The Synthesis of FeCl 3 -Modiﬁed Char from Phoenix Tree Fruit and Its Application for Hg 0 Adsorption in Flue Gas

: A sample of FeCl 3 -modiﬁed phoenix tree fruit char (MPTFC) was prepared using pyrolysis and a facile chemical immersion method; it was proposed as an effective sorbent for Hg 0 adsorption in ﬂue gas. The BET, SEM, FTIR, and XPS methods were adopted for the characterizations of the sorbents, and a series of Hg 0 adsorption tests were conducted on a bench-scale Hg 0 removal setup in the lab. The morphological analysis of the sorbent indicated that the hollow ﬁber in phoenix tree fruit (PTF) shifted to organized directional porous tubular columns in phoenix tree fruit char (PTFC) after pyrolysis. The surface area of MPTFC increased slightly in comparison with PTF and PTFC. The MPTFC showed excellent performance for Hg 0 adsorption at 200 ◦ C in ﬂue gas ambiance, and the Hg 0 removal efﬁciency approached 95% with 5% (wt.%) FeCl 3 modiﬁcation. The presence of O 2 may help to activate the MPTFC for Hg 0 adsorption in ﬂue gas, thus greatly promoting Hg 0 adsorption capability. NO had a positive effect on Hg 0 adsorption, while the presence of SO 2 in ﬂue gas restrained Hg 0 adsorption by MPTFC. Functional groups, such as C-Cl and Fe-O, were successfully decorated on the surface of PTFC by FeCl 3 modiﬁcation, which contributed greatly to Hg 0 adsorption. In addition, C=O, lattice oxygen (O α ), and adsorbed oxygen (O β ) also contributed to Hg 0 adsorption and oxidization. structure. Those alterations in morphology are highly suspected to result from the release of volatiles during pyrolysis. The violent gas explosion from the inside of the PTF caused the expansion of the layer structures, thus producing the organized tubular columns. Those changes beneﬁted PTFC in terms of adsorption because of the relatively higher number of porous structures. The micro-morphology of MPTFC is almost the same as that of PTFC, which indicates that the FeCl 3 modiﬁcations only functionalized the surface of the material.


Introduction
Mercury (Hg) in the atmosphere is characterized by hypertoxicity, bioaccumulation, and persistence; it is detrimental to both the environment and public health worldwide [1][2][3]. Most of the Hg in the ecosystem originates from human activities, and it is important to control anthropogenic Hg emissions. Coal-fired power plants are considered to be one of the major emission sources of Hg in the atmosphere [4]. To reduce Hg emissions from coal-fired electricity generation, the strictest emission limitations (GB13223-2011) in the world have been announced and put into practice by the Chinese Environmental Ministry. Besides the standards set for conventional sources of pollution, such as PMs, SO 2 , and NO, clear emission limitations have also been issued, for the first time, for pollution caused by non-conventional heavy metals such as Hg. Recently, a stricter standard with a maximum Hg emission concentration of 1 µg/m 3 was proposed for abating Hg emissions from coalfired power plants [5,6]. Herein, we state that controlling Hg emissions from coal-fired electricity generation is an urgent and significant issue.
Although Hg is present in coal in trace amounts, the total emissions are nevertheless very significant because of the heavy and long-term use of coal as the major source of energy in China [7][8][9][10]. Hg first evaporates as Hg 0 vapor during coal combustion; then, a series of chemical reactions occur over a wide range of temperature drops in the flue gas. Hg releases

Char Preparation from Phoenix Tree Fruit and FeCl 3 Modification
First, the char (PTFC) was prepared by pyrolysis from phoenix tree fruit (PTF) in an N 2 atmosphere at a flow rate of 300 mL/min for 30 min at 600 • C. After that, the phoenix tree fruit char (PTFC) was then modified with FeCl 3 by a facile chemical immersion method in solutions with different concentrations of FeCl 3 . Solid FeCl 3 was first diluted in DI water to produce FeCl 3 -containing solutions with weight percentages of 1%, 3%, 5%, and 7%. Then, 1 g PTFC was mixed with 3 mL of each of the FeCl 3 solutions of different concentrations in a beaker with magnetic stirring for 60 min. Then, the mixture system was ultrasound shocked for 10 min. After that, the mixtures were filtered and transferred to an oven to dry at 105 • C for 12 h. The modified PTFC (MPTFC) was obtained.

