistic effect on arsenic removal up to 73%, the emitted arsenic in flue gas was still much higher than mercury. This pushes more efforts to be made to control arsenic emission worldwide. However, the comprehensive and detailed studies on As control are quite limited. Only a few sorbents, including calcium-based compound,10−12 Fe2O3,13,14 and activated carbon,15 are reported as effective sorbents to suppress the release of arsenic to the atmosphere. Gas-phase arsenic might react with the added sorbents, then was transformed into larger and easier adsorbed particles, and thus was captured by ESP and other devices. Unfortunately, these specially added sorbents generally have high costs and are difficult to be separated from fly ashes after adsorption, thus bringing adverse impacts on fly ash reutilization. Fly ash is a waste byproduct of coal-fired the power system. Iron and calcium components contained in it are the effective substances for arsenic capture.16,17 Arsenic can be captured by fly ash particles through the formation of iron or calcium arsenates.18−21 Nevertheless, the contents of Fe and Ca in fly

1. INTRODUCTION Arsenic (As), as well as mercury and other hazardous heavy metals, is the most toxic air pollutant that threatens human health and ecological safety.1,2 Coal combustion in power plants for electricity generation is recognized as the major anthropogenic emission source of arsenic. During combustion or gasification of over 1000 °C, arsenic tends to be volatilized and released into flue gas. It is predominantly present in As3+ form as vapor [such as As2O3(g)].3,4 Upon cooling of the flue gas, arsenic vapor tends to concentrate in fine particles of fly ashes.5,6 Thus, some of them can be captured by an electrostatic precipitator (ESP) and other air pollution control devices. Tang et al. investigated the arsenic distribution in feed coal and its combustion byproducts from coal-fired power plants. On the basis of the overall material balances and the operation parameters of tested units, they found that the removal efficiency of As by the ESP unit was 83%, whereas that by the FGD process was 61%, and there was still 6% of As being emitted into the atmosphere.7 Considering the huge amount of coal consumption and the high average arsenic content in coal (about 5 μg/g) all over the world, the total emission of As will be enormous. Thus, it is imminent to effectively control arsenic emission. Tian et al. also reported the emissions of Hg, As, and other heavy metals. They found that the total emissions of Hg and As from coal-fired power plants in China in 2007 were calculated at 132 and 550 t, respectively.8 Arsenic emission in China is more than 4 times as much as mercury. In our previous study,9 we investigated the concentrations of gas-phase arsenic and © 2016 American Chemical Society

Received: July 7, 2016 Revised: August 10, 2016 Published: September 6, 2016 8746

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752

Article

Energy & Fuels

Figure 1. Schematic diagram of the experimental device. First, the samples of FA1, FA2, and FA3 were sieved to obtain particles with a size between 200 and 350 mesh. Second, about 500 g of the three fly ashes with the above size was dispersed in 2 L of water. The mixtures were stirred into slurry solutions, and the magnetic particles in fly ashes were obtained by a high-gradient tube magnet. The product type of the tube magnet was TRE-AM, purchased from DeuMagnet Technology (Shanghai) Co., Ltd. Third, the magnetic particles obtained were transferred into another beaker filled with water and continuously separated under the same magnetic field. This procedure was run repeatedly for about 3−6 times to recover the ferrospheres with a high purity of Fe. Finally, the obtained ferrosphere particles were oven-dried at 373 K for 6 h, further crushed in a carnelian mortar, and stored in the capped polyethylene bottle for use. Three corresponding ferrosphere samples were labeled as FS1, FS2, and FS3, respectively, which differed from their raw fly ash samples (FA1, FA2, and FA3). Previous studies applied the arsenic sources by heating of As2O3(s)18 or the oxidation of AsH3(g) using a hydride generator.10 Alternatively, in this study, the arsenic source was obtained through the vaporization and decomposition of the standard As(V) solution with a higher As(V) concentration of 1000 ± 4 mg/L. The standard As(V) solution was purchased from Inorganic Ventures (Christiansburg, VA), and its starting material is As2O5 and H2O. The As(V) standard solution was evenly dropped into the electric furnace by a mini-sized buret and quickly vaporized. O2(g) was partially released as a result of the instability of As2O5 under elevated temperatures, and thus, the gas-phase arsenic source of a mixture of As2O3(g) and As2O5(g) was generated (as shown in Figure 1). The application of the standard solution with a high As(V) concentration simplified arsenic generation equipment and enhanced safety and accuracy of experimental results. The use of the standard As(V) solution introduced gas impurities into the simulated flue gas at the same time, such as H2O(g) and O2(g), but the impact of them upon adsorption can be ignored, because the total amount of them was only a trace relative to other components in flue gas. 2.2. Adsorption Experiments. The experimental device used for arsenic capture mainly consists of two parts, i.e., the generation of As2O3(g) and As2O5(g) and the adsorption of them (Figure 1). These two parts were integrated in the same quartz tube reactor, which was 2.3 cm in diameter and 35 cm in height. The reactor was installed inside two electric furnaces and heated separately. During the experiments, the As(V) standard solution was dropped into the quartz wool at the upper part of the reactor by a mini-sized buret; the drop rate was carefully controlled by making several pre-experiments

