Effects of Coal-Fired Flue Gas Components on Mercury Removal by the Mechanochemical S-Modified Petroleum Coke.

In this work, the effects of coal-fired flue gas components (O2, CO2, SO2, and NO) on the Hg0 removal by the promising mercury removal adsorbent mechanochemical S-modified petroleum coke were characterized and analyzed in terms of the Hg0 removal efficiency, mercury adsorption capacity, and mercury mass balance. The results show that the mechanochemical S-modified petroleum coke with a theoretical sulfur content of 21% (named TSC-21) is the best candidate for mercury removal based on the Hg0 removal efficiency, Hg0 removal capacity, and difference ratio of Hg0 removal capacity (anti-interference ability) in the basic and full-component simulated flue gas atmosphere (N2 + O2 + CO2, N2 + O2 + CO2 + SO2 + NO). The maximum value (MV) and stable value (SV) of the Hg0 removal efficiency of TSC-21 in the basic simulated flue gas atmosphere are 99.25% (MV) and 91.17% (SV), respectively. O2, CO2, and NO all promote the Hg0 removal by the adsorbent, but they benefit the Hg0 oxidation while inhibiting the Hg0 adsorption. The promoting effect of O2 on the Hg0 removal by TSC-21 is affected by the reaction time, which is especially obvious after 1 min. The presence of SO2 inhibits the oxidation and adsorption of Hg0, which in turn reduces the Hg0 removal performance of the adsorbent. The improving effects on the oxidative escape of Hg0 by CO2 is higher than that by NO and O2. TSC-21 acts more as an oxidant than an adsorbent for Hg0 removal.

Introduction

Mercury emitted from coal combustion has attracted worldwide attention due to its high toxicity, volatility, environmental persistence, and biomass accumulation. Coal-fired power plants are considered one of the major anthropogenic emission sources of atmospheric mercury.[1−3] Mercury in coal-fired flue gas exists in three forms, including elemental mercury (Hg0), oxidized mercury (Hg2+,) and particle mercury (HgP).[4,5] HgP and Hg2+ can be captured using the dust removal unit (an electrostatic precipitator or a bag filter) and wet flue gas desulfurization system, respectively.[6,7] However, Hg0 is difficult to be removed and easily escapes into the atmosphere because of its volatility and insolubility.[8,9] The activated carbon injection (ACI) technology is considered to be a mature technology for mercury removal from coal-fired flue gas, while the high operation cost limits its wide application.[10−13]

Replacing activated carbon with inexpensive, high-performance adsorbents is a relatively common approach to reducing the cost of ACI technology. As a byproduct of the delayed coking process, the petroleum coke is usually considered as an economical and promising precursor of carbon-based adsorbents.[14] Xiao et al.[15,16] brominated petroleum coke using mechanochemical methods and found that the mercury removal efficiency of raw petroleum coke (RPC) was greatly improved with the highest value of above 99%. Chen et al.[17] used the density functional theory to analyze the mercury removal mechanism by brominated petroleum coke. It found that bromine on the petroleum coke surface enabled HgO and HgBr to be generated easily due to the increase of their adsorption energy and the decrease of their activation energy. She et al.[18] used the SO2 high-temperature impregnation method to modify petroleum coke, due to which the mercury adsorption capacity of the adsorbent increased from 3.41 to 29.54–58.08 μg/g. Zhu et al.[19] prepared the columnar elemental sulfur-impregnated activated petroleum coke. It was found that the elemental sulfur impregnation was dominant for Hg0 adsorption. Therefore, it can be seen that either bromine or sulfur modification can improve the mercury removal performance of RPC. For the bromine modification, the relatively higher cost and easy production of secondary pollution are the drawbacks of this method.[20,21] Mechanochemistry is a relatively novel and ideal modification method with the advantages of being a simple process, solvent-free, environmentally friendly, and highly efficient.[22,23] However, it is rarely used in the preparation of a petroleum coke-based mercury removal adsorbent combined with sulfur-containing modifiers. Based on our previous work,[24] the mechanochemical S-modified petroleum coke was a promising adsorbent for mercury removal from coal-fired flue gas. Zhang et al.[25] characterized the effects of the flue gas component on mercury removal by a sulfur-containing sorbent (used-Fe/SC120) at 90 °C, which indicated that O2 and SO2 inhibited the Hg0 removal due to the lost active sulfur sites and competitive adsorption, which was beneficial for NO in improving the Hg0 oxidation. Ma et al.[26] investigated the mercury removal performance of acid-treated activated coke at 160 °C in different flue gas atmospheres, which showed that NO could promote the mercury removal, while SO2 had varied influences. Huang et al.[11] carried out the mercury removal with bromide (NH4Br)-modified rice husk-activated carbon on a pilot-scale 0.3 MW circulating fluidized bed system. It found that increasing SO2 concentration inhibited the mercury removal efficiency, whereas higher NO concentration promoted that. Xu et al.[27] used the pyrolysis method to prepare the biomass adsorbent modified by the brominated flame retarded, which found that SO2, NO, O2, and HCl were favorable for the mercury removal. Li et al.[28] synthesized sulfur-abundant S/FeS2 by the hydrothermal method to remove the Hg0 from coal-fired flue gas at low temperature. This indicated that the presence of 50–150 ppm SO2 or 75 ppm NO had negligible effects on mercury removal by the adsorbent. Li et al.[29] studied the influence of acidic gases (CO2, SO2, NO, and HCl) on mercury removal by a raw activated carbon, which showed that NO and HCl could improve the mercury removal, while SO2 was the negative factor. It can be seen that there have been some studies on the influence of flue gas components on the mercury removal performance by the adsorbents. NO, O2, and HCl have a certain promoting effect on the mercury removal, while the influence of SO2 on that is doubtful. In fact, the effects of flue gas components on the mercury removal performance of adsorbents are related to the adsorption temperature, the concentration of flue gas components, and so on. Therefore, it is necessary to investigate the influence of flue gas components on the mechanochemical S-modified petroleum coke adsorbent developed in our previous work.[24]

