ROLE OF SURFACE FUNCTIONAL GROUPS IN THE CAPTURE OF
ELEMENTAL MERCURY AND MERCURIC CHLORIDE BY ACTIVATED CARBONS
S. Behrooz Ghorishi* and Robert M. Keeney
ARCADIS Geraghty & Miller, Inc., 4915 Prospectus Drive, Durham, NC 27713
Brian K. Gullett
U.S. Environmental Protection Agency, Office of Research and Development, National Risk
Management Research Laboratory, Air Pollution Prevention and Control Division (MD-65),
Research Triangle Park, NC 27711
Abstract
A laboratory-scale, fixed-bed apparatus was used to study the role of surface functional
groups (SFGs) in the capture of mercuric chloride (HgCl2) and elemental mercury (Hg°) in
nitrogen (N2) prior to flue gas atmosphere studies. This investigation focused on two activated
carbons (FGD and PC-100, Norit Americas, Inc.) with different physical and chemical
characteristics. SFGs may be acidic or alkaline. The acidic SFGs include carboxyl, lactone,
hydroxyl, and carbonyl functionalities while alkaline properties are believed to arise from two
types of active sites, pyrone and chromene structures. The acidic and alkaline SFGs of as-
received FGD and PC-100 were determined using surface titration techniques. An attempt was
made to correlate the number density of these SFGs to their ability to sorb Hg° and HgCl2. The
activated carbons were then treated with acid and alkaline washes to neutralize their alkaline and
acidic SFGs, respectively. The mercury capture capabilities of these treated carbons were then
compared to those of untreated, as-received samples. The initial hypothesis was that the number
of alkaline SFGs should correlate with HgCl2 capture. This was proven not to be the case; rather,
the concentration of surface chlorine (CI) sites was related to the HgCl2 and Hg° capture. Energy
Dispersive X-ray Spectroscopy (EDXS) analysis confirmed the existence of a correlation
between the increase in HgCl2 and Hg° uptake of acid-treated activated carbon to the increase
in surface CI sites. These CI sites were strongly bonded to the surface of carbon. It was
determined that acidic and alkaline SFGs play no role in the adsorption of Hg° and HgCl2 by
activated carbons.
Introduction
There is continuing concern over the anthropogenic emissions of mercury species from
combustion sources. One of the control technologies for removal of mercury from flue gas is the
adsorption of mercury species through injection of solid sorbents. The attractive features of
adsorption processes on dry sorbents has led researchers to evaluate adsorption kinetics and the
sorbent capacity of many different solid sorbents [ 1 -9]. The form of mercury species will dictate
the mechanism of its capture and its ultimate environmental fate. Similarly the form and
concentration of mercury species will determine the rate of its reactions with different solid
sorbents.
Typically, elemental mercury (Hg°) is the prevailing form of mercury in emissions from
1-Ghorishi

-------
coal combustion processes [10], although exceptions have been noted [11], Based on collected
information from the field using existing mercury speciation measurement methods, it appears
that combustion of western subbituminous and lignite coals results in a flue gas dominated by
Hg° and combustion of eastern bituminous coals leads to a flue gas dominated by oxidized forms
of mercury. The form of mercury species is important because oxidized forms of mercury, most
probably dominated by mercuric chloride (HgCl2), are more easily controlled via solid
adsorption [1-4,12] than the elemental (Hg°) form. Note that a recent study conducted by Olson
et al. [13] demonstrated that another form of oxidized mercury, mercury nitrate monohydrate
[Hg(N03)2*H20], can exist in a stable form in combustion flue gases. Oxidized forms of mercury
(chloride and nitrate) have higher solubilities in water which make them more amenable to
control by wet scrubbers. Research has been conducted to determine conditions that favor
formation of oxidized mercury species in combustion processes [14-16], Since the focus of this
study was on the capture of Hg° and HgCU by activated carbons, a brief review of adsorption
processes on activated carbons is warranted.
Mercury Species Adsorption by Activated Carbons
Activated carbons have been found to be effective sorbents for both Hg° and HgCl2 [ 1,
7,9,12,17]. However the projected annual cost for an activated carbon adsorption system is an
issue, not only because of the high cost of the sorbent but also because of its poor
utilization/selectivity formercury. Carbon-to-mercury weight ratios of3,000:1 to 100,000:1 have
been projected for various levels of mercury control [18,19]. Previous investigations [9,12,20-
22] have shown the role of active surface sites in the adsorption of Hg° and HgCl2 by carbon- and
non-carbon-based sorbents. The potential importance of reactive surface sites underscores the
need for comprehensive research on the characterization of active sites and surface functional
groups (SFGs) on the surface of activated carbons. Establishing the nature of those SFGs that
are instrumental in mercury capture by activated carbons may result in low-cost technologies for
mercury control.
A clear understanding of mercury reactions on the surface of activated carbons is crucial
to the design of more effective sorbents. These reactions may determine both the adsorptive
capacity and regeneration efficiency of activated carbons and ultimately their economic viability.
A thorough knowledge of carbon surface chemistry may also lead to the development of more
mercury-specific activated carbons. Mercury-reactive SFGs on activated carbons may include
organic oxygenated species and/or functional groups containing inorganic elements such as
chlorine (CI) or sulfur (S). Organic SFGs are formed through oxygen chemisorption on the
surface of activated carbons, forming carbon-oxygen functional groups [23,24]. These sites can
be acidic or alkaline. Acidic oxygenated SFGs include carboxyl, lactone, hydroxyl, and carbonyl
functionalities, while alkaline properties are believed to arise from two kinds of active sites:
pyrone and chromene structures [23]. The role of oxygenated SFGS on the absorption of organic
compounds has been investigated [23, 24]. Not much is known of the role of oxygenated SFGs
on mercury species adsorption. In a study conducted by Liu et al. [25], the acidic oxygenated
SFGs content of two activated carbons (one virgin and one S-impregnated) were intentionally
increased using a procedure described by Tessmer et al. [23]. They observed no impact on Hg°
uptake by treated activated carbons, thus concluding that acidic SFGs are not instrumental in Hg°
2-Ghorishi

