Preparation of gold- and chlorine-impregnated bead-type activated carbon for a mercury sorbent trap

Chemosphere 165 (2016) 470e477

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Preparation of gold- and chlorine-impregnated bead-type activated carbon for a mercury sorbent trap Young Cheol Song, Tai Gyu Lee* Department of Chemical and Biomolecular Engineering, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, South Korea

h i g h l i g h t s
Hg sorbent trap is prepared using impregnated bead-type activated carbon (BAC).
BAC is impregnated with chlorine and/or gold (BACC, BACAu, and BACClAu).
Hg spiking efficiency of BACClAu is 23% higher than that of virgin BAC.
Hg adsorption efficiency higher than 95% is obtained by BACClAu.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 July 2016 Received in revised form 3 September 2016 Accepted 6 September 2016 Available online 30 September 2016

This study aimed to develop a mercury (Hg) adsorption trap, which can be used to measure the concentration of elemental Hg in emissions from a Hg discharge facility, and evaluate its adsorption efficiency. The Hg spiking efficiency was compared by impregnating metallic and halogen materials that have high affinity for Hg into activated carbon (AC) to determine an accurate spiking method for Hg on AC. The Hg spiking efficiency was compared according to the type and content of the impregnated substances. AC impregnated with Cl and Au had a 15e20% higher Hg spiking efficiency compared to virgin AC. For Au impregnation at weight ratios of 0e20 wt% of adsorbent, spiking efficiencies of over 97% were observed under certain conditions. The Hg adsorption properties of the above adsorbent were determined experimentally, and the results were used to test the adsorption performance of Hg adsorption traps. © 2016 Elsevier Ltd. All rights reserved.

Handling Editor: Shane Snyder Keywords: Mercury Gold nanoparticle Adsorbent Activated carbon Impregnation

1. Introduction Stringent regulations and continuous and accurate monitoring of atmospheric Hg emissions are now mandated as a result of the Minamata Convention on Mercury (Hg) (UNEP, 2013). Despite the complexity of sample collection and pretreatment prior to measurement, wet processes for the examination of Hg emissions in air such as the Ontario Hydro Method (ASTM D6784, 2008) have been used for some time. A wet process involves the constant inhalation of gas in the stack by sampling equipment, followed by trapping in an absorption solution. The strong acids used as the absorption solution, such as KMnO4eH2SO4, should be contained in an impinger. However, wet processes could lead to

* Corresponding author. E-mail address: [email protected] (T.G. Lee). http://dx.doi.org/10.1016/j.chemosphere.2016.09.021 0045-6535/© 2016 Elsevier Ltd. All rights reserved.

large errors in the measured value because of the complex sample collection and pretreatment processes. Dry processes have been developed to overcome the shortcomings of wet processes. The most widely used dry process is US EPA Method 30 B (US EPA, 2007). Technical difficulties in obtaining both homogeneity and stability of the Hg-spiked reference section of the sorbent trap have prevented this method from being commercialized until recently (Zhang et al., 2016). Continuous emission monitoring (CEM) equipment was recently introduced and is advantageous in that it can determine the concentration of Hg discharged in real time. However, CEM is disadvantageous in terms of its high initial cost. US EPA Method 30 B uses an adsorption trap, which is easy to install and sample as well as inexpensive. However, because the adsorption trap cannot measure particulate Hg, it can be only used under low particulate concentrations. Nevertheless, the sorbent trap can be used by itself or to verify CEM readings and is comparatively easier than a typical wet