ashes are too limited, which causes a large amount of arsenic in flue gas to not be captured. Therefore, the main objective of this study is to isolate the Fe-enriched ferrospheres directly from fly ashes and further inject them back into the flue gas to enhance arsenic capture. This can make full use of the iron components contained in fly ashes, providing an effective and economic approach on the recovery of the useful product (ferrospheres) directly from fly ashes for arsenic capture. This study will also clarify interaction mechanisms between ferrospheres and As vapors. The sorption experiments were carried out using ferrospheres and fly ashes at temperatures ranging from 573 to 1073 K in simulated flue gas to find out the optimized temperature for arsenic capture and selection of the injection location. Changes of composition and main elements in ferrospheres before and after adsorption were also investigated using X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) to further understand the mechanism for arsenic capture. This research is expected to provide the theory and application guides for cheap and effective control of arsenic pollutants in flue gas only using fly ash itself.

2. EXPERIMENTAL SECTION 2.1. Materials. There were three fly ashes used in this study, FA1, FA2, and FA3. FA2 was taken from a coal-fired power station located in Hebei, China, which was generated from combustion of bituminous coal in a pulverized coal (PC) furnace. FA1 and FA3 were sampled from two different power stations located in Yunnan and Shanxi, China, respectively. They were both obtained from circulating fluidized bed (CFB) boilers; FA1 was generated from combustion of lignite, while FA3 was generated from that of bituminous coal. They were all collected by the ESPs. The samples were naturally air-dried at room temperature for at least 30 days before the experiments. To efficiently separate ferrospheres from fly ashes, two schemes of magnetic separation were investigated, in either dry or wet conditions.22 The effectiveness of separation methods was confirmed by the elemental composition analysis using X-ray fluorescence (XRF) spectrometry. Results revealed that the enriched Fe content of the obtained ferrospheres in a wet condition was much higher than that in a dry condition. Therefore, the wet magnetic separation procedure was applied in this study for the ferrosphere extraction from fly ashes. The details are described as follows: 8747

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752

Article

Energy & Fuels to achieve a uniform speed in the whole reaction time. The movable quartz wool was used to collect the As(V) standard solution to prevent it from directly dropping down on the adsorbent below. The standard solution was quickly vaporized and carried by the simulated flue gas through the fixed sorbent bed located at the lower part of the reactor. Different adsorption temperatures ranging from 573 to 1073 K could be achieved by controlling the furnace. Arsenic not retained in the sorbent was captured in impingers containing HNO3 and NaOH separately (0.5 N). At the beginning of experiments, 400 mg of fly ashes or ferrospheres was weighed and put onto the sorbent bed. Then, the tube reactor was placed inside the two furnaces which had been heated to setting temperatures. The carrier gas was passed through the reactor from top to bottom at a flow rate of 0.2 L min−1. The carrier gas was the simulated air consisting of 79% (v/v) N2 and 21% (v/v) O2. Then, the As(V) standard solution was dropped down to carry out the sorption experiment; its total amount was 0.5 mL. The sorption reaction lasted 1 h. The experimental conditions were listed in Table 1.