-------
capture. This is probably the only study of its kind and has not been confirmed by other
researchers. Moreover, the role of acidic SFGs on HgCl2 uptake, and the combined role of
alkaline SFGs on HgCl2 and Hg° uptake need to be investigated.
The nature of SFGs containing CI and S is not well understood. These SFGs may contain
covalently bound CI and S to the surface of carbon. The nature of these bonds is not known at
this time, although the role of CI and S in the capture of mercury species is well established.
Researchers have investigated mercury removal efficiencies of S-impregnated activated carbons
[1, 7, 25], These carbons have shown enhanced sorption capacity as compared to commercially
available virgin activated carbons, due to a higher content of active sulfur atoms. Questions
remain as to what constitutes an active S site and what is the nature of surface bonds between
active sulfur atoms and carbon. CI impregnation has shown the same effects. At temperatures
of 149-260°C, Teller and Quimby [26] found activated carbons impregnated with chloride salts
to have as much as 300 times greater Hg° removal capacity than virgin activated carbons.
However, the higher cost associated with using chemically impregnated activated carbons
requires their usage to be optimized. The future of understanding the nature of CI and S bonds
on the surface of activated carbons may lead to optimization of impregnation processes for
subsequent mercury capture.
Initially, the objective of this study was to determine the role of oxygenated alkaline
SFGs on the capture of Hg° and HgCl2. This approach was formulated based on our previous
observation [9,12] that indicated the importance of alkaline sites in the adsorption of HgCl2 by
a number of calcium-based sorbents. It was hypothesized that alkaline SFGs on the surface of
activated carbons play an important role in the capture of HgCl2. In this study, the acidic and
alkaline SFGs on the surface of two commercially available, virgin activated carbons were
identified and quantitated using the procedures described by Tessmer et al. [23], Subsequently,
using the same procedures, the acidic and alkaline SFGs were neutralized by various treatments
on activated carbons. The HgCl2 and Hg° uptake capabilities of treated activated carbons were
compared to those of untreated ones. The intent was to determine the role of each SFG on the
capture of Hg° and HgCI2. However, during the course of this study, it was found that SFGs
containing CI may play a much more important role than oxygenated SFGs. The focus of the
study was then shifted to understanding the role of Cl-containing sites. An analytical technique
was used to quantify surface-bound CI in the treated and the untreated activated carbons. An
attempt was made to correlate Hg° and HgCl, uptake with surface CI.
Determination of Oxygenated SFGs on Activated Carbons
Alkaline and acidic oxygenated SFGs were determined using the methods described by
Tessmer et al. [23]. Aqueous solutions of sodium bicarbonate (NaHC03, 0.05N), sodium
carbonate (NaC03,0.05N), sodium hydroxide (NaOH, 0.05 and 0.25N), and hydrogen chloride
(HC1, 0.05N) were prepared using deionized (DI) water. Each solution was individually added
to a predetermined amount (3 g) of activated carbon in a glass bottle. Each bottle was sealed and
allowed to equilibrate for 3 days with frequent shaking. Each set of bottles included a blank
solution without activated carbon. At the end of the equilibrium period, the activated carbon was
separated from the solution using a vacuum filter. The filtered activated carbon was washed
3-Ghorishi