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

process. To properly analyze the concentration of Hg in emitted gas, the amount of Hg spiked in the adsorbent (activated charcoal) should be accurate when producing the adsorption trap. Although existing adsorption traps have an efficiency of over 90% for highly concentrated spiked Hg, these traps lead to high error rates at low spiking concentrations. In this study, a halogen and transition metal (Kobiela et al., 2003), which have high binding affinity for Hg, were impregnated to compensate for the abovementioned disadvantage (Chingombe et al., 2005). Impregnating chlorine (Cl) as the deactivated catalyst (Zhang et al., 2016) activates the activated carbon (AC). Furthermore, the halogen, i.e., Cl, has strong binding affinity. Because gold (Au) is the most stable transition metal and has strong binding affinity for Hg, its binding with Hg is not broken at high temperatures (700  C) (Kobiela et al., 2003). 2. Experimental 2.1. Materials A sorbent trap involves an absorbent that is situated in a narrow pipe. To maintain the uniformity of the shape and size of the trap, bead-type activated carbon (BAC) was selected instead of powdered AC, which often causes a pressure drop. Aqua regia was used for Cl impregnation. Halogens restore disabled AC because they bond strongly to Hg. The reagents used in the experiment are detailed below (Granite et al., 1998). The fact that bonds between Hg and Au, as well as impregnated Au, are not broken at temperatures below 700  C indicates high bond strength (Portzer et al., 2004). A Au standard solution was prepared to impregnate Au. Overly impregnating Au can disrupt Hg absorption. A 1 ppm Au standard solution was prepared by diluting a Au stock solution (Kanto Chemical Co. Inc., Tokyo, Japan). 2.2. AC impregnation To increase the efficiency of Hg adsorption to BAC (Kanto Chemical Co. Inc., Tokyo, Japan), BAC was impregnated with 5% aqua regia and a 5 wt% Au standard solution; 5% aqua regia consisted of an HCl solution mixed with a solution of HNO3 at a ratio of 3 to 1, which was subsequently diluted. After soaking the BAC in 5% aqua regia, the sample was heated at 90  C for 30 min using a heating block (SCP Science, Quebec, Canada). After heating, the sample was baked at 90  C in an oven for 48 h. After adding 40 mL of deionized (DI) water to the sample after the baking was completed, a 5 wt% Au standard solution of BAC was bound to the sample. After stirring for 2 h using a stirrer at 350 rpm, the solution was filtered using a 0.45-mm membrane filter. Afterward, the sample was stored in a glass bottle after drying for 48 h at 110  C. BAC was impregnated using 1) 5% aqua regia, 2) a Au solution, and 3) both 5% aqua regia and a Au solution and named BACCl, BACAu, and BACClAu, respectively. 2.3. Hg spiking The Hg concentration in the spiked section of the sorbent trap that the experimenter was aware of must have been absorbed (US EPA, 2007). Therefore, the Hg vapor adsorption experiment is essential to ensure the accuracy of the data. Hg-spiked BACAu was produced as follows. After the insertion of BACAu was complete, the impregnation of Cl and Au in the Hg standard solution was conducted for 4 h using a stirrer at 350 rpm. After stirring, the samples were baked for 48 h at 90  C. The spiked concentration of Hg varied depending on the concentration and flow rate of Hg gas. The Hg concentration to be

471

spiked can be calculated using Eq. (1):

Cc ¼

Csol  Vsol Wc

(1)