Table 2. Chemical Compositions of Fly Ashes and Ferrospheres fly ashes properties (mass %) SiO2 Al2O3 Fe2O3 CaO SO3 TiO2 MgO K2O P2O5 SrO Na2O MnO ZnO

Table 1. Experimental Parameters condition sorbent weight (mg) carrier gas flow rate (L/min) As(V) standard solution feed rate (mL/min) gas-phase arsenic concentration in carrier gas (g/m3) adsorption temperatures (K) adsorption time (min)

ferrospheres

FA1

FA2

FA3

FS1

FS2

FS3

17.93 11.47 8.00 41.19 16.37 0.73 3.22 0.60 0.16 0.21

50.12 29.88 5.01 8.66 1.09 1.33 0.84 1.28 0.30 0.24 0.81 0.11 0.02

41.13 34.34 3.00 12.17 5.75 1.32 1.12 0.42 0.35 0.14 0.10 0.02 0.01

17.59 11.11 50.35 15.83 0.46 0.62 2.78 0.51 0.20 0.15 0.19 0.09 0.02

26.66 16.04 49.89 4.18 0.08 0.47 1.02 0.36 0.19 0.09 0.39 0.53 0.02

28.00 20.74 42.35 5.11 0.42 0.62 1.56 0.42 0.28 0.07 0.11 0.15 0.03

0.05 0.01

parameter 400 0.2 0.0083 0.0417

and FS3, which reached 50.35, 49.89, and 42.35% by mass, respectively, while contents of SiO2, Al2O3, and CaO decreased obviously. The significant enrichment of Fe in ferrospheres was obvious compared to their corresponding fly ashes, revealing that wet magnetic separation was an effective method to extract Fe-enriched ferrospheres from fly ashes. The ferrospheres accquired can be used as an adsorbent for removing certain pollutants. The morphology changes of fly ashes and their extracted ferrospheres are shown in Figure 2. Most FA1 and FA3 fly ashes exhibit irregular shape and non-uniform size, while FA2 appears to be close to an ideal sphere. This is mainly due to FA2 fly ash being from a pulverized coal boiler under a higher combustion temperature, while FA1 and FA3 are from circulating fluidized bed boilers under a lower combustion temperature. In comparison to the hard surface of the original ashes, all wet magnetic separated ferrospheres of FS1, FS2, and FS3 become loose between particles and rough and defective on their surfaces. This is obviously a result of water erosion. A number of iron microspheres are detected adhering to their surface or scattering around them. In the enlarged SEM images of FS1, FS2, and FS3 shown in Figure 2, the iron microspheres adhering to the surface can be observed more clearly. The XRD patterns of fly ashes and ferrospheres are shown in Figure 3. The main diffraction peaks of bulk fly ashes are assigned to quartz, mullite, and calcium- and iron-containing minerals. Calcium-containing minerals mainly exist in anhydrite, calcite, and lime, which are very prominent in FA1. This suggests that FA1 has a higher content of calcium; it is consistent with that in Table 2. Iron-containing minerals are identified as mainly magnetite and hematite. Contrasting with fly ashes, diffraction peaks of quartz and mullite greatly decrease in the XRD patterns of ferrospheres, while diffraction peaks of magnetite and hematite significantly increase, suggesting that they are the main compounds of ferrospheres.23 It is worth noting that the content of magnetite is much larger than that of hematite in FS2. This is not true for FS1 and FS3, where the differences between the two major iron compounds are quite smaller. This is mostly due to the fact that, in the corresponding fly ash of FA2, the content of magnetite is larger than that of hematite, while it is contrary for FA1 and FA3. 3.2. Adsorption Behavior. To compare arsenic removal characteristics by fly ashes and ferrospheres, test was conducted