-------
several times with fresh batches of DI water, dried, and stored in labeled bottles, then used for
Hg° and HgCU uptake studies. The filtrate was titrated using standardized acidic or alkaline
solutions. The amount of standardized alkaline or acidic solution consumed by each filtrate was
then related to the number of the alkaline or acidic SFGs on the surface of activated carbons [23],
SFGs on the surface of fresh activated carbons were quantitated in triplicate.
Acidic SFGs include carboxyl (-COOH), lactone (H2COCO), phenolic hydroxyl (-OH),
and carbonyl (-CO) groups. The amount of alkaline solution consumed by each acidic SFG was
calculated as the difference in the amount of standardized acidic solution required to titrate the
blank to a pH of 4.5 and the amount of acid required to titrate the filtrate to the same pH end
point. The carboxyl SFG concentration on the carbon surface was determined as the amount of
0.05N NaHC03 solution consumed by the carbon sample. The lactone SFG content was
calculated as the difference between the amounts of 0.05N Na2C03 and the amount of 0.05N
NaHC03 consumed by the carbon. The hydroxyl SFG content was determined as the difference
between the amount of 0.05N NaOH and the amount of 0.05N Na2C03 consumed by the carbon.
The carbonyl SFG content was found by subtracting the amount of 0.05N NaOH consumed by
the carbon from its 0.25N NaOH consumption. According to Tessmer et al. [23], alkaline SFGs
are believed to arise from the combination of two groups (pyrone and chromene) which cannot
be distinguished using this titration method [23], The total alkaline SFG content was evaluated
by equilibrating activated carbons with a 0.05N HC1 solution. The filtrate was titrated to a pH
of 11.5 using a 0.05N NaOH solution. The alkaline SFG content was calculated as the difference
between the amounts of NaOH required to reach the end point for the HC1 blank and for the
filtrate.
The above procedure was implemented on two activated carbons: DARCO FGD (FGD)
and DARCO PC-100 (PC-100) manufactured by Norit Americas, Inc. These two activated
carbons are described in detail elsewhere [12]. Briefly, FGD is a lignite-coal-based activated
carbon with a total surface area of about 500 m2/g, and an average pore diameter of 3.2 nm. PC-
100 is a bituminous-coal-based activated carbon with a total surface area of around 950 m2/g, and
an average pore diameter of 1.8 nm. Elemental compositions of these two activated carbons were
determined using the X-ray Fluorescence (XRF) technique. A significant difference in calcium
content was observed between these two carbons. The calcium content of FGD was 1.8 wt% and
that of PC-100 was 0.13 wt%. The CI and S levels in these two carbons were comparable at 0.23-
0.34 and 0.75-0.86 wt%, respectively. The SFG contents of treated and untreated FGD and PC-
100, expressed in |ieq/g, are summarized in Table 1.
4-Ghorishi

-------
Table 1. Acidic and alkaline SFGs in treated and untreated activated carbons (in (ieq/g)
SFGs
Untreated
Alkaline Treated
Acid Treated
Thermally Treated
FGD
PC-100
FGD
PC-100
FGD
PC-100
FGD
PC-100
Carboxyl
250±24
157±6
95
48
NA
NA
NA
NA
Lactone
217±11
94±6
<5
210
NA
NA ¦
NA
NA
Hydroxyl
182±24
285±11
<5
60
NA
NA
NA
NA
Carbonyl
2692±27
1817±18
767
<5
NA
NA
NA
NA
Alkaline*
1344±8
363±11
NA
NA
12.5
513
1670
828
NA: not app
icable; * total (pyrone and chromene)
The intention of acid treating (washing with 0.05N HC1) the activated carbons was to
neutralize their alkaline SFGs, and the intention of alkaline treating (with their respective
solution) was to neutralize specific acidic SFGs. A thermal treatment method was also used on
these activated carbons. The objective of this treatment was to increase alkaline SFGs and thus
influence the capture of HgCl2 according to the initial hypothesis. This thermal treatment process
is described in detail by Papirer et al. [27]. They showed that the amount of oxygen which
chemisorbs at 100°C on a pretreated (at 800°C) carbon corresponds to alkaline-like (pyrone)
groups which are formed after re-exposure to air. The pyrone group is generated by combination
of a heat-resistant oxygen group, which stays on the carbon surface even at 800 °C, and another
oxygen group created during air re-exposure. Based on this hypothesis, a thermal treatment
procedure was implemented on the activated carbons: activated carbons were heated to 800°C
in an inert atmosphere (helium) and maintained at this temperature for 1 hour. They were
subsequently cooled to room temperature in an inert atmosphere. Activated carbons were then
re-exposed to air at 100°C overnight.
As indicated in Table 1, alkaline SFGs increased in both thermally treated activated
carbons. This increase was more pronounced for PC-100. Untreated, as-received FGD had
higher concentrations of both acidic (except for hydroxyl) and alkaline SFGs as compared to PC-
100. The SFG contents of PC-100 are comparable to those reported by Tessmer et al. [23] for
a bituminous-coal-based activated carbon. As expected, alkaline-treatment of activated carbons
reduced the acidic sites in FGD and PC-100. The only exception was the lactone group in PC-
100 which was apparently increased by treatment with 0.05N solution of Na2C03. As expected,
acid treatment decreased the number of alkaline SFGs in FGD. However, acid treatment
increased the number of alkaline SFGs in PC-100. These anomalies cannot be explained.
Experimental
Apparatus
Detailed descriptions of the experimental setup, a fixed-bed reactor system (Figure 1),
have been presented previously [12], Briefly, pure Hg° liquid in a permeation tube (VICI
Metronics) was the source of Hg° vapor, and pure solid HgCl2 crystals in three size D diffusion
5-Ghorishi