where Cc is the Hg concentration (ng g1) in the Hg-spiked BACAu, Csol is the Hg concentration (mg L1) in the Hg standard solution, Vsol is the volume of the Hg standard solution, and Wc is the weight (g) of BACAu. 2.4. Hg adsorption Each experiment was conducted to analyze the adsorption properties of the different impregnated materials. The adsorption rate and adsorbed amount were confirmed according to the variations in the contact time and initial concentration. The adsorbed concentration and removal concentration were identified using experiments that were dependent on the contact time. The initial Hg concentration was 1000 ng mL1. The maximum adsorption capacity was confirmed based on the variation of the initial concentration of Au. The Au contents were 0, 5, 10, and 20 wt%; the Hg concentrations were 50, 100, 200, 300, 400, 500, 600, and 700 mg L1. Each sample was analyzed after a 12 h reaction. 2.5. Performance evaluation of the sorbent trap Hg adsorption in the gas phase was conducted to estimate the capacity of the sorbent trap. The Hg concentration was set using the standards for thermoelectric power plants and the cement industry (Granite et al., 2000). The inhalation flow rate and adsorption time mez-Gime nez were set as 0.4e1 lpm and 60 min, respectively (Go et al., 2015). A Hg0 calibration unit (Model 3310, Tekran Instruments Corp., Toronto, Canada) was used to generate a flue gas with a highly precise Hg0 concentration. After passing through the reactor, the adsorption efficiency was analyzed using a continuous, real-time, on-line, gas-phase Hg0 analyzer (VM-3000, Mercury Instruments, Karlsfeld, Germany). A cold vapor atomic absorption spectrometry (CVAAS) Hg analyzer (RA-915þ/RP-91, Lumex Ltd., St. Petersburg, Russia) was used to measure the Hg concentration on the AC in the sorbent trap. Fig. 1 shows the adsorption efficiency estimation and mimetic diagram. 3. Results and discussion 3.1. Characterization of the impregnated ACs The prepared BAC samples were analyzed using Fourier transform-infrared spectroscopy (FT-IR; PerkinElmer Spectrum 100 Series) to ensure impregnation. Virgin BAC, BACCl, and BACAu were analyzed. Under the same conditions, the peak changes were measured as a function of the Au content, for which the impregnated Au content varied from 5 to 10 wt%. The Au NPS peaks were observed at 1048-1203 cm1 and 2910-3100 cm1. No peak was observed for both virgin BAC and BACCl in the relevant ranges, whereas BACAu produced peaks near 1100 and 2900 cm1 (Fig. 2). There was no significant difference between the peaks as a function of the Au content. X-ray diffraction (XRD; MiniFlex, Rigaku Co., Tokyo, Japan) analysis was used to determine the impregnation status of BAC impregnated with Cl and Au. Virgin BAC was used as a reference. In the case of Cl, unique peaks were observed at 23, 28, and 80 (2q). In the case of Au, unique peaks were observed at 37, 44, and 69 (2q). These results were because of a quality decision. Fig. 3 shows the results of the XRD measurements. Cl produced peaks at 23, 28 and

472

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

Fig. 1. A schematic of experimental setup.

60 (2q), and Au produced peaks at 37 and 69 (2q); no peak was observed at 44 (2q) because AC produced its own peak at 44. The intensities of the peaks varied with the varying Au content. The multi-point (MP) method was used to determine the distribution and size of the modified BAC micropores. The MP method used N2 gas at 77 K and 102.91 kPa. Table 1 shows the changes in pore size, pore volume, and surface area before and after Au impregnation. After Au impregnation, the surface area decreased from 1340.6 to 1233.1 m2 g1. Furthermore, the micropore volume decreased from 0.5850 to 0.5369 cm2 g1. The similarity between the mean diameter before and after Au impregnation suggests that the decrease in specific surface area was likely due to micropore closure by Au molecules.

3.2. Hg spiking efficiency: comparison of the target and actual spiked amount Fig. 2. FT-IR spectra of Au loaded on BAC.

Fig. 3. XRD analysis result of impregnated BACs.

The Hg spiking was tested for each impregnated material to determine the best condition with the highest efficiency. One gram of BAC was used, and the test progressed with. 5% aqua regia at a Au (5 wt%) impregnation rate that was equal to the rate of Cl impregnation. The measured density of spiked Hg was 100 ng g1 spiked using the Hg standard solution. The experimental results are shown in Fig. 4. The analytical Hg concentration with Cl and Au impregnated (101.67 ng g1) exhibited a 23% higher spiking efficiency than the Hg analytical concentration with no impregnation in the AC (77.43 ng g1). Furthermore, a higher spiking efficiency was obtained for BACAu than for BACCl. The Hg spiking efficiency increased as BACClAu > BACAu > BACCl > virgin BAC (Korpiel and Vidic, 1997). Au with Cl impregnation showed the highest efficiency in the Hg spiking test. Cl plays a role in acidifying the AC (Fu et al., 2014). Au is combined at the surface of the AC and inside the micropores. Due to the possible blockage of the AC micropores by Au molecules (Ghaedi et al., 2014), an adequate amount of Au impregnation was investigated. After setting the AC standard as 1 g, the Hg spiking efficiency was compared for 0.5e20 wt% of impregnated Au. The experimental Hg concentration was the same, i.e., 100 ng g1. The results are shown in Fig. 4. At 5 wt% of impregnated Au, the Hg

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

473

Table 1 Characteristics of virgin BAC and BACAu. Samples

Adsorptive Adsorption Temperature (K) Surface Area (m2 g1) Total Pore Volume (cm3 g1) Micropore Volume (cm3 g1) Mean Pore Diameter (nm)

virgin BAC N2 BACAu

77

1340.6 1233.1

0.5776 0.5302

0.5850 0.5369

1.7235 1.7199

Fig. 4. Hg spiked vs. (a) impregnation materials and (b) Au contents.