573, 723, 873, and 1073 60

To ensure the accuracy of the experimental data, the amount of arsenic condensed in the reactor was determined by carrying out a blank experiment, cleaning the reactor using HNO3, and analyzing the concentration in solution. Arsenic captured in impingers was also analyzed after each experiment to guarantee the error between the arsenic inputs and outputs (including absorbed by the sorbent, condensed in the reactor, and captured in impingers) less than 20%. Several repeated runs were conducted to ensure the reproducibility of each run. The results indicated that the relative deviation of each run was generally within ±7.0%. All results within this deviation are regarded as unchanged. 2.3. Analytical Methods. To assess the adsorption characteristics of fly ashes and ferrospheres, the total amount of captured arsenic was determined by atomic fluorescence spectroscopy (AFS-9800, Beijing Haiguang Instruments Co., Ltd.) after digestion in a microwave oven (Berghof SpeedWave MWS-4, Germany) using the mixed acid of HNO3 and HF. Sorption capacity (milligrams of arsenic per gram of sorbent) and efficiency (percentage of arsenic retained) were then evaluated. The inorganic compositions of fly ashes and ferrospheres were analyzed with XRF spectroscopy (XRF-1800, Shimadzu Co., Japan). The analyses of the crystalline structure of the reaction products were performed by XRD (Rigaku Ultima III, Japan). Scanning electron microscopy (SEM, Carl Zeiss Supra 55, Germany) was applied regarding the detailed information about the microstructure of the particles. The main elements on the surface of ferrospheres before and after sorption were determined by XPS (Escalab 250, Thermo Fisher Scientific, Waltham, MA).

3. RESULTS AND DISCUSSION 3.1. Characteristics of the Adsorbents. The compositions of the inorganic components of fly ashes and ferrospheres are shown in Table 2. All three selected fly ashes (FA1, FA2, and FA3) had higher contents of SiO2 and Al2O3, whereas Fe2O3 displayed low contents of 8.00, 5.01, and 3.00% by mass, respectively. FA1 had a higher content of CaO compared to FA2 and FA3. After wet magnetic separation, the remarkably high total Fe2O3 contents appeared in ferrospheres of FS1, FS2, 8748

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752

Article

Energy & Fuels

Figure 2. SEM images of fly ashes and ferrospheres.

Figure 3. XRD patterns of fly ashes and ferrospheres [(Q, quartz (SiO2); M, mullite (3A12O3·2SiO2); A, anhydrite (CaSO4); C, calcite (CaCO3); L, lime (CaO); Ma, magnetite (Fe3O4); and H, hematite (Fe2O3)].

The following definition is employed to evaluate the removal efficiency of gas-phase arsenic in flue gas by fly ashes and ferrospheres:

under the same temperature of 873 K. As shown in Figure 4, the amounts of arsenic captured by three fly ashes FA1, FA2, and FA3 are 0.13, 0.30, and 0.25 mg/g, respectively, while those captured by the corresponding ferrospheres FS1, FS2, and FS3 are 0.29, 0.53, and 0.31 mg/g, respectively. The arsenic adsorption abilities of FS1 and FS2 are almost twice that of the original ashes, suggesting that ferrospheres are a much more effective adsorbent than their original ashes, and the reason is mainly attributed to the strong adsorption activity for arsenic by Fe compounds existing in them.

η=

mAs,ad mAs,in

× 100%

where mAs,ad is the mass of arsenic captured by the adsorbent and mAs,in is the total mass of arsenic introduced into the reaction system. Table 3 shows the arsenic removal efficiency 8749

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752

Article

Energy & Fuels

physical properties of the ferrosphere particles, such as the particle size, specific surface area, pore structure, and distribution of the Fe compound. FS2 possesses the most excellent arsenic adsorption ability, which is exactly due to the average distributed uniformity, small grain size without aggregation, and thus the uniform distribution of the Fe compound. It is also supposed that the adsorption ability of ferrospheres may be related to the crystal structure of Fe compounds. The magnetite content in FS2 is found much larger, which might possess better chemical activity and affinity to arsenic than hematite. This deserves to be further studied. 3.3. Effect of the Adsorption Temperature. FS2 was taken as an example to investigate the effect of the adsorption temperature on the arsenic removal. Adsorption experiments of FS2 at different temperatures were carried out. As shown in Figure 5, the arsenic capture by FS2 was strongly affected by

Figure 4. Amount of arsenic captured by fly ashes and ferrospheres at 873 K.