In this work, the effects of coal-fired flue gas components on mercury removal by the mechanochemical S-modified petroleum coke were characterized on a simulated flue gas (SFG) fixed bed mercury removal test bench, which were analyzed in terms of Hg0 removal efficiency, mercury adsorption capacity, and the mercury mass balance. The mercury temperature-programed desorption (Hg-TPD) analysis was used to obtain the mercury speciation and mercury adsorption on adsorbents after use.[30] The main contents include (1) screening of optimal mechanochemical S-modified petroleum coke; (2) effect of each flue gas component on mercury removal performance; and (3) comparative analysis based on the mercury mass balance. The main purpose is to comprehensively evaluate the mercury removal performance of mechanochemical S-modified petroleum coke and provide technical support and theoretical guidance for its industrial application.

Experimental Section

Sample Preparation

A kind of high-sulfur petroleum coke was selected to be the precursor of the mercury removal adsorbent, the proximate and elemental analyses of which are shown in Table . It is shown that carbon is the main content, and the sulfur content (5.89 wt %) is high. This facilitates the preparation of high-performance mercury removal adsorbents.[31] The high-sulfur petroleum coke was modified with the elemental sulfur (S) having a purity greater than 99.9% using the mechanochemical preparation method. The omni-directional planetary ball mill was the main equipment in the sample preparation process. The rotation speed of 600 rpm, the revolution speed of 300 rpm, and the milling time of 60 min were selected for the adsorbent preparation. The material of the grinding ball was zirconia, and the mass ratio of balls to the mixture of high-sulfur petroleum coke and S was 15:1. The theoretical sulfur content (TSC) was used as the basis for the quantification of high-sulfur petroleum coke and S in the mechanochemical S-modified petroleum coke adsorbent. It was defined as the ratio of the sum mass of sulfur in the petroleum coke and the modifier to the sum mass of the petroleum coke and the modifier, given in percentage. In this work, the mechanochemical S-modified petroleum coke adsorbents with different TSCs were prepared, which were named as TSC-9, TSC-13, TSC-17, TSC-21, and TSC-25, respectively. For example, TSC-9 represented the mechanochemical S-modified petroleum coke adsorbent for which the TSC was 9%.

Table 1

Proximate and Elemental Analysis of the High-Sulfur Petroleum Cokea

proximate analysis (wt %)				elemental analysis (wt %)
Mad	Aad	Vad	FCad	Cd	Hd	Od	Nd	Sd	Cld
0.52	0.19	9.83	89.46	87.30	3.49	1.90	1.23	5.89	0.01

Note: ad, air-dried basis; d, dried basis.