BACAu, BACClAu, and virgin BAC. BACCl had the fastest time because the impregnated aqua regia widened the micropores, thus preventing Au impregnation. After observing this result, no chemical adsorption was observed after the early stage of physical adsorption (Crockett and Kinnison, 1979). The time required to reach adsorption equilibrium for powdered AC and BAC was approximately 4 and 11 h, respectively. Regarding the Hg removal efficiency, BACClAu had the highest removal efficiency, followed by virgin BAC, BACCl, and then BACAu. The effect of the amount of impregnated Au on Hg adsorption was investigated (Fig. 5). The BACCl samples were further impregnated using 0, 5, 10, and 20 wt% Au solutions. The initial concentrations of the Hg solutions were 50, 100, 200, 300, 400, 500, 600, and 700 mg L1. The contact/soaking time was 12 h. Increases in the Au content resulted in an increased Hg spiking amount.

spiking reached an efficiency rate of 99.65%. As the rate of impregnated Au increased, a higher Hg concentration than the spiking concentration was measured. Based on this result, the surface chemistry had a greater effect than the texture because of the high binding strength between Hg and Au. 3.3. Hg spiking capacity: comparison of the aqueous Hg adsorption of various BAC samples To spike Hg into the AC of the sorbent trap, an Hg standard solution was used to spike liquid Hg into the AC. The time to reach adsorption equilibrium to produce AC was measured during this experiment. The initial concentration of the Hg solution was 1000 ng mL1. The contact/soaking times were 10, 20, 30, 40, 50, 60, 120, 240, 480, 720, and 1440 min. The results are shown in Table 2. Although the different samples differed slightly according to the impregnated material and type of AC, a dramatic increase in the adsorbed quantity in the first 40 min was typically observed. BACCl had the fastest adsorption equilibrium time of 50 min, followed by

3.4. Kinetics of Hg adsorption by BACClAu Pseudo-1st-order and pseudo-2nd-order models were tested to

Table 2 Effect of contact time on Hg adsorption. Initial Hg concentration 1000 ng mL1 Contact time (min) Sample

0

10

20

30

40

50

60

120

240

480

720

1440

Hg Concentration(ng mL1)

ACClAu BACClAu BACCl BACAu virgin BAC

1000 1000 1000 1000 1000

163 32.8 27.8 50.4 54.8

155 29.2 25.9 38.8 53.5

144 27.4 23.4 38.8 50.3

90.8 26.7 23.7 35.8 47.8

85 24.9 19.1 36 45.1

79.2 23.8 18.2 29.9 40.7

61.7 15.4 18.9 25 31.3

43.3 13.6 18.8 24.2 15.2

27 0.73 16.2 22.7 6.52

7.55 0.53 16 21.8 5.79

0.32 0.3 15.2 21.6 5.85

Hg Adsorbed (ng g1)

ACClAu BACClAu BACCl BACAu virgin BAC

0 0 0 0 0

8370.00 9672.00 9722.00 9496.00 9452.00

8450.00 9708.00 9741.00 9612.00 9465.00

8560.00 9726.00 9766.00 9612.00 9497.00

9092.00 9733.00 9763.00 9642.00 9522.00

9150.00 9751.00 9809.00 9640.00 9549.00

9208.00 9762.00 9818.00 9701.00 9593.00

9383.00 9846.00 9811.00 9750.00 9687.00

9567.00 9864.00 9812.00 9758.00 9848.00

9730.00 9992.70 9838.00 9773.00 9934.80

9924.50 9994.70 9840.00 9782.00 9942.10

9996.80 9997.00 9848.00 9784.00 9941.50

474

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

Fig. 5. Hg adsorbed by BACAu vs. Au contents.