-------
vials (VICI Metronics) was the source of HgCl2 vapor. The concentration of HgCl2 or Hg° in the
gas stream was controlled by adjusting the temperature of the permeation tube's or diffusion
vial's water bath. HgCl2 or Hg° vapor generated was delivered to the fixed bed by an N2 stream
at a constant total system flow rate of 300 cm3/min (at standard temperature and pressure, STP).
The activated carbon (treated or untreated) to be studied (0.1 g) was mixed to a total of 2 g with
sand and placed in the down-flow vertical reactor between two quartz wool plugs and maintained
at the desired bed temperature by a temperature controller. A furnace, kept at 920° C, was
downstream of the reactor to convert any oxidized mercury (Hg++, as in HgCl2) to Hg°. The
presence of the furnace enabled detection of oxidized, non-adsorbed Hg° or non-adsorbed HgCl2
as gas-phase Hg° by the on-line ultraviolet (UV) Hg° analyzer. Thus, continuous Hg° or HgCl2
capture data by the fixed bed of sorbent could be acquired.
Methodology
Initial scoping tests were performed to determine the optimum conditions for HgCl2
uptake, so that a valid comparison could be made between treated and untreated activated
carbons. The optimum conditions were determined to be a bed temperature of 140°C, inlet
HgCl2 concentration of 38 ppbv in N2, and an exposure time of 2 hours. In the case of Hg°,
scoping tests determined that an optimum comparison between these sorbents could be made at
a bed temperature of 100°C, inlet Hg° concentration of 40 ppbv in N2, and an exposure time of
4 hours. During the HgCl2 or Hg° contact period, the exit concentration of mercury was
continuously monitored. The instantaneous removal of HgCl2 or Hg° at any time (Q,) was
obtained as given by:
Q, = [(Hgin-Hg0Ut)/Hg in] x 100
HgCl2 or Hg° uptake was determined by integrating and evaluating the area under the removal
curve. HgCl2 or Hg° uptake was defined as a cumulative removal of HgCl2 or Hg° up to time t
(4 hours for Hg° and 2 hours for HgCl2) and was expressed as a weight ratio of mercury species
uptake to active sorbent (ng Hg/g sorbent). Selected experiments conducted during this test
program were run in duplicate and indicated a range of ±5 percent about the mean in the
experimental results. Blank tests indicated the empty reactor capture of Hg° and HgCl2 to be 0.02
and <1.0 |a.g, respectively. These values are significantly lower than uptakes exhibited by the
sorbents.
Analytical
An analytical technique was employed to accurately measure the amount of surface CI
in treated and untreated activated carbons. Scanning Electron Microscopy and Energy Dispersive
X-Ray Spectroscopy (SEM/EDXS) analysis was performed with a Jeol JSM-6400 interfaced
with a SUN SPARC workstation 5. The EDXS detector was a sealed window PRISM detector
mounted horizontally. Samples were mounted onto carbon stubs by double-sided tape. The
sample was placed in excess on the tape and pressed down lightly with a clean spatula; the
excess sample was removed by gently tapping the sample stub. Three separate areas for each
sample were magnified to 1000X, and an image and spectra were obtained (only one area was
analyzed for the thermally treated FGD and the FGD treated with 0.05 N and 0.25 N NaOH).
Spectra were acquired in PRISM quantitative mode at a takeoff angle of 31 °, a working distance
of 16 mm, and an accelerating voltage of approximately 15 KeV. Quantitative analysis was
6-Ghorishi

-------
performed using the UNIX-based IMIX software. This analysis was performed in the standard-
less mode using the "Z-A-F" calculation [28]. ZAF comes from the "atomic number" (Z),
"absorption" (A), and, fluorescence" (F). Moreover, the spectra were normalized to account for
the slight differences in accelerating voltage between the spectra. Each element present in the
samples was calculated in parts per million on a weight basis. Due to the low accelerating
voltage, it can be assumed that these elements are on the surface of the carbons within a depth
of 1 to 1.5 |j.m (see the nomograph on page 46 of reference 28). Further studies and computer
modeling should result in more accurate estimates of the depth of electron penetration and the
depth of excitation of the X-rays for the individual elements.
The existence of mercury bonds on the surface of different sorbents was probed using an
analytical technique known as X-Ray Absorption Fine Structure (XAFS) spectroscopy [29].
Huggins et al. [29] have shown that different activated carbons (sulfur activated, iodine
activated, and lignite-derived activated carbon; FGD) sorbed mercury from an experimental flue
gas by three different mechanisms (formation of Hg-S, Hg-I, and Hg-Cl bonds, respectively).
XAFS analysis is described in detail elsewhere [29].
Results and Discussion
The initial hypothesis, and the basis for this study, was that alkaline SFGs on the surface
of activated carbons are the active sites for the capture of acidic HgCl2. HgCl2 uptake by
untreated (as-received) and treated activated carbons was measured in the fixed bed. The uptake
results after a 2-hour exposure (expressed in |j.g HgCl2/g sorbent) are presented in Figure 2. An
unexpected behavior was observed. Acid treating the activated carbons (see 0.05N HC1)
promoted the HgCl2 capture by both activated carbons as compared to the untreated baseline.
Note that the uptake exhibited by HCl-treated activated carbons corresponds to complete
removal of inlet HgCl2 during the exposure period (2 hours). As mentioned earlier, HC1
treatment neutralized alkaline SFGs in FGD and to a slight extent increased them in PC-100. It
appears that, in contrast to the initial hypothesis, alkaline SFGs do not provide active sites for
the capture of HgCl2. Noted that the treatment of activated carbons did not change the physical
properties of these sorbents; e.g., total surface area and pore size distribution. The increase in
HgCl2 uptake exhibited by the HCl-treated activated carbons is believed to be due to an increase
in CI content of these treated carbons. This treatment process may have acted as a Cl-
impregnation process which is a proven method for enhancing mercury uptake capabilities of
activated carbons [26], The increase in CI content of HCl-treated activated carbons is discussed
later. As an additional experiment, samples of HCl-treated activated carbons were exhaustively
washed with DI water (total wash time of 96 hours with four batches of fresh DI water) to
remove any loosely bound surface CI. The exhaustively washed HCl-treated activated carbons
still exhibited a superior HgCl2 removal (see Figure 2), indicating that the active CI sites created
by HC1 treatment is of a chemisorbed nature.
Thermal treatment of activated carbons was also used to further show that alkaline
oxygenated SFGs do not play a role in the capture of HgCl2. The activated carbons were
thermally treated to increase their alkaline SFGs (see Table 1). The HgCl2 uptake by thermally
treated activated carbons was compared to the untreated activated carbons (Figure 2). The
7-Ghorishi

