Effects of Sulfur , Nitric Acid , and Thermal Treatments on the Properties and Mercury Adsorption of Activated Carbons from Bituminous Coals

The influences of sulfur, HNO3, and thermal treatments on the properties and Hg adsorption of activated carbon (AC) derived from bituminous coals were evaluated. These ACs were impregnated with either polysulfide or elemental sulfur at 120, 200, and 600°C. Additional ACs were treated with HNO3 at 95°C to incorporate oxygen functionalities, and then subjected to temperature-programmed desorption to 950°C to remove surface oxygen groups and create nascent active sites on the AC surface. Polysulfide and elemental sulfur impregnation at < 200°C did not improve Hg adsorption. However, improvement in HgCl2 adsorption by up to 97% was observed. Elemental sulfur impregnation at 600°C enhanced Hg and HgCl2 adsorption by up to 42% and 404%, respectively, for ACs derived from a low-organic-sulfur coal. Improvements in Hg adsorption for ACs from high-organic-sulfur coals were not observed after sulfur impregnation. HNO3 and thermal treatments reduced Hg-active groups such as sulfur. Oxygen surface groups and the nascent carbon sites appeared to be more reactive towards other flue gas components than Hg, because HNO3 and thermal treatments led to a reduction in Hg and HgCl2 adsorption. These results suggest that AC’s physical and chemical properties after sulfur treatment influence Hg adsorption, while HgCl2 adsorption is mainly affected by AC’s chemical characteristics.


INTRODUCTION
Hg consisting of Hg 0 , Hg 2+ , and organic Hg is a toxic pollutant that bioaccumulates in human and wildlife through the food chain (Bittrich et al., 2011a, b).The United States' Clean Air Act Amendments of 1990 list Hg as one of the original 188 hazardous air pollutants (HAPs).Hg can travel globally upon release to the environment from both natural and anthropogenic sources.Coal-fired power plants (CFPPs) have been identified as the largest single anthropogenic source category in most countries (Pacyna et al., 2010;Fang et al., 2012;Wu et al., 2012).Low-concentration Hg (1-10 ppb v ) in CFPP flue gas is difficult to separate from the flue gas stream.Numerous approaches have been developed to remove low-concentration Hg from coal combustion flue gases.These approaches are expected to respond to the new Mercury and Air Toxics Standards (MATS) that were announced by USEPA in March 2011 and finalized in December 11, 2011 for coal-and oil-fired electric generating units (U.S. Environmental Protection Agency website, 2011).
Adsorption by porous carbonaceous materials is a promising technique to remove low-concentration Hg from CFPP gas streams.Efforts have been made to increase removal efficiency of activated carbon (AC) for both Hg 0 and Hg 2+ via a carbon fixed bed or a powder injection approach (Staudt and Jozewicz, 2003).Sulfur impregnation has been shown to increase the Hg 0 adsorption capacity of ACs (Krishnan et al., 1994;Korpiel and Vidic, 1997;Liu et al., 1998Liu et al., , 2000;;Hsi et al., 2001Hsi et al., , 2002;;Vitolo and Seggiani, 2002;Feng et al., 2006a, b;Ho et al., 2008).Adsorption onto sulfur-impregnated ACs at low temperature (i.e., < 150°C) was suggested to be the most mature technology for direct capture of gaseous Hg in coal gasification processes (Peise et al., 2008).
Previous research demonstrated that carbonaceous adsorbents with large Hg 0 adsorption capacities (> 5,000 μg-Hg per g-adsorbent; μg/g) can be prepared from select organic precursors by carbonizing, activating, and impregnating them with elemental sulfur between 200 and 650°C (Hsi et al., 2001(Hsi et al., , 2002(Hsi et al., , 2011)).Impregnation at 400°C appeared optimal since the AC's porosity was preserved while active sulfur sites were added.Increasing bulk sulfur content did not necessarily ensure enhanced Hg 0 uptake.Feng et al. (2006b) agreed and reported that excess sulfur may block or fill the adsorbent's pores, hindering Hg 0 adsorption.They proposed that a sulfur monolayer maximizes Hg 0 adsorption capacity in microporous ACs.
Surface oxygen groups on AC can also be effective for enhancing Hg 0 adsorption.These groups are generally in the form of carboxylic, lactonic, phenolic, or quinonic on the AC surface (Bansal et al., 1988).Matsumura (1974) demonstrated that Hg 0 adsorption capacities at 30°C increased approximately 20 times after oxidizing ACs with nitric acid.Li et al. (2003) also proposed that AC's Hg 0 adsorption capacity was correlated with lactone and carbonyl surface.
Removal of Hg 2+ (e.g., HgCl 2 , HgO) is also of interest because 50 to 80% of total vapor-phase Hg in flue gases generated from burning bituminous coals or high chlorine coals is in an ionic form (Carpi, 1997).Lancia et al. (1993) and Ghorishi and Sedman (1998) showed that temperature, HgCl 2 concentration, and relative gas-solid velocity strongly influenced the adsorption of HgCl 2 in fixed beds of Ca(OH) 2 .They concluded that physisorption was the dominant mechanism for adsorption of HgCl 2 .Karatza et al. (1998) also proposed that physisorption was the main mechanism for adsorption of HgCl 2 onto fly ash at < 150°C.However, at > 150°C, the HgCl 2 adsorption mechanism may fall into a "transition zone" between physisorption and chemisorption, based on a study using a mixture of AC and Ca(OH) 2 (Sorbalit; Märker-Gruppe, Germany).
In our previous study, ACs were prepared from highsulfur coals and had comparable Hg 0 and HgCl 2 adsorption performance to that of a commercial AC (Hsi et al., 1998).This study further discusses the impacts of sulfur impregnation (elemental sulfur or polysulfide), HNO 3 treatment, followed by thermal treatments on the physical/ chemical properties and Hg 0 and HgCl 2 adsorption by ACs derived from high-and low-organic-sulfur bituminous coals.Results from this work provide insight into the effectiveness of inherent and impregnated sulfur on Hg adsorption and an understanding of the role of AC surface oxygen groups and nascent carbon sites on Hg 0 and HgCl 2 adsorption in a simulated coal combustion flue gas.