Mercury Removal Test

The schematic diagram of the SFG fixed bed mercury removal experimental device is shown in Figure . It consisted of a gas supply unit, a mercury generator, a flue gas preheater, a fixed bed reactor, a system for online monitoring of Hg0, and an exhaust gas treatment unit. Several mass flow meters (Beijing Sevenstar D07-19B, China) were used to control the flow rate of the SFG. The Hg0 concentration was measured and recorded using the online mercury concentration analyzer (Lumex RA-915M, Canada). The flow rate of the SFG was set as 1 L/min, and the initial Hg0 concentration was 51.5 ± 1.5 μg/m3. The air velocity was about 0.15 m/s in the fixed bed. The amount of the adsorbent used for each set of mercury removal experiment was 100 mg, the particle size of which was about 200–400 μm. In the experimental process, N2 was used to carry Hg0 and balance the total flow. The temperatures of the preheated SFG and Hg0 adsorption were all kept at 150 °C.

Figure 1

Schematic diagram of the SFG fixed bed mercury removal experimental device.

The experimental conditions designed in this work are shown in Table . The component of SFG-1 was a basic SFG under the ideal combustion condition of carbon and air, which included only N2, O2, and CO2. The component of SFG-2 was the full-component SFG, in which the concentrations of O2, CO2, SO2, and NO are all typical values for the coal combustion.[1,32,33] SFG-1 and SFG-2 were selected for the screening of the optimal mechanochemical S-modified petroleum coke. The operating conditions of SFG-3 to SFG-7 were used to study the effect of each flue gas composition on the mercury removal performance of the optimal mechanochemical S-modified petroleum coke.

Table 2

Experimental Conditions Designed in This Work

no.	simulated flue gas components
SFG-1	N2 + 6% O2 + 12% CO2
SFG-2	N2 + 6% O2 + 12% CO2 + 800 ppm SO2 + 250 ppm NO
SFG-3	N2
SFG-4	N2 + 6% O2
SFG-5	N2 + 12% CO2
SFG-6	N2 + 800 ppm SO2
SFG-7	N2 + 250 ppm NO

Evaluation Index and Relevant Characterizations

Hg0 removal efficiency and Hg0 removal capacity were adopted to evaluate the mercury removal performance of the adsorbent, which are defined in Formulas and 2 shown as follows.where ηt represents the Hg0 removal efficiency, given in percentage; qt represents the Hg0 removal capacity, given in micrograms per gram; Cin and Cout-e represent the Hg0 concentrations at the inlet and outlet of the fixed bed, respectively, given in micrograms per cubic meter; QSFG represents the total flow of the SFG, given in cubic meters per minute; mcoke represents the mass of the adsorbent, given in grams; and t1 represents the time for the mercury removal, given in minutes.

For the screening of the optimal mechanochemical S-modified petroleum coke in the atmosphere of SFG-1 and SFG-2, the difference ratio of Hg0 removal capacity was introduced, as shown in Formula .where r represents the difference ratio of Hg0 removal capacity, given in percentage and qSFG–1 and qSFG–2 represent the Hg0 removal capacities in the atmospheres of SFG-1 and SFG-2, respectively, given in micrograms per gram.

For the influence of O2, CO2, SO2, and NO on the mercury removal performance of the typical sample, the Hg-TPD test was carried out. The temperature-programed furnace (OTF-1200X, China) and the on-line mercury concentration analyzer (Lumex RA-915M, Canada) were used to analyze the mercury forms form the same adsorbent in different atmospheres. The samples were heated from room temperature to 700 °C with a heating rate of 5 °C/min in the temperature-programed furnace in N2 with a flow rate of 100 mL/min. According to the Hg-TPD curve, the mercury adsorption capacity could be calculated based on Formula .where qa represents the amount of Hg0 released during the Hg-TPD test, given in micrograms per gram; Cad represents the Hg0 concentration released from the adsorbent in the Hg-TPD process, given in micrograms per cubic meter; QTPD represents the flow of N2, given in cubic meters per minute; mcoke represents the mass of the adsorbent, given in grams; and t2 represents the time of the Hg-TPD test, given in minutes.