Analytical Methods
Several characterization methods, namely BET, SEM, EDS, FTIR, and XPS, were adopted for the characterization of MPTFC. The specific area (BET) analysis was adopted for measuring the surface area of the sorbents, and the BET analysis was conducted using the Micrometrics TriStar II 3020 from the United States of America. The SEM analysis method was used for micro-morphology observations, and the tests were conducted on the QUANTA200 from the FEI company at Wuhan University. In addition, the energy dispersive spectrometer (EDS) method was used to identify the elemental distributions in the sorbents in order to confirm the success of the Fe-Cl impregnations. The EDS analysis was conducted using the Horiba 7021-H. Further, the Nicolet 6700 Fourier transform infrared (FTIR) spectrometer was adopted to identify the molecular surface structure and the functional groups on the surface of the sorbents. The FTIR measurements were conducted at Wuhan University. In addition, X-ray photoelectron spectroscopy (XPS) was used for elemental valence state characterizations on the surface of the sorbents. The XPS equipment we used in the experiments was the G2f20 from TECNAI; the reference calibration junction energy was 284.6 eV, and the error was ±0.3 eV.

Hg Adsorption Experiments
The Hg 0 removal setup is shown in Figure 1. The experimental system can be divided into five major parts, namely the gas feeding and mixture system, the Hg 0 generation system, the main reaction system, the Hg 0 monitoring system, and the exhaust purification system.

Analytical Methods
Several characterization methods, namely BET, SEM, EDS, FTIR, and XPS, were adopted for the characterization of MPTFC. The specific area (BET) analysis was adopted for measuring the surface area of the sorbents, and the BET analysis was conducted using the Micrometrics TriStar II 3020 from the United States of America. The SEM analysis method was used for micro-morphology observations, and the tests were conducted on the QUANTA200 from the FEI company at Wuhan University. In addition, the energy dispersive spectrometer (EDS) method was used to identify the elemental distributions in the sorbents in order to confirm the success of the Fe-Cl impregnations. The EDS analysis was conducted using the Horiba 7021-H. Further, the Nicolet 6700 Fourier transform infrared (FTIR) spectrometer was adopted to identify the molecular surface structure and the functional groups on the surface of the sorbents. The FTIR measurements were conducted at Wuhan University. In addition, X-ray photoelectron spectroscopy (XPS) was used for elemental valence state characterizations on the surface of the sorbents. The XPS equipment we used in the experiments was the G2f20 from TECNAI; the reference calibration junction energy was 284.6 eV, and the error was ±0.3 eV.

Hg Adsorption Experiments
The Hg 0 removal setup is shown in Figure 1. The experimental system can be divided into five major parts, namely the gas feeding and mixture system, the Hg 0 generation system, the main reaction system, the Hg 0 monitoring system, and the exhaust purification system.

Figure 1.
Bench-scale experimental system for Hg 0 adsorption. The gas feeding system chiefly consisted of sources of O2, N2, NO, and SO2 with a mass flow controller for control of the flow rate. The simulated flue gas was mostly composed of the 4 gases listed above, with proportions chosen to simulate the components of flue gas. The Hg 0 generation system used a Dynacalibrator Model 500 mercury penetration tube. The steady Hg 0 generation system also consisted of a water bath, N2 as a carrier gas with a flow rate of 300 mL/min, a U-shaped tube for Hg 0 penetration, and a tube holder. The initial Hg 0 concentration was maintained at 50μg/m 3 by controlling the temperature of the water bath at 60 °C. The main part of the adsorption reaction system was a vertical fixed bed, which mainly consisted of an electric furnace and an annular quartz tube reactor. The diameter of the inner tube was 14 mm with a quartz sieve plate. During Figure 1. Bench-scale experimental system for Hg 0 adsorption. The gas feeding system chiefly consisted of sources of O 2 , N 2 , NO, and SO 2 with a mass flow controller for control of the flow rate. The simulated flue gas was mostly composed of the 4 gases listed above, with proportions chosen to simulate the components of flue gas. The Hg 0 generation system used a Dynacalibrator Model 500 mercury penetration tube. The steady Hg 0 generation system also consisted of a water bath, N 2 as a carrier gas with a flow rate of 300 mL/min, a U-shaped tube for Hg 0 penetration, and a tube holder. The initial Hg 0 concentration was maintained at 50µg/m 3 by controlling the temperature of the water bath at 60 • C. The main part of the adsorption reaction system was a vertical fixed bed, which mainly consisted of an electric furnace and an annular quartz tube reactor. The diameter of the inner tube was 14 mm with a quartz sieve plate. During the reactions, the sorbents were placed in the plate covered by quartz wool to inhibit the blowing of the sorbent during gas feeding. During the experiment, the Hg 0 vapor passed through the sorbents, and the Hg 0 adsorption reactions occurred in the gas-solid reaction system. A VM3000 from Mercury Instruments in Germany was used for Hg 0 online monitoring, and the detection precision was about 0.1 µg/m 3 . The exhaust first passed through activated carbon to dispose of the remaining Hg 0 before it was sucked into a vent.
A series of experiments were conducted, and the detailed experimental arrangements are as listed in Table 1. The Hg 0 removal efficiency η is calculated by using the following Equation (1): where η is the Hg 0 removal efficiency; C t and C 0 stand for the Hg 0 concentration at t seconds and the initial Hg 0 concentration, respectively.