-------
increase in alkaline SFGs did not result in an increase in HgCl2 uptake; rather, thermal treatment
decreased the HgCU uptake. This is believed to be due to the loss of surface CI sites during the
thermal treatment process (see Table 2 and its discussion).
The effect of acidic oxygenated SFGs on HgCl2 uptake was studied by comparing
alkaline-treated with untreated activated carbons. The activated carbons were treated with 0.05N
NaHC03 to neutralize carboxyl groups; 0.05N Na2C03 to neutralize carboxyl and lactone groups;
0.05N NaOH to neutralize carboxyl, lactone, and hydroxyl SFGs; and 0.25N NaOH to neutralize
all acidic sites. The HgCl, uptake by these alkaline-treated activated carbons is shown in Figure
2. It was demonstrated that acidic SFGs play no role in capturing HgCl2. The decline in HgCl2
uptake by these treated activated carbons may also be due to the loss of surface CI sites.
Not much is known about the nature of active sites for Hg° capture. Hg° uptake by the
acid-, alkaline-, and thermally treated activated carbons was evaluated and compared to that of
untreated ones. Results are shown in Figure 3. Treated activated carbons exhibited similar
behavior in capturing Hg° as they did in capturing HgCl2. HCl-treated activated carbons showed
complete removal of Hg° during the 4 hours of exposure. HC1 treatment significantly improved
Hg°, as well as HgCl2, uptake. This observation can once again be attributed to an increase in
surface CI content of HCl-treated activated carbons. The exhaustively washed HCl-treated
activated carbons still exhibited a superior Hg° (as well as HgCl2) removal.
Results showed that acidic and alkaline oxygenated SFGs are not instrumental in the
capture of both mercury species (Hg° and HgCl2). Neutralization or generation of these SFGs on
activated carbons had little or no effect on mercury species capture. However, it seems that Cl-
containing SFGs play an important role in Hg° and HgCl2 capture. Chemical impregnation of
virgin activated carbons using chlorine compounds has been shown to be effective in increasing
Hg° and HgCl2 uptake [26,30], Quimby [30] showed that treatment of carbons with HC1 solution
(1:1 weight ratio) can improve mercury adsorption capacity. He observed significant
improvement using copper chloride solution as the impregnating agent. It has also been shown
that a flue gas stream containing HC1 vapor can serve to impregnate activated carbons in-situ and
improve mercury removal capacity [22, 30]. This in-situ impregnation is suspected to play a
major role in mercury vapor emission control from municipal, hazardous, and hospital waste
incinerators. Granite et al. [31] also showed that HCl-treated activated carbons exhibit a large
capacity (up to 4 mg/g) for Hg° and HgCl2. They hypothesized that mercury species will
primarily form the tetrachloromercury complex, HgCl42", on the surface of HCl-treated activated
carbon. The presence of this surface compound has not been verified.
The HCl-treated activated carbons in this study exhibited similar behavior. Their
improved Hg° and HgCl2 removal capabilities may be due to an increase in Cl-containing SFGs.
Both Hg° and HgCl2 in the flue gas are attracted to these sites. As mentioned, and to test this
hypothesis, an analytical SEM/EDXS method was employed to measure surface CI content of
treated and untreated activated carbons (Table 2). A sample of SEM/EDXS spectra for HCl-
treated and untreated activated carbons is shown in Figures 4 (for FGD) and 5 (for PC-100).
SEM/EDXS analysis showed that the surface of FGD and PC-100 carbons is predominately
8-Ghorishi