Results and Discussion

Screening of the Optimal Mechanochemical S-Modified Petroleum Coke

The mercury removal performance of mechanochemical S-modified petroleum coke with different TSCs in the atmospheres of SFG-1 and SFG-2 is shown in Figure . From Figure a, Hg0 removal efficiency of mechanochemical S modified petroleum coke with different TSCs achieves the maximum value (MV, 74.58–99.25%) quickly and decreases smoothly to their respective stable value (SV, about 57.00–96.80%) in the atmosphere of SFG-1. Both the MV and SV of Hg0 removal efficiency of the adsorbent change regularly with the increase of TSC in the atmosphere of SFG-1. However, the MV and SV of the Hg0 removal efficiency of the adsorbent do not have obvious regularity with the increase in the TSCs in the atmosphere of SFG-2. This indicates that the mercury removal performance of the adsorbent is affected by the full-component SFG. For the SFG-2 atmosphere, the Hg0 removal efficiency of TSC-21 is slightly higher than that of TSC-9 in the initial reaction stage (within about 20 min) and then is almost the same with further increase in the reaction time. This shows that the excess S addition does not significantly improve the mercury removal ability of the adsorbent, which is also reflected in the SFG-1 atmosphere. Comparing the Hg0 removal efficiencies of TSC-9 and TSC-21 in the SFG-1 and SFG -2 atmospheres, it can be seen that 800 ppm SO2 and 250 ppm NO have a more obvious inhibitory effect on the mercury removal ability of TSC-21, which may originate from the high SO2 concentration in the SFG-2 atmosphere.[34,35] The excessive addition of the S modifier can increase the surface active sites of the adsorbent while also deteriorating the surface pore structure. The high concentration of SO2 in the SFG-2 atmosphere may form SO42- under the action of surface active sites (such as oxygen-containing functional groups) and O2 in the flue gas, which hinders the Hg0 removal by the active sites. These results in the lower SV of Hg0 removal efficiency of TSC-25 in the SFG-2 atmosphere. From Figure b, the increase in the amplitude of the Hg0 removal efficiency of the adsorbent with the TSC increasing from 9 to 17% changes obviously from 74.58% (MV) and 57.00% (SV) to 98.47% (MV) and 91.00% (SV), respectively, in the atmosphere of SFG-1, which will tend to be stable with a further increase in the TSC with the values of 98.47%–99.25% (MV) and 91.00%–95.80% (SV). For Hg0 removal efficiency in the atmosphere of SFG-2, the changing ranges of the MV and SV are relatively smaller with the increase of TSC, which are 47.79–55.23% and 25.60–34.70%, respectively. It can be seen that the presence of NO and SO2 in the SFG not only reduces the mercury removal efficiency of the adsorbent but also causes the mercury removal performance of the adsorbent to be irregular with the increase of TSC. The Hg0 mercury removal capacity and difference ratio of the adsorbent with different TSCs is summarized in Figure c, which shows that the difference ratio of Hg0 removal capacity follows the order of TSC-9 (34.34%) < TSC-21 (55.74%)

Figure 2

Mercury removal performance of mechanochemical S-modified petroleum coke with different TSCs in the atmospheres of SFG-1 and SFG-2. (a) Hg0 removal efficiency vs time; (b) Maximum and stable values of Hg0 removal efficiency vs TSC; and (c) Hg0 mercury removal capacity and difference ratio vs TSC.

In addition, the mercury removal work of some modified carbon-based adsorbents is listed in Table . It can be found that these modified carbon-based adsorbents can achieve the Hg0 removal efficiency of about 90%, among which 40ZIS/CN, 1M-500, and TSC-21 all have the values of above 96%. In the case of similar mercury removal efficiency, mechanochemistry has a more convenient preparation method than impregnation or impregnation and pyrolysis, with a simple operation and a short preparation period. Comparing FA-MC-Br and TSC-21, it can be seen that the industrial byproduct petroleum coke has excellent potential as a support material for high-performance mercury removal adsorbents. Therefore, in view of the mercury removal ability, preparation method, and support material, TSC-21 has a good application prospect for flue gas mercury removal.

Table 3

Comparison of Mercury Removal Performance between Modified Carbon-Based Adsorbents and TSC-21

sample	precursor	modifier	preparation method	mercury removal conditions	ηt %	ref
40ZIS/CN	g-C3N4 nanosheet	ZnIn2S4	impregnation	∼82.7 μg/m3 Hg0, N2, 120 °C	∼98.87%,	(36)
BC-8S2Cl2-IM	sawdust coke	S2Cl2	impregnation	50 μg/m3 Hg0, N2 + 6% O2 + 12% CO2, 150 °C	91.94%	(37)
1M-500	sewage sludge	ZnCl2	impregnation and pyrolysis	70 μg/m3 Hg0, N2, 140 °C	∼96.5%	(38)
FA-MC-Br	fly ash	NH4Br	mechanochemistry	54 μg/m3 Hg0, N2 + 4% O2, 150 °C	∼88%	(39)
TSC-21	petroleum coke	elemental sulfur	mechanochemistry	51.5 μg/m3 Hg0, N2 + 6% O2 + 12% CO2, 150 °C	∼99.25%	this work