Properties of the Phoenix Tree Fruit
The proximate and ultimate analyses of PTF are shown in Table 2. It can be observed that the major components are volatiles (72.08%) and fixed carbon (15.93%). However, the moisture and ash contents are relatively low, only accounting for 5.41% and 6.58%, respectively. The high content of volatiles in the materials was the basis for concluding that a flourishing porous structure would result when the char was prepared by pyrolysis; this would enhance adsorption capability. The major elements in PTF are C and O, at around 66% and 35%, respectively. Meanwhile, the contents of H, N, and S are relatively low. The abundance of C and O in PTF creates more chances to form oxygen-containing functional groups during pyrolysis. Those oxygen-containing groups are believed to benefit the promotion of Hg 0 adsorption capability. Meanwhile, the lower contents of N and S reduce the possibility of N-and S-bearing gas emissions during pyrolysis, thus reducing the potential environmental pollution caused by pyrolysis.

SEM Observations
The SEM observations for PTF, PTFC, and MPTFC are shown in Figure 2. It can be seen that the original PTF has a hollow fiber shape with compact layer structures, and the diameter of the fiber is about 20 µm. Both the surface and the inner layer of the fiber seem to be smooth and without pores. However, the micro-morphology of PTFC is very different in comparison to that of PTF. The hollow fiber shape shifts to organized directional tubular columns. A single tubular structure is about 10 µm in diameter, while the column is about 50 µm in diameter. Some cracks and pores can also be observed on the surface of the tubular structure. Those alterations in morphology are highly suspected to result from the release of volatiles during pyrolysis. The violent gas explosion from the inside of the PTF caused the expansion of the layer structures, thus producing the organized tubular columns. Those changes benefited PTFC in terms of adsorption because of the relatively higher number of porous structures. The micro-morphology of MPTFC is almost the same as that of PTFC, which indicates that the FeCl 3 modifications only functionalized the surface of the material.
diameter of the fiber is about 20 μm. Both the surface and the inner layer of the fiber seem to be smooth and without pores. However, the micro-morphology of PTFC is very different in comparison to that of PTF. The hollow fiber shape shifts to organized directional tubular columns. A single tubular structure is about 10 μm in diameter, while the column is about 50 μm in diameter. Some cracks and pores can also be observed on the surface of the tubular structure. Those alterations in morphology are highly suspected to result from the release of volatiles during pyrolysis. The violent gas explosion from the inside of the PTF caused the expansion of the layer structures, thus producing the organized tubular columns. Those changes benefited PTFC in terms of adsorption because of the relatively higher number of porous structures. The micro-morphology of MPTFC is almost the same as that of PTFC, which indicates that the FeCl3 modifications only functionalized the surface of the material.