-------
composed of carbon (C) and oxygen (O), with trace levels of calcium (Ca), aluminum (Al),
silicon (Si), iron (Fe), magnesium (Mg), and sulfur (S). The surface of FGD carbon contains
significantly more Ca than the surface of PC-100 carbon (Table 2).
The results of the SEM/EDXS analysis showed that the level of CI on the surface of the
HCl-treated carbons is significantly higher than on the untreated, alkaline-, and thermally treated
carbons (Table 2, and Figures 4 and 5). The exhaustively washed HCl-treated carbons also
showed higher levels of CI than the untreated carbons, but a lower level of CI compared to the
non-exhaustively washed HCl-treated carbons. It can be hypothesized that the exhaustive wash
of carbons removed all the loosely bound surface CI and established the amount of those CI
atoms that are strongly bound to the surface. The exhaustively washed, HCl-treated carbons and
non-exhaustively washed HCl-treated carbons removed the same amount of HgCl2 and Hg°
(Figures 2 and 3), suggesting that the strongly bound surface CI provides active sites for the
capture of mercury species.
XAFS analysis was also performed on HCl-treated and untreated activated carbons to
determine the nature of mercury bonds on the surface of activated carbons. Possible formation
of Hg-Cl type bonds were detected on the surface of HCl-treated and untreated FGD when
exposed to Hg° and HgCl2 in the flue gas. However, in the case of PC-100 (both treated and
untreated), the presence of Hg-S type bonds may have been detected. XRF analysis on PC-100
and FGD indicated identical sulfur content (0.75-0.85 wt%). Sulfur in PC-100 may be more
active than that in FGD. Note that XAFS analysis is a qualitative method, and the interpretation
of the results should be regarded with caution. Other analytical techniques (such as infrared
spectroscopy) need to be developed to detect mercury surface species.
Table 2. Concentration of CI and Ca on treated and untreated activated carbons
Treatment
CI (ppmw)
Ca (ppmw)
FGD
PC-100
FGD
PC-100
As-received (untreated)
NC
150±71
10267±2272
267±58
0.05N HC1
833±153
1900±608
1033±351
250±71
0.05N HC1 (exhaustively washed)
167±58
633±58
800±173
100
0.05N NaHC03
NC
NC
14000±2921
367±58
0.05N Na2C03
NC
NC
13700±2193
533±208
0.05N NaOH
100
NC
14200
400±100
0.25N NaOH
100
NC
15500
333±58
Thermally treated
NC
NC
13600
367±115
sfC: Not Calculated, below system/standard quantitation limit (100 ppmw).
9-Ghorishi

-------
Summary
The content of oxygenated acidic and alkaline surface functional groups (SFGs) on the
surface of two activated carbons was manipulated to investigate their role in Hg° and HgCl2
capture. Acidic SFGs on the surface of activated carbons were neutralized by a variety of
alkaline washes. The alkaline-treated activated carbon showed no enhancement in Hg° and HgCl2
capture, thus indicating that acidic SFGs play no role in capturing mercury species. The alkaline
SFGs contents were increased by a thermal treatment process. The thermally treated activated
carbons did not exhibit any improvement with regard to their Hg° and HgCl2 capture capabilities
as compared to the untreated ones. The activated carbons were then treated with a very dilute
HC1 solution to decrease their alkaline SFGs content. The HCl-treated activated carbon showed
a very significant improvement in its Hg° and HgCl2 capture capabilities. This observation was
contrary to the initial hypothesis that alkaline sites are needed to capture acidic HgCl2 from the
flue gas. It was then hypothesize that HC1 treatment increases the number of active surface
chlorine sites, which subsequently enhance Hg° and HgCl2 capture. An analytical technique ,
Energy Dispersive X-ray Spectroscopy (EDXS), was used to quantify surface CI sites. A strong
correlation between the increased amount of surface CI and Hg°/HgCl2 uptake enhancement was
observed. The role of SFGs containing CI atoms in providing Hg°/HgCl2 active sites was
established. Future investigation using SEM/EDXS and Fourier Transform Infrared (FTIR) will
focus on understanding the nature of CI bonds on the surface of carbon, so that more effective
mercury species sorbents can be manufactured.
Acknowledgments
The authors would like to acknowledge the advisory role of Wojciech Jozewicz (of
ARCADIS) and experimental assistance from Wojciech Kozlowski (formerly with ARCADIS)
during these studies.
References
1.	Krishnan, S. V., Gullett, B. K. & Jozewicz, W. (1994) Sorption of Elemental Mercury by
Activated Carbons. Environmental Science & Technology, 28, p 1506.
2.	Krishnan, S. V., Gullett, B. K. & Jozewicz, W. (1995) Mercury Control by Injection of
Activated Carbon and Calcium-Based Sorbents. Presented at the Solid Waste Management
Thermal Treatment & Waste-To-Energy Technologies Conference, USEPA/ORD/AEERL and
A&WMA, Washington, DC.
3.	Krishnan, S. V., Bakhteyar, H. & Sedman, C. B. (1996) Mercury Sorption Mechanisms and
Control by Calcium-Based Sorbents. Presented at the 89th Air & Waste Management
Association Annual Meeting, Nashville, TN.
4.	Lancia, A., Musmarra, D., Pepe, F. & Volpicelli, G. (1997) Control of Emissions of Metallic
Mercury from MSW Incineration by Adsorption on Sorbalit1M. The 1997 International
Conference on Incineration and Thermal Treatment Technologies, Oakland, CA.
10-Ghorishi