Table 3. Capture Efficiency of Arsenic by Fly Ashes and Ferrospheres at 873 K fly ashes

capture efficiency, η (%)

ferrospheres

capture efficiency, η (%)

FA1 FA2 FA3

10.68 24.15 19.57

FS1 FS2 FS3

23.26 42.75 25.03

by fly ashes and ferrospheres. As expected, removal efficiencies of FS1 and FS2 are almost 2 times that of the original ashes. FS2 achieves the highest removal efficiency of 42.75%, which is nearly half of the total amount of arsenic introduced into flue gas. An evaluation of calcium contents in the three original fly ashes reveals the remarkably high content of the total calcium content of 41.19% in FA1; however, the corresponding removal efficiency for arsenic is the lowest in the three fly ashes, which is only 10.68%. Similarly, the calcium content in FS1 is also the highest, while the corresponding removal efficiency is the lowest in all three ferrospheres. This seems to contradict previous reports,3,12,18 bearing in mind that it was reported that arsenic could be captured by limestone, which was suggested through a chemical reaction between CaO and As2O3 resulting in Ca3(AsO4)2. Nevertheless, the results of this study suggest that the calcium composition in fly ash does not greatly influence the adsorption ability for arsenic; arsenic removal efficiency hardly increases with the calcium content in fly ash. This may be ascribed to the various crystal structures of calcium compounds in fly ash. Anhydrite and calcite are the main forms of calcium compounds, while less lime is observed in FA1. Both the reaction between CaSO4 and As2O3 and that between CaCO3 and As2O3 were found much more difficult to occur compared to lime.10,18 For example, Chen et al. found that CaSO4 was able to absorb As2O3(g) only at high temperatures in 1023−1323 K, while the reaction between CaO and As2O3(g) occurred at a lower temperature range.10 Therefore, the arsenic removal by fly ashes in this study is more attributed to the Fe active component in them and less likely attributed to calcium compounds. It can be observed that Fe contents in all three ferrospheres are almost similar; however, their arsenic adsorption abilities and arsenic removal efficiencies are significantly different. One can speculate that the arsenic capture by ferrospheres should be associated with not only the Fe components but also the

Figure 5. Arsenic capture by FS2 at different temperatures.

the adsorption temperatures. Few arsenic was captured at 573 K (0.043 mg/g, 3.44%). The capture of arsenic was remarkably enhanced with the temperature increasing from 573 to 873 K (0.534 mg/g, 42.75%), while it was strongly decreased with the temperature further increasing from 873 to 1073 K (0.208 mg/ g, 16.64%). The mechanism of arsenic capture by ferrospheres was quite complicated. Arsenic vapors might be stabilized in the ferrosphere particles through condensation, physical adsorption, and chemical oxidation.10 To further understand the mechanism of arsenic capture by ferrospheres, the reaction products obtained at temperatures from 573 to 1073 K were reheated to 1073 K under N2 for 30 min to test their thermal stability. The results showed that captured arsenic on FS2 at lower temperatures was weaker and quite thermally unstable. Captured arsenic at 573 and 723 K was released by 37.3 and 13.5% during thermal treatment, respectively, implying that captured arsenic at lower adsorption temperatures was predominantly through direct condensation and/or physical adsorption. This could be explained that As2O3 has a high melting point and boiling point (about 588 and 740 K, respectively) and As2O5 is unstable to heat; it start to release O2 and is converted into As2O3 at its melting point of 573 K.24,25 Thereby, below or around the above temperature range (from 573 to 723 K), As2O3 and As2O5 vapors would come into a saturated or supersaturated state in flue gas and, thus, intend to be condensed on the surface of ferrosphere particles. Captured arsenic at high temperatures of 873 and 1073 K was much more stable than that at low temperatures; less arsenic was released when it was heated, implying a strong 8750

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752

Article

Energy & Fuels

ferrosphere products obtained at all three adsoption temperatures. This may be because the amount of arsenic adsorbed is too small and lower than the detection limits of the XRD technique. Therefore, XPS analysis was further carried out to find traces of iron arsenate in FS2 ferrospheres. Figure 7 displays the XPS spectra of FS2 ferrospheres before and after the adsorption reaction at 873 K. Figure 8 further

chemical association rather than physical adsorption. This is agreeable with principles of chemical adsorption. An activation energy is required for chemical adsorption to occur; thus, only enough energy input can overcome reaction barriers of chemical adsorption (herein, the temperature rising to about or above 723 K). A further increase of the adsorption temperature from 723 to 873 K promoted the chemical adsorption of arsenic vapors; this is because more adsorption sites of adsorbent as well as more gas-phase arsenic molecules are activated and available for the adsorption reaction. As a result, the amount of captured arsenic increased rapidly, and when it reached 873 K, the adsorption amount achieved the maximum. Therefore, to ensure a high efficiency for arsenic capture by ferrospheres, the optimized temperature should be around 873 K. However, the adsorption amount of arsenic decreased as the temperature further rose from 873 to 1073 K. This is agreeable with the fact that the chemical adsorption of As2O3(g) and As2O5 (g) onto ferrospheres is an exothermic reaction in nature. Thermodynamic calculation confirms it. Various studies have suggested that reactions of Fe compounds with As are as follows (eqs 1 and 2):18,20 hematite with As(g) Fe2O3 + As2 O3(g) + O2 (g) → 2Fe(AsO4 )

Figure 7. XPS spectra of FS2 ferrospheres before and after the adsorption reaction.