Preparation of Raw ACs
Two high-organic-sulfur (IBC107 and C2) and one loworganic-sulfur (IBC109) Illinois bituminous coals were used to prepare the AC samples (Table 1).IBC107 and IBC109 coals were obtained from the Illinois Basin Coal Sample Program and C2 was obtained from an active coal mine in Illinois.As-received coal samples were ground and sieved to between 0.21 and 1.00 mm in diameter.Benchscale production of ACs was performed in a custom 5-cm ID fluidized-bed reactor with oxidation (air, 225°C), carbonization (N 2 , 400°C), and then steam activation (50% H 2 O/50% N 2 , 825°C) (Hsi et al., 1998).Properties of the ACs are presented in Table 2.
Sample AC109-2 was treated with polysulfide solution at 120°C and ambient pressure for 6 h.The polysulfide solution was prepared by mixing reagent-grade Na 2 S•9H 2 O with powdered elemental sulfur with a 2.5:1 mass ratio in deionized water and then heating the solution to 120°C until all the sulfur dissolved.The prepared ACs were subsequently washed with deionized water to remove residual polysulfide solution and dried in air at 105°C.The sulfur content of resulting samples were set to be 2 or 5 wt%, controlled by the concentration of polysulfide in solution used for impregnation.The resulting ACs were designated as AC109-2-120S-2 and AC109-2-120S-5 (Table 2).