Effect of Each Flue Gas Component on Mercury Removal Performance

Effect of O2

The effect of O2 on the mercury removal performance of TSC-21 is shown in Figure . In Figure a, it is shown that the presence of 6% O2 improves the Hg0 removal efficiency of TSC-21 in the N2 atmosphere, which increases from 73.41% (MV) and 37.73% (SV) in SFG-3 to 99.37% (MV) and 98.32% (SV) in SFG-4. The Hg0 removal efficiency in the atmosphere of SFG-3 is higher than that in SFG-4 within the first 3 min, which both reach a large value at around 7 min in the two atmospheres. After 7 min, the Hg0 removal efficiency of TSC-21 in the atmosphere of SFG-4 increases smoothly from 92.84 to 98.32% (SV), while that in the atmosphere of SFG-3 decreases relatively quickly from 73.41% (MV) to 37.73% (SV). In Figure b, it is shown that the Hg0 removal capacity (86.09 μg/g) of TSC-21 in the atmosphere of SFG-4 is nearly 2 times that (46.41 μg/g) in SFG-3 within the reaction time of 180 min. The reason of this promotion effect in the presence of O2 is the heterogeneous reaction between Hg0 and O2 occurring on the adsorbent surface.[40] Moreover, the Hg0 removal capacity of TSC-21 in the atmosphere of SFG-4 within the first 7 min is lower than that in the atmosphere of SFG-3. It can be found that the reaction time is an important factor affecting O2 in promoting or inhibiting the Hg0 removal performance of TSC-21. This may be due to the competitive adsorption of O2 and Hg0 on the adsorbent surface in the initial stage under the O2 atmosphere, which inhibits the contact of Hg0 with the adsorbent surface. After 3 min, the adsorbed O2 has been transformed to active sites favorable for the Hg0 oxidation on the adsorbent surface, which thereby enhances its mercury removal ability.

Figure 3

Effect of O2 on the mercury removal performance of TSC-21. (a) Hg0 removal efficiency vs time and (b) Hg0 removal capacity vs time.

The Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-4 are shown in Figure . From Figure a, the obvious peak of the Hg-TPD curve in the two atmospheres occurs near 300 °C. According to previous research studies,[41−43] the mercury compound corresponding to this decomposition peak may be HgS or HgO. Considering the TSC-21 adsorbent containing a certain amount of sulfur, HgS should be the dominated mercury form. From Figure b, it is shown that the mercury adsorption capacity of TSC-21 in the atmosphere of SFG-4 (4.23 μg/g) is lower than that in the atmosphere of SFG-3 (4.46 μg/g). For the SFG-3 atmosphere, the Hg0 removal capacity of TSC-21 (46.41 μg/g) is much larger than its mercury adsorption capacity (4.46 μg/g), which indicates that a large amount of Hg0 is oxidized in the mercury removal process in the N2 atmosphere. This may be caused by the presence of oxygen-containing functional groups on the adsorbent surface [such as C–O and C–O–C (1250–1500 cm–1) and C=O (1500–1750 cm–1),[15] as shown in Figure .[44] Combining with Figures and 4, it can be seen that the presence of O2 is beneficial for the improvement of the mercury removal performance of TSC-21 (including Hg0 removal efficiency and Hg0 removal capacity), where the mercury adsorption capacity of TSC-21 in the SFG-3 atmosphere is higher than that in the SFG-4 atmosphere. This shows that the existence of O2 mainly promotes the heterogeneous reaction of Hg0 on the adsorbent surface, but the Hg2+ after the heterogeneous oxidation will leave the adsorbent and enter the flue gas leaving the fixed bed. Moreover, the presence of O2 plays a dominate role in the Hg0 oxidation process.

Figure 4

Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-4. (a) Hg-TPD results and (b) mercury adsorption capacity.

Figure 5

Fourier transform infrared spectra of TSC-21.

Effect of CO2

The effect of CO2 on the mercury removal performance of TSC-21 is shown in Figure . It can be seen that the presence of CO2 is beneficial for the improvement of the mercury removal performance of TSC-21, where Hg0 removal efficiency (MV, 96.72%) and Hg0 removal capacity (80.71 μg/g) in SFG-5 is higher than that in SFG- 3. As shown in Figure a, the Hg0 removal efficiency of TSC-21 in the atmosphere of SFG-5 quickly reaches 96.72% (MV) and then slowly decreases to 71.57% (SV). The Hg0 removal efficiency of the two atmospheres is similar before 1 min, but the promotion effect of CO2 is obvious after 1 min. From Figure b, the Hg0 removal capacity of TSC-21 in the SFG-3 and SFG-5 atmospheres has an inflection point at around 1 min, where the distance between them gradually widened with the increase in the reaction time after 1 min. Therefore, the promoting effect of CO2 on the Hg0 removal of TSC-21 is not affected by the reaction time.