-------
5.	Metzger, M. & Braun, H. (1987) In-Situ Mercury Speciation in Flue Gas by Liquid and Solid
Sorption Systems. Chemosphere, Vol. 16, No. 4, pp 821-832.
6.	Otani, Y., Emi, H., Kanaoka, C., Uchijima, I. & Nishino, H. (1988) Removal of Mercury
Vapor from Air with Sulfur-Impregnated Adsorbents. Environmental Science & Technology, 22,
pp 708-711.
7.	Vidic, R. D. & McLaughlin, J. B. (1996) Uptake of Elemental Mercury Vapors by Activated
Carbons. Journal of Air & Waste Management Association, 46, pp 241-250.
8.	Lancia, A., Musmarra, D., Pepe, F. & Volpicelli, G. (1993) Adsorption of Mercuric Chloride
Vapors from Incinerator Flue Gases on Calcium Hydroxide Particles. Combustion Science and
Technology, 93, p 277.
9.	Gullett, B.K., Ghorishi, S.B., Keeney, R.M. & Huggins, F.E. (2000) Mercuric Chloride
Capture by Alkaline Sorbents. Presented at the 93rd Air & Waste Management Association
Annual Meeting, Salt Lake City, UT.
10.	Devito, M. S., Tunati, P. R., Carlson, R. J. & Bloom, N. (1993) Sampling and Analysis of
Mercury in Combustion Flue Gas. Presented at the EPRI's Second International Conference on
Managing Hazardous Waste Air Pollutants, Washington, DC.
11.	Laudal, D. L., Heidt, M. K., Brown, T. D., Nott, B. R. & Prestbo, E. P. (1996) Mercury
Speciation: A Comparison Between EPA Method 29 and Other Sampling Methods. Presented
at the 89th Air & Waste Management Association Annual Meeting, Nashville, TN.
12.	Ghorishi, S.B. & Gullett, B.K. (1998) Sorption of Mercury Species by Activated Carbons
and Calcium-Based Sorbents: Effect of Temperature, Mercury Concentration and Acid Gases.
Waste Management & Research, 16(6), 582-593.
13.	Olson, E.S., Sharma, R.K., Miller, S.J. & Dunham, G.E. (1999) Identification of the
Breakthrough Oxidized Mercury Species from Sorbents in Flue Gas. In the Proceedings of the
Mercury in the Environment Speciality Conference, Minneapolis, MN.
14.	Livengood, C. D. & Mendelsohn, M. H. (1997) Enhanced Control of Mercury Emissions
Through Modified Speciation. Paper 97-WA72A.01 Presented at Air & Waste Management
Association's 90th Annual Meeting & Exhibition, Toronto, Ontario, Canada.
15.	Senior, C. L., Bool, L. E., Huffman, G. P., Huggins, F. E., Shah, N., Sarofim, A., Olmez, I.
& Zeng, T. (1997) A Fundamental Study of Mercury Partitioning in Coal-Fired Power Plant Flue
Gas. Paper 97-WA72B.08 Presented at Air & Waste Management Association's 90th Annual
Meeting & Exhibition, Toronto, Ontario, Canada.
11-Ghorishi

-------
16.	Ghorishi, S.B., Lee, C.W. & Kilgroe, J.D. (1999) Mercury Speciation in Combustion
Systems: Studies with Simulated Flue Gases and Model Fly Ashes. Paper 99-651 presented at
the 92nd Air & Waste Management Association Annual Meeting & Exhibition, St. Louis, MO.
17.	Lancia, A., Musmarra, D., Pepe, F. & Volpicelli, G. (1996) Adsorption of Metallic Mercury
on Activated Carbon. Twenty-Sixth Symposium (International) on Combustion, The
Combustion Institute, Pittsburgh, PA, p. 2439.
18.	Keating, M.H. & Mahaffey, K.R. (1997) Mercury Study Report to Congress, Vols. I-VIII,
EPA-452/R-97-003; U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards and Office of Research and Development; Washington, DC, December 1997, also
available at .
19.	Brown, T.D.; Smith, D.N.; Hargis, R.A. & O'Dowd, W.J. (1999) Mercury Measurement and
Its Control: What We Know, Have Learned, and Need to Further Investigate. 1999 Critical
Review, Presented at the Air & Waste Management Association 92nd Annual Meeting and
Exhibition, St. Louis, MO, June 23.
20.	Ghorishi, S.B. & Sedman, C.B. (1998) Low Concentration Mercury Sorption Mechanisms
and Control by Calcium-Based Sorbents: Application in Coal-Fired Processes. Journal of Air
& Waste Management Association, 48, pp. 1191-1198.
21.	Ghorishi, S.B., Gullett, B.K., Jozewicz, W. & Kozlowski, W. (1999) Role of HC1 in
Adsorption of Elemental Mercury Vapor by Calcium-Based Sorbents. In the Proceedings of the
2nd EPRI-DOE-EPA Combined Utility Air Pollutant Control Symposium, Atlanta, GA.
22.	Ghorishi, S.B. & Gullett, B.K. (1997) Fixed-Bed Control of Mercury; Role of Acid Gases
and a Comparison Between Carbon-Based, Calcium-Based, and Coal Fly Ash Sorbents. In the
Proceedings of the 1st EPRI-DOE-EPA Combined Utility Air Pollutant Control Symposium,
Washington, DC.
23.	Tessmer, C.H., Vidic, R.K. & Uranowski, L.J. (1997) Impact of Oxygen-Containing Surface
Functional Groups on Activated Carbons Adsorption of Phenols. Environmental Science and
Technology, Vol. 13, No. 7, p 1872.
24.	Coughlin, R.W. & Ezra, F.S. (1968) Role of Surface Acidity in the Adsorption of Organic
Pollutants on the Surface of Carbon. Environmental Science and Technology, Vol. 2, No. 4, pp
291-297.
25.	Liu, W., Vidic, R.K. & Brown, T.D. (2000) Impact of Flue Gas Conditions on Mercury
Uptake by Sulfur-Impregnated Activated Carbon. Environmental Science and Technology, Vol.
34, No.l, pp 154-159.
12-Ghorishi