Preparation of ACs with HNO 3 and Thermal Treatments
AC was subjected to a HNO 3 treatment to introduce oxygen groups onto the surface.Sample AC-C2 of 10 g was mixed with 100 mL of 45 vol% (v/v) reagent-grade HNO 3(aq) and heated to 95°C at atmospheric pressure for 2.5 h.The resulting sample (designated as AC-C2-HNO 3 ) was rinsed with deionized water and dried at 105°C in air.
Samples AC-C2 and AC-C2-HNO 3 were subjected to temperature-programmed desorption (TPD) at a linear heating rate of 5 °C/min to 950°C in UHP N 2 .TPD is used not only to quantify oxygen functional groups but also to generate nascent sites on the AC's surface.Thermal treatment at 950°C in oxygen-lean conditions develops nascent carbon sites due to desorption of surface oxygen functional groups.The surface oxygen groups are generally formed on edge active sites that are much more reactive than the carbon atoms in the interior of the basal planes.Therefore, surface oxygen groups are predominantly located on the edges and their removal leaves carbon nascent sites containing unpaired electrons or residual valences, and therefore are more reactive than the basal planes for adsorbing organic and inorganic components, including hydrocarbons and SO 2 (Hoffman et al., 1984;Daley et al., 1997;Lizzio and DeBarr, 1997).
ACs treated with HNO 3 were heated in a 2.5-cm ID stainless steel fixed-bed reactor with 0.5 standard L/min UHP N 2 .
The TPD process removes oxygen functional groups from the AC's surface as CO and CO 2 (Lizzio and DeBarr, 1997).CO 2 is released by decomposition of carboxylic or lactone groups at temperatures below 700°C.CO is released from decomposition of phenols, ethers, carbonyls and quinones at temperatures between 700 and 1000°C.The effluent TPD gas was analyzed for CO and CO 2 concentrations with an infrared analyzer (Rosemount Model 880).The total amount of surface oxygen (as wt% O 2 ) was determined from the amounts of CO and CO 2 evolved during a TPD experiment based on the concentration and gas flow rate.The resulting AC (designated as sample AC-C2-HNO 3 -TPD) was collected for characterization and Hg adsorption tests.
During the HNO 3 and TPD treatments, some of the sulfur in AC was removed.The percent sulfur removal was calculated by: 0 % 100 100 ( ) ( ) where S 0 is the sulfur mass fraction of the original sample, S f is the sulfur mass fraction of the product, and Y is the production yield, which is the product mass divided by the precursor mass.

Sample Characterization
Total surface area (S BET ), total pore volume (V t ), micropore (pore width < 2 nm) surface area (S micro ), micropore volume (V micro ), and pore size distribution of all ACs were determined by N 2 adsorption at 77 K (Micromeritics ASAP2400 analyzer).Samples were degassed at 0.013-0.026atm vacuum and 150°C for 24 h before the N 2 adsorption measurements occurred between 0.001 and 1 atm.S BET was calculated by the Brunauer-Emmett-Teller equation based on ASTM method D4820-96a.S micro and V micro were calculated from t-plot analyses using the Jura-Harkins equation: t = [13.99/(0.0340-log (p/p 0 )] 0.5 (Lippens and de Boer, 1965).
The range of relative pressures used to determine S micro and V micro was based on thickness t values between 0.45 and 0.8 nm.Micropore size distribution was determined using the 3-D model (Sun et al., 1998).Ultimate and sulfur analyses for all samples were measured by LECO MAC-d SC-32 systems according to ASTM methods D5373-93 and D4239-94, respectively.Least-squares analysis of sulfur K-edge X-ray absorption near-edge structure (S-XANES) spectra was used to quantify sulfur groups in select ACs.A detailed description of the principles and application of S-XANES spectroscopy is available elsewhere (Huggins et al., 1993).