Figure 6

Effect of CO2 on the mercury removal performance of TSC-21. (a) Hg0 removal efficiency vs time and (b) Hg0 removal capacity vs time.

The Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-5 are shown in Figure . From Figure a, the Hg-TPD curve of the used TSC-21 in the atmosphere of SFG-5 agrees well with that in the atmosphere of SFG-3. The peak temperature of the Hg-TPD curve in the two atmospheres is around 300 °C, which corresponds to HgS combined with the reaction conditions. Thus, the presence of CO2 does not promote the generation of new mercury compounds, where HgS is still the main mercury form on the used TSC-21 in the atmosphere of SFG-5. From Figure b, the mercury adsorption capacity (2.37 μg/g) of TSC-21 in the SFG-5 atmosphere decrease obviously compared to that (4.46 μg/g) in the SFG-3 atmosphere. In the SFG-5 atmosphere, CO2 will partially fill the surface microporous structure of the mechanochemical S-modified petroleum coke adsorbent on one hand, which inhibits the adsorption of flue gas mercury.[40] On the other hand, CO2 will react with the carbon on the adsorbent surface to form oxygen-containing functional groups, which is favorable for Hg0 oxidation.[40,45−47] Therefore, for mechanochemical S-modified petroleum coke, the presence of CO2 promotes the Hg0 oxidation on the adsorbent surface but inhibits the adsorption of flue gas mercury on the adsorbent. It results in the improvement of mercury removal ability of TSC-21 but a decrease in the mercury adsorption capacity.

Figure 7

Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-5. (a) Hg-TPD results and (b) mercury adsorption capacity.

Effect of SO2

The effect of SO2 on the mercury removal performance of TSC-21 is shown in Figure . From Figure a, the Hg0 removal efficiency of TSC-21 decreases from 73.41% (MV) and 37.73 (SV) in the SFG-3 atmosphere to 53.80% (MV) and 23.11% (SV) in the SFG-6 atmosphere. The curves of Hg0 removal efficiency of TSC-21 in the two atmospheres almost coincide within the first 1 min, which will reach the respective MV and SV at different changing rates after that. From Figure b, the Hg0 removal capacity of TSC-21 decreases from 46.41 μg/g in the SFG-3 atmosphere to 27.59 μg/g in the SFG-6 atmosphere within the same reaction time of 180 min. There is the same inflection point at the reaction time of 1 min, which is also the demarcation point of Hg0 removal capacity with different changing rates in the atmosphere of N2 and N2 + SO2, respectively. Thus, the presence of SO2 has a significant inhibitory effect on the mercury removal of mechanochemical S-modified petroleum coke, and the inhibitory effect is obvious after the reaction time of 1 min. This may be because there is a competitive adsorption between SO2 and Hg0 on the adsorbent surface in the initial stage, but the surface active sites still play a role in the mercury removal process. With the prolongation of the reaction time, the surface active sites are further covered. There also exists the reduction of oxidized mercury by SO2, which leads to a significant decrease in the mercury removal ability.

Figure 8

Effect of SO2 on the mercury removal performance of TSC-21. (a) Hg0 removal efficiency vs time and (b) Hg0 removal capacity vs time.

The Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-6 are shown in Figure . From Figure a, the Hg-TPD curves of the used TSC-21 in the two atmospheres of SFG-3 and SFG-6 have good similarities. The presence of SO2 does not promote the generation of sulfur-containing mercury compounds, where HgS is still the main mercury compound on the used TSC-21 in the presence of SO2. From Figure b, the mercury adsorption capacity (4.46 μg/g) of the adsorbent in the SFG-3 atmosphere is higher than that (3.93 μg/g) in the SFG-6 atmosphere. This shows that the presence of SO2 can inhibit the flue gas mercury adsorption. The inhibitory effect of SO2 on the mercury removal performance of mechanochemical S-modified petroleum coke may be due to the following reasons. (1) SO2 competes with Hg0 for adsorption on the adsorbent surface, which occupies part of the active sites for mercury removal.[25,48] (2) HgO generated by the reaction between Hg0 and the oxygen functional group on the adsorbent surface can be reduced to Hg0 by SO2, as described in Formula .[29,49] The generated SO3 will further inhibit the adsorption of Hg0 on the adsorbent surface.[50] Compared to the Hg0 removal capacity and mercury adsorption capacity of TSC-21 in the SFG-3 atmosphere, it can be found that the decrease (18.82 μg/g) in the Hg0 removal capacity is higher than that (0.53 μg/g) in the mercury adsorption capacity in the SFG-6 atmosphere. This indicates that the presence of SO2 reduces both the oxidation and adsorption of mercury, which leads to a decrease in the Hg0 removal efficiency.