BET Analysis of the Sorbents
The BET(BET surface area,SBET,m 2 /g)analysis results are shown in Table 3. The results reveal that the surface area of PTF is around 2.4 m 2 /g, and the surface area of PTFC is around 3.4 m 2 /g. The increase in the surface area from PTF to PTFC is highly suspected to be linked to the volatilization of the volatile matter, including CO2 and H2O. The release of those gas components produces porous structures, thus increasing the surface area. However, the surface area of MPTFC is larger in comparison with PTF and PTFC. The concentrations of the precursor also have some impact on the surface area. When the concentration of FeCl3 increased from 3% (wt.%) to 5% (wt.%), the surface area increased from 5.8 m 2 /g to 10.6 m 2 /g. However, when the concentration increased to 7% (wt.%), the surface area decreased slightly to 8.4 m 2 /g; this is highly suspected to be linked to the blocking of the pores and tubular channels by the higher concentration of the Fe-Cl precursor. To evaluate the effects of temperature on the performance efficiency of PTFC for Hg 0 removal, we conducted the Hg 0 adsorption experiments at different temperatures in the N2 ambiance. The Hg 0 removal efficiencies as a function of temperature are shown in Figure 3. It can be seen that the optimal temperature for Hg 0 adsorption by PTFC was 150 °C; the efficiency was greater than 50%. As the temperature increased, the Hg 0 removal efficiency of PTFC decreased, falling to 16% when the temperature increased to 300 °C. The results indicate that physical adsorption may dominate the Hg 0 adsorption of PTFC in N2

BET Analysis of the Sorbents
The BET(BET surface area, SBET, m 2 /g) analysis results are shown in Table 3. The results reveal that the surface area of PTF is around 2.4 m 2 /g, and the surface area of PTFC is around 3.4 m 2 /g. The increase in the surface area from PTF to PTFC is highly suspected to be linked to the volatilization of the volatile matter, including CO 2 and H 2 O. The release of those gas components produces porous structures, thus increasing the surface area. However, the surface area of MPTFC is larger in comparison with PTF and PTFC. The concentrations of the precursor also have some impact on the surface area. When the concentration of FeCl 3 increased from 3% (wt.%) to 5% (wt.%), the surface area increased from 5.8 m 2 /g to 10.6 m 2 /g. However, when the concentration increased to 7% (wt.%), the surface area decreased slightly to 8.4 m 2 /g; this is highly suspected to be linked to the blocking of the pores and tubular channels by the higher concentration of the Fe-Cl precursor. To evaluate the effects of temperature on the performance efficiency of PTFC for Hg 0 removal, we conducted the Hg 0 adsorption experiments at different temperatures in the N 2 ambiance. The Hg 0 removal efficiencies as a function of temperature are shown in Figure 3. It can be seen that the optimal temperature for Hg 0 adsorption by PTFC was 150 • C; the efficiency was greater than 50%. As the temperature increased, the Hg 0 removal efficiency of PTFC decreased, falling to 16% when the temperature increased to 300 • C. The results indicate that physical adsorption may dominate the Hg 0 adsorption of PTFC in N 2 +8% O 2 ambiance. Hg 0 was physically adsorbed on the surface of the PTFC and then diffused in the tubular columns in the gas flow fields. The binding force between the Hg 0 and the PTFC was weak, thus, it was desorbed as temperature increased as a result of the intensification of molecular motion. +8% O2 ambiance. Hg 0 was physically adsorbed on the surface of the PTFC and then diffused in the tubular columns in the gas flow fields. The binding force between the Hg 0 and the PTFC was weak, thus, it was desorbed as temperature increased as a result of the intensification of molecular motion.  Figure 4. It can be seen that FeCl3 modification of PTFC can greatly promote Hg 0 adsorption capability in comparison with PTFC. However, temperature also has an impact on MPTFC's efficacy for Hg 0 removal. Its Hg 0 removal efficiency increased from 73% to 95% when the temperature increased from 150 °C to 200 °C, respectively. However, the Hg 0 removal efficiency decreased to 83% and 74% when the temperature continued increasing to 250 °C and 300 °C, respectively. It can be concluded that the optimal temperature for Hg 0 adsorption by MPTFC is 200 °C, which is slightly different from that of PTFC. It can be inferred that the Hg 0 adsorption mechanisms are different for PTFC and MPTFC. Regarding PTFC, Hg 0 physical adsorption on the surface of the sorbent is more likely to be the dominating reaction. However, high temperature intensifies molecular thermal motion, thus weakening the surface force of Hg 0 adsorption by PTFC [30,31]. However, when the Fe-Cl functional groups are introduced on the surface of the sorbent, both physical and chemical adsorption reactions occur. The increase in temperature in a certain range promotes chemisorption reactions, thus enhancing the Hg 0 adsorption capability.  Figure 4. It can be seen that FeCl 3 modification of PTFC can greatly promote Hg 0 adsorption capability in comparison with PTFC. However, temperature also has an impact on MPTFC's efficacy for Hg 0 removal. Its Hg 0 removal efficiency increased from 73% to 95% when the temperature increased from 150 • C to 200 • C, respectively. However, the Hg 0 removal efficiency decreased to 83% and 74% when the temperature continued increasing to 250 • C and 300 • C, respectively. It can be concluded that the optimal temperature for Hg 0 adsorption by MPTFC is 200 • C, which is slightly different from that of PTFC. It can be inferred that the Hg 0 adsorption mechanisms are different for PTFC and MPTFC. Regarding PTFC, Hg 0 physical adsorption on the surface of the sorbent is more likely to be the dominating reaction. However, high temperature intensifies molecular thermal motion, thus weakening the surface force of Hg 0 adsorption by PTFC [30,31]. However, when the Fe-Cl functional groups are introduced on the surface of the sorbent, both physical and chemical adsorption reactions occur. The increase in temperature in a certain range promotes chemisorption reactions, thus enhancing the Hg 0 adsorption capability.
Atmosphere 2022, 13, x FOR PEER REVIEW 6 of 14 +8% O2 ambiance. Hg 0 was physically adsorbed on the surface of the PTFC and then diffused in the tubular columns in the gas flow fields. The binding force between the Hg 0 and the PTFC was weak, thus, it was desorbed as temperature increased as a result of the intensification of molecular motion.