Hg Adsorption Test
Bench-scale Hg 0 and HgCl 2 adsorption tests were performed at URS Corporation (Austin, TX) (Carey et al., 1998) and National Taipei University of Technology, Taiwan (Hsi et al., 2011).Briefly, Hg adsorption tests were performed at 150°C in a fixed-bed column (1.3 cm ID) that contained 20 -50 mg adsorbent in 10 g quartz sand (Fig. 1).The simulated CFPP gas stream fed to the fixed bed reactor was controlled at 1-1.2 standard (25°C and 1 atm) L/min with 1600 ppm v SO 2 , 50 ppm v HCl, 12% CO 2 , 7% H 2 O, 6% O 2 and 50 ± 10 Hg 0 or HgCl 2 μg Nm -3 .Hg 0 and HgCl 2 were generated with certified Hg 0 and HgCl 2 permeation tubes (VICI Metronics) in a gas generator at 70 ± 0.1°C to ensure constant Hg diffusion rates.The effluent gas from the fixed-bed column flowed through heated lines to an impinger containing SnCl 2 that reduced any oxidizing Hg compounds to Hg 0 .The gas then flowed through an aqueous buffer solution (Na 2 CO 3 ) to remove SO 2 and HCl and through a moisture trap (i.e., a nefion tube) to remove H 2 O, thus protecting the downstream detector system.Gas exiting the impinger solutions flowed through a gold amalgamation column, where the Hg 0 in the gas was adsorbed and concentrated (< 100°C).After a fixed period of time, Hg 0 was thermally desorbed from the gold (> 750°C) and sent as a concentrated Hg 0 stream to a cold-vapor atomic absorption or a cold-vapor atomic fluorescence spectrophotometer for analysis.
Cumulative Hg adsorption (μg/g) at breakthrough was determined by: ' , , 0 ( ) where m i is the mass of adsorbed Hg, m adsorbent is the total mass of adsorbent, t' is the adsorption time to achieve 100% breakthrough (i.e., equilibrium), C i , in is the inlet Hg concentration, C i , out is the outlet Hg concentration at time t, Q g is the gas flow rate, and Δt is the time interval between measurements during the breakthrough test.Adsorption results for HgCl 2 were normalized to exclude the weight of adsorbed chlorine by multiplying the HgCl 2 adsorption capacities by 0.739, which is the molecular weight of Hg 0 divided by that of HgCl 2 .

Sulfur Impregnation at 120 and 200°C
Table 2 presents the chemical and physical characteristics of the original, sulfur-, HNO 3 -, and thermal-treated ACs.Sample AC109-2 contained 0.81 wt% sulfur and its total sulfur content increased by 160 to 640% after polysulfide or elemental sulfur impregnation at 120 or 200°C.The sulfur contents of sample AC-109-2-200 prepared at 200°C were close to the total amount of sulfur introduced and inherent sulfur content of sample AC109-2, indicating that most of the added sulfur remained in the AC product.Polysulfide impregnation at 120°C was less effective, causing a smaller than expected increase in the bulk sulfur content.
ACs derived from high-organic-sulfur coal had larger Hg 0 and HgCl 2 adsorption capacities than those from the low-organic-sulfur coal (Table 2), in agreement with our earlier work (Hsi et al., 1998).Note that sample AC109-2 showed a marked decrease in equilibrium Hg 0 adsorption capacity after sulfur impregnation at 120 or 200°C.AC109-2 had an equilibrium Hg 0 adsorption capacity of 840 μg/g, while the equilibrium capacity of the sulfur-treated samples was between 63 and 699 μg/g, reflecting a 17-93% decrease in adsorption capacity.Polysulfide impregnation caused an even larger reduction in Hg 0 adsorption capacity than elemental sulfur impregnation.The decrease in Hg 0 adsorption of treated ACs is explained by observing the forms of added sulfur and the changes in the AC's physical properties (Table 2).When elemental sulfur is impregnated into ACs at low temperatures (≤ 200°C), 99.96 mol% of the sulfur molecules exist as S 8 (77 mol%) or S 6 (23 mol%) rings (Tuller, 1954;Berkovitz, 1965).These sulfur forms are less reactive due to a lack of terminal sulfur atoms (Liu et al., 1998).In addition, S 6 and S 8 rings with widths between 7.6 and 8.4 Å depending on if the molecule exists as a ring or chain may be too large to penetrate into AC's microporous structure, especially the ultra-micropores (pore width < 7 Å) (Hsi et al., 2002).Consequently, these molecules may completely fill or block micropores (Table 2).
Results also suggested that polysulfide molecules appeared to be less reactive than S 6 and S 8 for Hg 0 adsorption.These results indicate that the pronounced reduction in the Hg 0 adsorption capacity after low-temperature sulfur impregnation may stem from the changes in both physical and chemical properties of AC after low-temperature sulfur impregnation.On the contrary, 120 and 200°C sulfur impregnation (except for AC109-2-120S-2) improved the equilibrium HgCl 2 capacities by up to 97% (Table 2).These results suggest that HgCl 2 adsorption is more affected by the chemical properties of AC (i.e., surface or bulk sulfur content) than the physical properties of AC (i.e., porosity).