Figure 9

Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-6. (a) Hg-TPD results and (b) mercury adsorption capacity.

Effect of NO

The effect of NO on the mercury removal performance of TSC-21 is shown in Figure . From Figure a, the presence of NO is beneficial for the improvement of the Hg0 removal efficiency, which will increase from 73.41% (MV) and 37.73 (SV) in the SFG-3 atmosphere to 86.63% (MV) and 79.27% (SV) in the SFG-7 atmosphere. The main promoting effect on the increasing rate of Hg0 removal efficiency occurs after the reaction time of 1 min. The Hg0 removal efficiency of TSC-21 in the atmosphere of NO is stable in the range of 77%–80% after the reaction time of 30 min. From Figure b, the Hg0 removal capacity (72.69 μg/g) in the atmosphere of SFG-7 is higher than that (46.41 μg/g) in the atmosphere of SFG-3 in the whole reaction process. There is also an inflection point in the Hg0 removal capacity curve, which exists around the reaction time of 1 min. Similar to CO2, the effect of NO on the mercury removal of TSC-21 is also not limited by the reaction time, which always has the positive factor for the mercury removal performance.

Figure 10

Effect of NO on the mercury removal performance of TSC-21. (a) Hg0 removal efficiency vs time and (b) Hg0 removal capacity vs time.

The Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-7 are shown in Figure . From Figure a, it is shown that the Hg-TPD curve of the used TSC-21 in the SFG-7 atmosphere has two obvious peaks at around 312 °C and 375 °C, which usually correspond to HgO and/or HgS.[5,41,42] In addition, a small peak of the Hg-TPD curve in the range of 445–500 °C in the SFG-7 atmosphere indicates that there is a relatively small proportion of Hg(NO3)2.[41,42] From Figure b, it can be found that the mercury adsorption capacity (3.16 μg/g) of TSC-21 in the SFG-7 atmosphere is still lower than that (4.46 μg/g) in the SFG-3 atmosphere. When NO exists in the SFG, NO will form NO2 under the action of surface oxygen (O*) on the carbon-based adsorbent surface, and the NO2 can promote the oxidation of mercury to form HgO and Hg(NO3)2, as shown in Formula −8.[51] In addition, the generated mercury compound Hg(NO3)2 has a certain volatility.[52] Therefore, the addition of NO promotes the Hg0 oxidation by the mechanochemical S-modified petroleum coke, while generating a more volatile mercury compound [Hg(NO3)2]. NO occupies the adsorption site of flue gas mercury on the adsorbent surface. This eventually led to an improvement in the mercury removal performance but a decrease in the mercury adsorption capacity.

Figure 11

Hg-TPD results and mercury adsorption capacity of the used TSC-21 in SFG-3 and SFG-7. (a) Hg-TPD results and (b) mercury adsorption capacity.

Comparative Analysis Based on the Mercury Mass Balance

The mercury mass balance for mercury removal by the adsorbent is established, which is shown in Figure . The mercury mass balance in the Hg0 removal process can be described by Formulas and 10.where min represents the Hg0 mass flow at the inlet of the fixed bed, given in micrograms per minute; mout represents the mercury mass flow at the outlet of the fixed bed, given in micrograms per minute; mout-e and mout-ox represent the mass flow of Hg0 and Hg2+ at the outlet of the fixed bed, respectively, given in micrograms per minute; and mad represents the mass of mercury adsorbed on the adsorbent per unit time, given in micrograms per minute.

Figure 12

Schematic diagram of the mercury mass balance for mercury removal by the adsorbent.

Then, the amount of Hg2+ escaped is introduced and described by Formula .

The escaping rate of Hg2+ is introduced and described by Formula .where qo represents the escaping amount of Hg2+, given in micrograms per gram and ro represents the escaping rate of Hg2+, given in percentage.