Effects of Temperature on Hg 0 Removal by MPTFC
The Hg 0 adsorption capabilities of MPTFC as a function of temperature are shown in Figure 4. It can be seen that FeCl3 modification of PTFC can greatly promote Hg 0 adsorption capability in comparison with PTFC. However, temperature also has an impact on MPTFC's efficacy for Hg 0 removal. Its Hg 0 removal efficiency increased from 73% to 95% when the temperature increased from 150 °C to 200 °C, respectively. However, the Hg 0 removal efficiency decreased to 83% and 74% when the temperature continued increasing to 250 °C and 300 °C, respectively. It can be concluded that the optimal temperature for Hg 0 adsorption by MPTFC is 200 °C, which is slightly different from that of PTFC. It can be inferred that the Hg 0 adsorption mechanisms are different for PTFC and MPTFC. Regarding PTFC, Hg 0 physical adsorption on the surface of the sorbent is more likely to be the dominating reaction. However, high temperature intensifies molecular thermal motion, thus weakening the surface force of Hg 0 adsorption by PTFC [30,31]. However, when the Fe-Cl functional groups are introduced on the surface of the sorbent, both physical and chemical adsorption reactions occur. The increase in temperature in a certain range promotes chemisorption reactions, thus enhancing the Hg 0 adsorption capability.

Effects of FeCl 3 Concentrations on MPTFC for Hg 0 Removal
The FeCl 3 precursor we used for the PTFC modifications was prepared in a concentration gradient from 1% (wt.%) to 7% (wt.%) with intervals of 2% (wt.%). The series of MPTFCs with different levels of FeCl 3 modifications was tested for Hg 0 adsorption capability at 200 • C in N 2 + 8%O 2 ambiance. The Hg 0 removal efficiencies for all MPTFCs are shown in Figure 5.

Effects of FeCl3 Concentrations on MPTFC for Hg 0 Removal
The FeCl3 precursor we used for the PTFC modifications was prepared in a concentration gradient from 1% (wt.%) to 7% (wt.%) with intervals of 2% (wt.%). The series of MPTFCs with different levels of FeCl3 modifications was tested for Hg 0 adsorption capability at 200 °C in N2 + 8%O2 ambiance. The Hg 0 removal efficiencies for all MPTFCs are shown in Figure 5. The introduction of FeCl3 to PTFC can help promote Hg 0 adsorption capability. The Hg 0 removal capabilities of MPTFC showed an increasing trend with increasing concentrations of FeCl3 from 1% (wt.%) to 5% (wt.%). The corresponding Hg 0 removal efficiencies increased from 48% to 95%. However, Hg 0 removal efficiency decreased to 84% when the FeCl3 concentration continued increasing to 7% (wt.%). This result indicates that the most optimal concentration of FeCl3 as a precursor for PTFC modification was 5%(wt.%). It can be seen that the presence of Fe 3+ and Clhad positive consequences for Hg 0 oxidization, thus enhancing the chemisorption reactions for Hg 0 on the surface. Fe 3+ has a strong oxidization capability in reactions, while Clis easy to complex with Hg-containing compounds, thus causing Hg 0 oxidization in the reaction. In addition, the modification of PTFC with FeCl3 also enlarged the surface area of the sorbents, which contributed to Hg 0 removal. The BET analysis also indicated that the surface area of MPTFC with the 5% FeCl3 pretreatment was the highest in comparison with the other MPTFCs.