Sulfur Impregnation at 600°C
We showed that sulfur impregnation at temperature ≥ 400°C enhanced Hg 0 adsorption of ACs (Hsi et al., 2001(Hsi et al., , 2002(Hsi et al., , 2011)).In this study, we chose 600°C as the impregnation temperature to deposit more organic sulfur functionalities on the AC surface (Hsi et al., 2001).By doing so, the Hg adsorption results can be compared to those from ACs derived from high-organic-sulfur coals.Sulfur impregnation at 600°C increased the sulfur content of ACs from 0.81-1.64wt% (AC107, AC109-1, and AC109-2) to 12-13 wt% (AC107-600S, AC109-1-600S, and AC109-2-600S) (Table 2).Notably, these samples showed a smaller decrease in S BET , V t and a smaller change in micropore size distribution compared to those of samples treated at 120 and 200°C.Clearly, sulfur impregnation at 600°C has less influence on the physical properties of ACs than the low temperature sulfur treatments used in this work.
Unlike the materials prepared at 120 and 200°C, loworganic-sulfur ACs (i.e., AC109-1 and AC109-2) impregnated with sulfur at 600°C showed considerable improvements in both Hg 0 and HgCl 2 adsorption capacities by up to 42% and 404%, respectively (Table 2).The same impregnation process used on the high-organic-sulfur AC (AC107) resulted in no improvement in equilibrium Hg 0 adsorption capacities, reducing from 1925 to 1643 μg/g, suggesting that ACs from high-organic-sulfur coals do not benefit from additional sulfur impregnation at these conditions (Hsi et al., 1998).However, when the total sulfur content of AC107 increased from 1.64 to 12.7 wt% after sulfur impregnation, the equilibrium HgCl 2 adsorption capacities increased by 50%, from 463 to 692 μg/g.These findings again demonstrate the difference in adsorption mechanisms for Hg 0 and HgCl 2 with sulfur-impregnated ACs.
The form and reactivity of sulfur at high temperature (e.g., 600°C) is different than those at low temperatures (e.g., 120 and 200°C), causing differences in the properties and Hg adsorption capacities of the tested ACs.At 600°C, a large portion of sulfur molecules are in the form of S 2 -S 4 (Berkovitz, 1965), which was estimated to have a diameter between 5.2 and 6.9 Å based on their molar volumes and assuming spherical sulfur molecules (Hsi et al., 2002).Unlike the aforementioned S 6 and S 8 rings that are suspected to mainly physisorb onto the AC surface at < 200°C, the smaller-sized S 2 -S 4 molecules that possess active, terminal sulfur atoms can enter micropores, react with the carbon matrix, and form sulfur-carbon complexes.AC's existing atoms (e.g., sulfur, oxygen, hydrogen), functional groups, and unsaturated sites impact the effectiveness of sulfur impregnation (Puri and Hazra, 1971).At 600°C, impregnated sulfur can substitute onto existing functionalities, bond to unsaturated carbon atoms, or occupy activated sites due to carbon degradation during extended thermal treatment, as was shown using H 2 S, CS 2 , or SO 2 gases (Puri and Hazra, 1971;Blayden and Patrick, 1967).These reactions form organic sulfur on the AC surface.ACs impregnated with elemental sulfur at 600°C may undergo similar sulfur impregnation mechanisms as those using H 2 S, CS 2 or SO 2 gases.S-XANES results showed that the content of organic thiophene, sulfone, and sulfoxide groups increased after 600°C sulfur impregnation (Table 3), supporting that carbon-sulfur bonds formed during the high temperature impregnation.S-XANES results also confirm that organic sulfur plays a critical role in Hg 0 adsorption since AC107 has negligible elemental sulfur content but high equilibrium Hg 0 adsorption.
The above results indicate that sulfur impregnation at 600°C can help increase Hg 0 adsorption capacity of a loworganic-sulfur ACs because it provides active organic sulfur sites without sacrificing the porous structure of the AC.Such approach is able to take advantage of the micropores to concentrate the Hg in the pores and organic sulfur functional groups to then react with the Hg.For HgCl 2 adsorption, it appears that sulfur groups, rather than micropores have a greater impact on HgCl 2 adsorption capacity.