(1)

magnetite with As(g) illustrates the fine XPS spectra of 2p orbits of both Fe and Ca. The XPS peaks of Si, Al, Fe, Ca, and O elements were prominent, suggesting that the ferrospheres were mainly composed of the above five elements. In comparison to FS2 ferrospheres before the adsorption reaction, a new peak at 44− 45 eV was observed in the ferrospheres after the adsorption at 873 K, which was attributed to the typical peak of As 3d. It is comfirmed that the arsenic vapor was captured by the ferrospheres through the surface adsorption. Moreover, the Fe 2p3/2 peak showed a small shift (about 0.83 eV) toward a high binding energy state (from 710.86 to 711.69 eV), as shown in Figure 8a, implying that the electrons were transferred from Fe atoms to arsenates influenced by the strong electronegativity of arsenates, which led to the relaxation effect of Fe 2p to a higher binding energy position. From these results, it is obvious that the arsenic vapor could be captured by the Fe active component in ferrospheres and iron arsenates were formed through the chemical reactions between iron oxides and arsenic oxide (g). Meanwhile, XPS peaks of Si 2p, Al 2p, and Ca 2p appeared with almost no shift before and after the adsoption, suggesting that Si, Al, and Ca compounds in ferrospheres nearly did not participate in the adsorption reaction for arsenic. Therefore, it is finally confirmed that the major contributor to arsenic capture by ferrospheres is the Fe active component in them and not Si, Al, and Ca compounds.

3Fe3O4 + 4As2 O3(g) + 4O2 (g) → 6Fe(AsO4 ) + Fe3(AsO4 )2

(2)

We used HSC Chemistry 6.0 to evaluate the reaction enthalpy of eqs 1 and 2; the enthalpy change ΔH obtained was below zero, suggesting that the above adsorption reactions were exothermic reactions. When the temperature is very high, the chemical adsorption can quickly tend to balance, yet the equilibrium adsorption capacity decreases with the temperature increasing; therefore, an excessive increase of the temperature is not conducive to arsenic adsorption. 3.4. Mechanism of Arsenic Capture by Ferrospheres. Figure 6 shows the XRD patterns of FS2 ferrospheres after the adsorption reactions at 723, 873, and 1073 K. Unfortunately, the formation of iron arsenate was not detected in any of the

4. CONCLUSION This study reveals that the wet magnetic separation is an effective way to extract ferrospheres with a higher Fe content from fly ashes. The ferrospheres can achieve a much higher removal efficiency of gas-phase arsenic from flue gas than the raw ashes. The arsenic capture by ferrospheres was enhanced initially but decreased then with the temperature increasing, showing that the condensation/physical adsorption predominated in temperatures ranging between 573 and 723 K, while

Figure 6. XRD patterns of FS2 ferrospheres before and after the adsorption reaction [Q, quartz (SiO2); M, mullite (3A12O3·2SiO2); Ma, magnetite (Fe3O4); and H, hematite (Fe2O3)]. 8751