Effects of O2 on Hg 0 Removal by the Sorbents
Here, we discuss the effects of the presence of O2 on the adsorption capabilities of MPTFC. The MPTFC modified with 5% FeCl3 was tested for its Hg 0 adsorption capability in N2 ambiance at different temperatures. The results are shown in Figure 6. It can be observed that MPTFC did not perform as well as in the ambiance containing O2. The Hg 0 removal efficiency decreased significantly in comparison with that observed in the 8% O2 + N2 ambiance. This result indicates that O2 played a key role in Hg 0 oxidization in the reactions. The highest Hg 0 removal efficiency was observed at 150 °C, with an efficiency of less than 35%. The Hg 0 removal efficiency was only 17% at 200 °C. It can also be inferred that physical adsorption probably dominated the Hg 0 removal reactions in the absence of O2. Though Fe 3+ and Clhave very strong oxidization abilities, their reactions with Hg 0 can The introduction of FeCl 3 to PTFC can help promote Hg 0 adsorption capability. The Hg 0 removal capabilities of MPTFC showed an increasing trend with increasing concentrations of FeCl 3 from 1% (wt.%) to 5% (wt.%). The corresponding Hg 0 removal efficiencies increased from 48% to 95%. However, Hg 0 removal efficiency decreased to 84% when the FeCl 3 concentration continued increasing to 7% (wt.%). This result indicates that the most optimal concentration of FeCl 3 as a precursor for PTFC modification was 5% (wt.%). It can be seen that the presence of Fe 3+ and Cl − had positive consequences for Hg 0 oxidization, thus enhancing the chemisorption reactions for Hg 0 on the surface. Fe 3+ has a strong oxidization capability in reactions, while Cl − is easy to complex with Hg-containing compounds, thus causing Hg 0 oxidization in the reaction. In addition, the modification of PTFC with FeCl 3 also enlarged the surface area of the sorbents, which contributed to Hg 0 removal. The BET analysis also indicated that the surface area of MPTFC with the 5% FeCl 3 pretreatment was the highest in comparison with the other MPTFCs.

Effects of O 2 on Hg 0 Removal by the Sorbents
Here, we discuss the effects of the presence of O 2 on the adsorption capabilities of MPTFC. The MPTFC modified with 5% FeCl 3 was tested for its Hg 0 adsorption capability in N 2 ambiance at different temperatures. The results are shown in Figure 6. It can be observed that MPTFC did not perform as well as in the ambiance containing O 2 . The Hg 0 removal efficiency decreased significantly in comparison with that observed in the 8% O 2 + N 2 ambiance. This result indicates that O 2 played a key role in Hg 0 oxidization in the reactions. The highest Hg 0 removal efficiency was observed at 150 • C, with an efficiency of less than 35%. The Hg 0 removal efficiency was only 17% at 200 • C. It can also be inferred that physical adsorption probably dominated the Hg 0 removal reactions in the absence of O 2 . Though Fe 3+ and Cl − have very strong oxidization abilities, their reactions with Hg 0 can only be activated and accelerated with the presence of O 2 . The presence of O 2 in the feeding gas serves as a supplement for adsorbed oxygen and lattice oxygen in reactions, which is of significance for Hg 0 oxidization. only be activated and accelerated with the presence of O2. The presence of O2 in the feeding gas serves as a supplement for adsorbed oxygen and lattice oxygen in reactions, which is of significance for Hg 0 oxidization. Figure 6. Hg 0 adsorption capabilities of MPTFC in N2 ambiance.