Effects of HNO 3 Treatment
HNO 3 treatment is known to increase surface oxygen groups on ACs.Additionally, HNO 3 treatment is also an ASTM standard method (1988) to remove pyrite and sulfate from sulfur-containing materials, including coals.HNO 3 treatment oxidizes organic sulfide and thiophene into sulfoxide and sulfone in lignite coals (Maes et al., 1996).The dissolution and transformation of sulfur functional groups after HNO 3 treatment may influence the Hg 0 /HgCl 2 adsorption reactivity and capacity of ACs.In this study, the oxygen content of AC was estimated based on TPD results (Fig. 3).HNO 3 treatments increased the oxygen content of AC-C2 from 1.9 to 16.9 wt% as O 2 .The surface oxygen groups evolved from the AC as CO 2 with a peak at 270°C and shoulders between 100 and 800°C.These surface oxygen groups are products of decomposition of carboxylic or lactonic groups (Bansal et al., 1988).CO was released with a peak at 690°C and shoulders between 300 and 950°C, which is the product of decomposition of phenolic or quinonic groups.The bulk sulfur content of AC-C2 decreased from 1.26 to 0.84 wt% after HNO 3 treatment (Table 2).This decrease resulted both from a mass dilution effect of adding 17 wt% oxygen to the AC and from the dissolution of sulfur functional groups by HNO 3 .For AC-C2-HNO 3 (yield = 0.84 based on the starting mass of AC-C2), approximately 44 wt% of the precursor's total sulfur was dissolved in HNO 3 .
HNO 3 treatment also altered the physical properties of ACs, but the changes were not so significant as those caused by low temperature elemental sulfur impregnation (Table 2).S BET of AC-C2 decreased from 702 to 590 m 2 /g most likely due to the addition of oxygen groups into AC's pores.
HNO 3 treatment significantly decreased the Hg adsorption capacities.The equilibrium Hg 0 adsorption capacities for AC-C2 decreased from 1758 to 334 μg/g after HNO 3 treatment (Table 2).Also the equilibrium HgCl 2 adsorption capacities of AC-C2 decreased from 657 to 64 μg/g.These results indicate that the addition of surface oxygen groups combined with the removal of almost half of the original sulfur content decreases Hg adsorption from the coal-combustion flue gas.S-XANES results for samples AC-C2 and AC-C2-HNO 3 suggested that sulfate, sulfone and sulfoxide were unlikely to be the active functional groups for Hg 0 adsorption because the quantities of these groups increased or remained constant after HNO 3 treatment, while Hg adsorption markedly decreased.Elemental sulfur is an oxidation product of pyrite (Duran et al., 1986), which is present in the AC precursors (pyrite sulfur = 0.4-1.1 wt%, Table 1).Thiophene is commonly present in coal-based ACs.Elemental sulfur, thiophene, and organic sulfide content of the AC decreased after HNO 3 treatment (Table 3).It appears, therefore, that elemental sulfur, thiophene, and organic sulfide are the active functional groups for Hg 0 adsorption because  the decrease in the concentration of these groups is consistent with the decrease in Hg 0 adsorption after HNO 3 treatment.
Note that an increase in AC's H 2 O and SO 2 adsorption capacity after HNO 3 treatment may also contribute to the observed decrease in Hg 0 adsorption capacity (Daley et al., 1997;Lizzio and DeBarr, 1997;Sullivan et al., 2007).AC-C2 had an H 2 O content of approximately 5 wt% while AC-C2-HNO 3 contained 14 wt% H 2 O. Hg 0 is insoluble in H 2 O (solubility = 6 × 10 -5 g/L at 25°C) (Schuster, 1991).As a result, only 0.008 μg/g Hg 0 can partition into the HNO 3treated AC if H 2 O saturated the porous structure and Hg 0 only interacted with the water in the pores (i.e., no interaction with the solid phase); potentially increasing the resistance of Hg 0 diffusion into the pores of HNO 3 -treated AC.