-------
26.	Teller, A.J. & Quimby, J.M. (1991) Mercury Removal from Incineration Flue Gas. Paper
91-35.5 presented at the 84th Air and Waste Management Association's Annual Meeting and
Exhibition, Vancouver, BC.
27.	Papirer, E., Li, S. & Donnet, J.B. (1987) Contribution to the Study of Basic Surface Groups
on Carbon. Carbon, Vol. 25, No. 2, pp 243-247.
28.	Russ, John C. (1991) Fundamental of Energy Dispersive X-ray Analysis. Monographs in
Materials, Butterworths, London, England.
29.	Huggins, F.E., Huffman, G.P., Dunham, G.E. & Senior, C.L. (1999) XAFS Examination of
Mercury Sorption on Three Activated Carbons, Energy & Fuels, 13, p. 114.
30.	Quimby, J.M. (1993) Mercury Emission Control from Combustion Systems, Paper 93-MP-
5.03, Presented at Air and Waste Management Association's Annual Meeting, Denver, CO.
31.	Granite, E.J., Pennline, H.W. & Hargis, R.A. (2000) Novel Sorbents for Mercury Removal
from Flue Gas. Ind. Eng. Chem. Res., Vol. 39, No. 4, pp 1020-1029.
Mercury Generation
System
N,
Carbon Trap
-o
3-Way Valve 3-Way Valve
Water Bath
Carbon Trap
Data Acquisition
<0
CO
OJ
Q_
I
CO
-O
H
Rotameter
Elemental
Mercury
Analyzer
3-Way Valve
-o
Furnace
(O
a>
a:
ca
•6
a>
Figure 1. Schematic of the bench-scale, fixed-bed adsorption system.
13-Ghorishi

-------
treatment agent
* exhaustively washed with Dl water
Figure 2. HgCl2 uptake by treated and untreated activated carbons, bed temperature of 140 °C,
inlet HgCl2 concentration of 38 ppbv, O.lg sorbent/2 g sand, and exposure period of 2 hours.
treatment agent
* exhaustively washed with Dl water
Figure 3. Hg° uptake by treated and untreated activated carbons, bed temperature of 100° C, inlet
HgO concentration of 40 ppbv, 0.1 g sorbent/2 g sand, and exposure period of 4 hours.
14-Ghorishi

-------
Ca
Intensity
Untreated
HCl-treated
ci
2.0
2.5
3. 0
3.5
4.0
4.5
keV
Figure 4. EDXS spectra of HCl-treated and untreated activated carbon FGD.
Intensity
HCl-treated
ci
Untreated
.._ * *
2.0
2.5
3.0
3.5
4.0
4.5
keV
Figure 5. EDXS spectra of HCl-treated and untreated activated carbon PC-100.
15-Ghorishi

-------
MRMRT -RTP-P-R49 TECHNICAL REPORT DATA
in ruvi ru-^ rti *r- sr ot& (Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA/600/A-00/103
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Role of Surface Functional Groups in the Capture of
Elemental Mercury and Mercuric Chloride by
Activated Carbons
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
S.B. Ghorishi and R. M. Reeney (ARCADIS), and
B. R. Gullett (EPA)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
ARCADIS Geraghty and Miller, Inc.
4915 Prospectus Erive
Durham, NC 27713
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-C-99-201, WAs 0-014
and 1-022
12. SPONSORING AGENCY NAME AND ADORESS
EPA, Office of Research and Development
Air Pollution Prevention and Control Division
Research Triangle Park, NC 27711
i™i^rcf^RaTpAeNrD;Pi5?§§-0W
14. SPONSORING AGENCY CODE
EPA/600/13
15.supplementary notes APPCD project officer is Brian K. Gullett, Mail Drop bfc>, yiy/
541-1534. For presentation at Conference on Air Quality II, Tysons Corner, McLean,
VA, 9/19-21/00.
16.abstractpaper discusses using a laboratory-scale, fixed-bed apparatus to study
the role of surface functional groups (SFGs) in the capture of mercuric chloride
(HgC12) and elemental mercury (Hgo) in nitrogen (N2) prior to flue gas atmosphere
studies. The study focused on two activated carbons (FGD and PC-100, Norit Ameri-
cas, Inc.) with different physical and chemical characteristics. SFGs may be acidic
or alkaline. Acidic SFGs include carboxyl, lactone, hydroxyl, and carbonyl function-
alities, while alkaline properties are believed to arise from two types of active
sites, pyrone and chromene structures. The acidic and alkaline SFGs of as-received
FGD and PC-100 were determined using surface titration. An attempt was made to
correlate the number density of these SFGs to their ability to sorb Hgo and HgC12.
The activated carbons were then treated with acid and alkaline washes to neutralize
their alkaline and acidic SFGs, respectively. The mercury capture capabilities of
these treated carbons were then compared to those of untreated as-received samples.
The initial hypothesis was that the number of of alkaline SFGs should correlate with
HgC12 capture. This was proved not to be the case; rather, the concentration of sur-
face chlorine (CI) sites was related to the HgC12 and Hgo capture. Acidic and alka-
line SFGs play no role in the adsorption of Hgo and HgC12 by activated carbons.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Mercury (Metal)
Activated Carbon
Nitrogen
Flue Gases
Pollution Control
Stationary Sources
Mercuric Chloride
Surface Functional
Groups (SFGs)
13 B
07B
11G
2 IB
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21j.^iO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9*73)

-------