Effects of Flue Gas Composition on the Hg 0 Removal Performance of MPTFC
In order to evaluate the impact of the flue gas components on Hg 0 adsorption of MPTFC, certain concentrations of NO and SO2 were introduced into the reaction system in the experiments; the results are shown in Figure 7. It can be observed that the gas components had some impact on Hg 0 removal capability. Specifically, Hg 0 removal efficiency in N2 was around 17%, whereas Hg 0 removal efficiency was boosted to around 95% when O2 was introduced into the reaction system. When 200 ppm NO was added into the reaction system, Hg 0 removal efficiency increased to 98%. However, Hg 0 removal efficiency decreased to 71% when 500 ppm SO2 was added into the reaction system, indicating the inhibiting effects of the SO2 injection. When both NO and SO2 were present in the reaction system, an improvement in Hg 0 adsorption capability was observed, and the Hg 0 removal efficiency increased to 87% in comparison with the SO2-only ambiance. It can also be observed that an increase of NO in the flue gas from 200 ppm to 500 ppm inhibited Hg 0 adsorption capability slightly, which was probably due to competitive adsorption between NO and Hg 0 on the surface of the MPTFC. When NO was at a low concentration, the presence of NO promoted Hg 0 adsorption capability. This result was probably because the NO adsorbed on the surface formed NO2 and NO, and those activated species reacted with Hg 0 to form HgO and Hg (NO3)2.  Figure 7. It can be observed that the gas components had some impact on Hg 0 removal capability. Specifically, Hg 0 removal efficiency in N 2 was around 17%, whereas Hg 0 removal efficiency was boosted to around 95% when O 2 was introduced into the reaction system. When 200 ppm NO was added into the reaction system, Hg 0 removal efficiency increased to 98%. However, Hg 0 removal efficiency decreased to 71% when 500 ppm SO 2 was added into the reaction system, indicating the inhibiting effects of the SO 2 injection. When both NO and SO 2 were present in the reaction system, an improvement in Hg 0 adsorption capability was observed, and the Hg 0 removal efficiency increased to 87% in comparison with the SO 2 -only ambiance. It can also be observed that an increase of NO in the flue gas from 200 ppm to 500 ppm inhibited Hg 0 adsorption capability slightly, which was probably due to competitive adsorption between NO and Hg 0 on the surface of the MPTFC. When NO was at a low concentration, the presence of NO promoted Hg 0 adsorption capability. This result was probably because the NO adsorbed on the surface formed NO 2 and NO, and those activated species reacted with Hg 0 to form HgO and Hg (NO 3 ) 2 . Figure 8 shows the FTIR spectrums of PTFC, MPTFC, and spent MPTFC(S). It is evident that, after MPTFC was modified using FeCl 3 as a precursor, an obvious in-plane bending vibration of the C-H bond can be found at 1396 cm −1 in comparison with PTFC. In addition, the C-Cl bonding of MPTFC can also be found in the wavenumber of 686 cm −1 , indicating that Cl − was successfully modified on the surface of PTFC. There is also a Fe-O characteristic peak showing in the wavenumber of 592 cm −1 for MPTFC(S), which is not identified for PTFC. This result indicates that the chemical reactions between FeCl 3 and PTFC occurred in the sample preparation. The presence of C-Cl and Fe-O in the sample indicates that FeCl 3 transformed into Cl-containing and Fe-containing active functional groups. Both of them contribute to Hg 0 adsorption. Further, it can also be observed that some of the carbon-containing functional groups disappeared after the Hg 0 adsorption reactions, indicating possible reactions between Hg 0 and those carbon-containing functional groups.  Figure 8 shows the FTIR spectrums of PTFC, MPTFC, and spent MPTFC(S). It is evident that, after MPTFC was modified using FeCl3 as a precursor, an obvious in-plane bending vibration of the C-H bond can be found at 1396 cm −1 in comparison with PTFC. In addition, the C-Cl bonding of MPTFC can also be found in the wavenumber of 686 cm −1 , indicating that Clwas successfully modified on the surface of PTFC. There is also a Fe-O characteristic peak showing in the wavenumber of 592 cm −1 for MPTFC(S), which is not identified for PTFC. This result indicates that the chemical reactions between FeCl3 and PTFC occurred in the sample preparation. The presence of C-Cl and Fe-O in the sample indicates that FeCl3 transformed into Cl-containing and Fe-containing active functional groups. Both of them contribute to Hg 0 adsorption. Further, it can also be observed that some of the carbon-containing functional groups disappeared after the Hg 0 adsorption reactions, indicating possible reactions between Hg 0 and those carbon-containing functional groups.    Figure 8 shows the FTIR spectrums of PTFC, MPTFC, and spent MPTFC(S). It is evident that, after MPTFC was modified using FeCl3 as a precursor, an obvious in-plane bending vibration of the C-H bond can be found at 1396 cm −1 in comparison with PTFC. In addition, the C-Cl bonding of MPTFC can also be found in the wavenumber of 686 cm −1 , indicating that Clwas successfully modified on the surface of PTFC. There is also a Fe-O characteristic peak showing in the wavenumber of 592 cm −1 for MPTFC(S), which is not identified for PTFC. This result indicates that the chemical reactions between FeCl3 and PTFC occurred in the sample preparation. The presence of C-Cl and Fe-O in the sample indicates that FeCl3 transformed into Cl-containing and Fe-containing active functional groups. Both of them contribute to Hg 0 adsorption. Further, it can also be observed that some of the carbon-containing functional groups disappeared after the Hg 0 adsorption reactions, indicating possible reactions between Hg 0 and those carbon-containing functional groups.