Effects of Thermal Treatment of the HNO 3 -Treated Samples
Thermal treatment at 950°C desorbs oxygen and sulfur functional groups from ACs.Although the sulfur content of the TPD product (i.e., AC-C2-HNO 3 -TPD) increased from 0.84 to 0.96 wt% (Table 2), the sulfur removal due to TPD treatment (sample yield = 0.51 based on the mass of AC-C2-HNO 3 ) was 42%.S BET , V t , and V micro increased after thermal treatment due to desorption of oxygen functional groups (Table 2).The equilibrium Hg 0 adsorption capacities for AC-C2-HNO 3 -TPD were 355 μg/g (Table 2), which is comparable to that of AC-C2-HNO 3 .Compared to AC-C2 (1758 μg/g), however, the nascent carbon sites generated after releasing surface oxygen and sulfur groups were not active for Hg 0 adsorption in the simulated coal-combustion flue gas.Nevertheless, the increase in the amount of carbon nascent sites resulted in a greater HgCl 2 adsorption capacity than they were covered with the surface oxygen groups (Table 2).
Both the HNO 3 -and thermal-treated ACs exhibited lower Hg adsorption capacities than untreated, high-organic-sulfur coal-derived ACs.These observations disagree with an earlier study, in which oxygen functional groups were shown to improve Hg 0 adsorption of ACs in a N 2 environment (Matsumura, 1974).In our study, Hg adsorption tests were performed in a simulated coal-combustion flue gas as described above.It is highly likely that some of the flue gas components (e.g., SO 2 and H 2 O) preferentially react with surface oxygen groups and nascent carbon sites (Hoffman et al., 1984;Daley et al., 1997;Lizzio and DeBarr, 1997), resulting in a decrease in Hg adsorption capacity due to competitive adsorption/reaction at the AC's active sites.The interactions between Hg, flue gas components, and surface functional groups on the AC are complex.It remains to be determined what are the factors contributing to the adsorption equilibrium and kinetics of Hg onto ACs and whether a detailed understanding of these factors can be applied to produce highly efficient and cost-effective carbonaceous adsorbents to remove vapor-phase Hg in flue gases of coal-fired power plants.

CONCLUSIONS
This study describes how sulfur, HNO 3 , and thermal treatments impact the Hg adsorption of coal-based activated carbon (AC) in a simulated coal-combustion flue gas.The results showed that equilibrium Hg 0 and HgCl 2 adsorption capacities for low-organic-sulfur ACs in general increased after high-temperature sulfur impregnation.Low-temperature sulfur impregnation (120°C and 200°C) using polysulfide or elemental sulfur decreased Hg 0 adsorption but increased HgCl 2 adsorption.These data suggest that AC's physical and chemical properties influence Hg 0 capture.HgCl 2 adsorption, however, is primarily controlled by the chemical characteristics (e.g., sulfur content) of ACs for the conditions reported here.HNO 3 and thermal treatments removed active sulfur functional groups and caused a decrease in Hg capture.Elemental sulfur, thiophene, and organic sulfide appeared to be the active functional groups contributing to adsorption of Hg onto AC.HNO 3 and thermal treatment of ACs resulted in decreasing Hg 0 and HgCl 2 adsorption capacity, suggesting that Hg and flue gas components potentially compete for the same nascent carbon and surface oxygen groups adsorption sites.

Table 1 .
Ultimate and sulfur analyses of coal samples.

Table 3 .
Sulfur functional groups of original and sulfur-impregnated activated carbons quantified by S-XANES examinations.