The mercury removal performances of TSC-21 in the atmospheres of SFG-3 to SFG-7 are summarized in Table . It shows that the escaping rates of Hg2+ in the five different atmospheres are 85.76–97.06%, where the rate in the N2 + 12% CO2 atmosphere has the highest value, while that in the N2+ 800 ppm SO2 atmosphere has the lowest value. Zhou et al.[30] conducted an experimental study on the effects of flue gas components on the oxidation and adsorption of Hg0 by the NH4Br-modified fly ash, which used 10% SnCl2 to reduce the Hg2+ in the exhaust gas to Hg0 and measured the Hg2+ concentration in the exhaust gas combined with the subtraction method. It also found a certain percentage of oxidative escaping mercury. The mercury adsorption capacities of TSC-21 in the five different atmospheres are 2.37–4.46 μg/g, which are all smaller than that in the pure N2 atmosphere (4.46 μg/g). However, the Hg2+ escaping amounts in the five different atmospheres are in the range 23.66–81.86 μg/g, which is higher than the mercury adsorption capacity. For the mechanochemical S-modified petroleum coke, the poor surface structure is the main reason for its lower mercury adsorption capacity. The oxygen-containing functional groups and active sulfur on the adsorbent surface are the main internal reasons for its high Hg0 oxidation ability. Overall, TSC-21 acts more as an oxidant than an adsorbent for Hg0 removal. From the Hg0 removal capacity (qt), it can be seen that O2, CO2, and NO all promote Hg0 removal by the mechanochemical S-modified petroleum coke, while SO2 plays an inhibitory role. The escaping rate of Hg2+ (97.06%) in the N2 + 12% CO2 atmosphere is greater than that in the atmosphere of N2 + 250 ppm NO (95.65%) ≈ N2 + 6% O2 (95.09%), which indicates that the improving effects on the oxidative escaping of Hg0 by CO2 is higher than that by NO and O2. The Hg0 removal capacity (80.71 μg/g) in the N2 + 12% CO2 atmosphere is larger than that (72.69 μg/g) in the atmosphere of N2 + 250 ppm NO, while the mercury adsorption capacity on the adsorbent in the two atmospheres has the opposite trend. This further illustrates that CO2 promotes the oxidation of mercury but inhibits the adsorption of flue gas mercury on the adsorbent.

Table 4

Mercury Removal Performance of TSC-21 in the Atmospheres of SFG-3 to SFG-7

no.	SFG components	qt μg/g	qa μg/g	qo μg/g	ro %
SFG-3	N2	46.41	4.46	41.95	90.39
SFG-4	N2 + 6% O2	86.09	4.23	81.86	95.09
SFG-5	N2 + 12% CO2	80.71	2.37	78.34	97.06
SFG-6	N2 + 800 ppm SO2	27.59	3.93	23.66	85.76
SFG-7	N2 + 250 ppm NO	72.69	3.16	69.53	95.65

Conclusions

Considering the Hg0 removal efficiency, Hg0 removal capacity, and difference ratio of Hg0 removal capacity (anti-interference ability) in the SFG-1 and SFG-2 atmospheres, the mechanochemical S-modified petroleum coke adsorbent with the TSC of 21% is the best candidate for mercury removal. The MV and SV of Hg0 removal efficiency of TSC-21 in the atmospheres of SFG-1 and SFG-2 are 99.25% (MV) and 91.17% (SV) and 55.23% (MV) and 34.69% (SV), respectively. O2, CO2, and NO all promote the Hg0 removal by TSC-21, and the promotion effect of O2 on Hg0 removal is limited by the reaction time (there is an obvious promotion effect after the reaction time of 1 min). All these three components promote the Hg0 oxidation on the TSC-21 surface but inhibit the adsorption of flue gas mercury on the adsorbent. SO2 has an obvious inhibitory effect on the mercury removal from TSC-21 especially after 1 min of reaction time, which results from the decrease in both the oxidation and adsorption of Hg0. The escaping rates of Hg2+ in the five different atmospheres (SFG-3 to SFG-7) are 85.76%–97.06%, and TSC-21 acts more as an oxidant than an adsorbent for Hg0 removal. The improving effects on the oxidative escape of Hg0 by CO2 is higher than that by NO and O2.

In the follow-up research, the pore structure of the adsorbent should be improved as much as possible to provide more physical adsorption or chemical active sites to further improve the Hg0 removal ability. Based on the position where the traditional ACI technology is used in the coal-fired power plant, the escaping of oxidized mercury will be captured by the subsequent dust collectors or wet desulfurization units. Therefore, it is necessary to comprehensively analyze the environmental stability, emission requirement, and reusability of fly ash and the used adsorbent, desulfurization wastewater, and desulfurization gypsum after mercury removal by the adsorbent injection, in addition to paying attention to the mercury concentration in the flue gas emitted to the atmosphere.