XPS Analysis of MPTFC before and after the Reactions
For a deep understating of the Hg 0 adsorption mechanisms of MPTFC, XPS analysis was adopted for the identified C 1s, O 1s, Cl 2p, Fe 2p, and Hg 4f spectrums for the sorbents before and after the reactions. The results are displayed in Figure 9. The C 1s spectrums for MPTFC before and after the Hg 0 adsorption experiments are shown in Figure 9a,b. Four major peaks can be observed, which are assigned as follows: C-C or C-H bonding at 284.3-284.8 eV, C-O or C-OH bonding at 284.9-285.6 eV, C-Cl bonding at 286.0-286.9 eV, and C=O bonding at 288.5-289.1 eV [32]. The relative contents of those C-bonding functional groups before and after the reactions are shown in Table 4. It can be observed that the content of the C-Cl and C=O functional groups decreased slightly after Hg 0 adsorption, while the contents of the C-O, C-C, and C-H functional groups increased after the reactions. This result indicates that C=O probably transformed into C-O or C-OH and further transformed into C-C or C-H groups. The decrease in C-Cl indicates that C-Cl was probably involved in the Hg 0 adsorption process in the reaction. These results are in accordance with the FTIR results. The Fe 2p spectrums of MPTFC before and after the reactions are shown in Figure  9g,h. The Fe 2+ assigned to both 711.3-711.4 eV and 724.82 eV can be observed [36]. In addition, Fe 3+ at 713.9-714 eV, 728.2-728.3 eV, and 732.7-732.8 eV can also be seen. Making a comparison between MPTFC before and after the reactions, it can be found that the content of Fe 2+ increased from 54.24% to 57.93%, while that of Fe 3+ decreased from 45.76% to 42.25%. This result indicates that Fe 3+ was involved in Hg 0 removal during the experiment and then transformed to Fe 2+ . The Fe-bearing oxides, probably Fe2O3 produced by FeCl3 decomposition, may have provided lattice oxygen (Oα) that reacted with Hg 0 on the surface of the sorbent, thus forming the HgO-and Fe 2+ -bearing oxides, such as FeO and Fe3O4 [37].
The Hg 4f spectrums for spent MPTFC are shown in Figure 9i. Several peaks can be observed. Characteristic peaks of Hg 2+ can be observed at 101 eV and 104.35 eV [38], while Hg 0 can be found at 99.76 eV. The obvious peaks of Hg 2+ confirm that Hg 0 was oxidized to form Hg 2+ . The presence of Hg 0 is highly suspected to result from physical adsorption on the surface of MPTFC. The presence of Clwith Hg 2+ is highly suspected to easily form  The O 1s spectrums for MPTFC before and after the Hg 0 adsorption experiments are shown in Figure 9c,d. It can be observed that lattice oxygen (O α ), adsorbed oxygen (O β ), and H 2 O molecular oxygen (O γ ) can all be identified in the spectrums [33], which are assigned as 530.1-530.9 eV, 531.3-532.0 eV, and 532.6-533.5 eV, respectively. It can be observed that the content of O α reduced from 20.9% to 17.29%, while O β decreased from