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752

Article

Energy & Fuels

Figure 8. XPS spectra of Fe 2p and Ca 2p of FS2 ferrospheres before and after the adsorption reaction: (1) before adsorption and (2) after adsorption, at 873 K. (11) Diaz-Somoano, M.; Martinez-Tarazona, M. R. Environ. Sci. Technol. 2004, 38 (3), 899−903. (12) Jadhav, R.; Fan, L. Environ. Sci. Technol. 2001, 35 (4), 794−799. (13) Zhang, Y.; Wang, C.; Li, W.; Liu, H.; Zhang, Y.; Hack, P.; Pan, W. Energy Fuels 2015, 29 (10), 6578−6585. (14) Chen, W. F.; Parette, R.; Zou, J.; Cannon, F. S.; Dempsey, B. A. Water Res. 2007, 41 (9), 1851−1858. (15) López-Antón, M.; Díaz-Somoano, M.; Fierro, J.; MartínezTarazona, M. R. Fuel Process. Technol. 2007, 88 (8), 799−805. (16) Sterling, R. O.; Helble, J. J. Chemosphere 2003, 51 (10), 1111− 1119. (17) Xu, W.; Wang, H.; Zhu, T.; Kuang, J.; Jing, P. J. Environ. Sci. 2013, 25 (2), 393−398. (18) Lopez-anton, M.; Diaz-somoano, M.; Spears, D.; Martineztarazona, M. R. Environ. Sci. Technol. 2006, 40 (12), 3947−3951. (19) Zhang, K.; Zhang, D.; Zhang, K. Water Sci. Technol. 2016, 73 (8), 1954−1962. (20) Huggins, F. E.; Senior, C. L.; Chu, P.; Ladwig, K.; Huffman, G. P. Environ. Sci. Technol. 2007, 41, 3284−3289. (21) Shah, P.; Strezov, V.; Prince, K.; Nelson, P. F. Fuel 2008, 87, 1859−1869. (22) Xue, Q. F.; Lu, S. G. J. Zhejiang Univ., Sci., A 2008, 9 (11), 1595−1600. (23) Zhao, Y. C.; Zhang, J. Y.; Sun, J. M.; Bai, X. F.; Zheng, C. G. Energy Fuels 2006, 20 (4), 1490−1497. (24) Wang, J.; Tomita, A. Energy Fuels 2003, 17, 954−960. (25) Sakulpitakphon, T.; Hower, J. C.; Trimble, A. S.; Schram, W. H.; Thomas, G. A. Energy Fuels 2003, 17, 1028−1033.

the chemical adsorption predominated at further higher temperatures. The optimal temperature to maintain a high arsenic capture efficiency using ferrospheres should be around 873 K. The major contributor to arsenic capture by ferrospheres is the Fe active component in them rather than Si, Al, and Ca compounds. Therefore, separating ferrospheres from fly ashes and further injecting them back into the flue not only can achieve a much higher removal efficiency for gas-phase arsenic but also provides an effective and economic approach on the recovery of the useful product directly from fly ashes.



AUTHOR INFORMATION

Corresponding Author

*Telephone: 86-13520707519. E-mail: [email protected] and/ or [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors are grateful for financial support of this research from the National Natural Science Foundation of China (51306052). The authors also acknowledge the Institute for Energy Environment Science and Engineering of North China Electric Power University for supplying the microwave digestion system.



REFERENCES

(1) Hughes, M. F. Toxicol. Lett. 2002, 133 (1), 1−16. (2) Sharma, V. K.; Sohn, M. Environ. Int. 2009, 35 (4), 743−759. (3) Mahuli, S.; Agnihotri, R.; Chauk, S.; Ghosh-Dastidar, A.; Fan, L. Environ. Sci. Technol. 1997, 31 (11), 3226−3231. (4) Winter, R.; Mallepalli, R.; Hellem, K.; Szydlo, S. Combust. Sci. Technol. 1994, 101 (3), 45−48. (5) Lundholm, K.; Boström, D.; Nordin, A.; Shchukarev, A. Environ. Sci. Technol. 2007, 41, 6534−6540. (6) Contreras, M. L.; Arostegui, J. M.; Armesto, L. Fuel 2009, 88, 539−546. (7) Tang, Q.; Liu, G.; Yan, Z.; Sun, R. Fuel 2012, 95, 334−339. (8) Tian, H.; Wang, Y.; Xue, Z.; Qu, Y.; Chai, F.; Hao, J. Sci. Total Environ. 2011, 409, 3078−3081. (9) Zhang, K.; Zhang, K.; Pan, W. P. J. Fuel Chem. Technol. 2013, 41, 839−844 (in Chinese). (10) Chen, D.; Hu, H.; Xu, Z.; Liu, H.; Cao, J.; Shen, J.; Yao, H. Chem. Eng. J. 2015, 267, 201−206. 8752

DOI: 10.1021/acs.energyfuels.6b01637 Energy Fuels 2016, 30, 8746−8752