Fig. 6. Pseudo-1st-order Hg removal by impregnated BAC at 25  C.

study the kinetics of Hg adsorption by BACClAu. The pseudo-1storder model is as follows (Ho and McKay, 2000):

dQt ¼ k1 ðQe  Qt Þ dt

(2)

  Qt ¼ Qe 1  ek1 t

(3)

lnðQe  Qt Þ ¼ lnQt  k1 t

(4)

where Qe is the quantity (ng g1) of Hg adsorbed onto the AC, Qt is the quantity (ng g1) of Hg adsorbed during time t, and k1 is the rate constant (min1). The pseudo-2nd-order model is as follows:

dQt ¼ k2 ðQe  Qt Þ2 dt

(5) Fig. 7. Pseudo-2nd-order Hg removal by impregnated BAC at 25  C.

t 1 1 ¼ þ t Qe k2 Qe2 Qe

(6)

where k2 is the rate constant (g mg1 min1). Table 3 shows the results obtained by applying the test data (Table 2) from the adsorption experiment according to the impregnated materials' contact times. Figs. 6 and 7 display these results in diagram form. When applying the data from the Hg adsorption test to the pseudo-1st-order and pseudo-2nd-order models, the correlation coefficient of the latter was considerably higher. According to this result, the Hg adsorption of BACClAu was based on the pseudo-2ndorder model. Furthermore, the absorption equilibrium based on the pseudo-2nd-order model was considerably lower than that based

on the pseudo-1st-order kinetic model. The Hg adsorption rates of BACCl, BACAu, and virgin BAC were 9,850, 9,780, and 9960 ng g1, respectively; the samples impregnated with Cl and Au had lower adsorption rates than BACClAu. 3.5. Hg adsorption isotherm of BACClAu The Langmuir and Freundlich isotherms were used to determine the isothermal adsorption characteristics of Hg impregnated in each material (Gupta and Ganesan, 2015).

Table 3 Experimental pseudo first-order and pseudo second-order kinetic parameters for Hg adsorption. 1st order

Adsorbent Qea ClAu

AC BACClAu BACCl BACAu virgin BAC a b

Experimental. Theoretical.

(ng g

10000

1

)

2

2nd order

R (%)

Qeb

72.01 78.18 37.70 65.81 90.27

1680.76 578.825 147.835 295.037 961.217

(ng g

1

)

k1 (min 0.0045 0.0085 0.005 0.0075 0.0107

1

)

R2 (%)

Qeb (ng g1)

k2 (g ng1 min1)

99.99 99.99 99.99 99.99 99.99

10000 10010 9850 9780 9960

0.0186 0.07901 0.283808 0.188875 0.057079

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

475

To measure the adsorption equilibrium as a function of the amount of Au (wt%) and the initial Hg dosage, 30 mL solutions with Hg concentrations of 50, 100, 200, 300, 400, 500, 600, and 700 mg L1 and 0.1 g of the adsorption medium were injected into a 50 mL vial. The concentration equilibrium and amount of adsorption were determined by measuring the Hg concentration remaining in solution after a reaction time of 12 h and removal of solids using a DigiFILTER. Eqs. (7) and (8) show the Langmuir isotherm and its re-arranged form, respectively, where Q is the amount (mg g1) of Hg adsorbed, Ce is the Hg concentration (mg L1) at equilibrium, Qmax is the maximum amount of adsorption (mg g1), and KL is the equilibrium constant.

Qmax kl Ce 1 þ kl Ce

(7)

Ce 1 1 ¼ þ Ce Qmax kl kl Q

(8)

Q¼

Fig. 8. Langmuir adsorption isotherm of Hg removal by impregnated BAC.

The Freundlich isotherm, where Kf and n are constants for a given adsorbate and adsorbent at a particular temperature, respectively, is given below: 1

Q ¼ Kf Cen

(9)

1 logQ ¼ logKf þ logCe n

(10)

Based on the parameters in Table 4, the Langmuir isotherm exhibited a higher average R2 value related to the Au content (92.92%) compared to Freundlich isotherm's average value (89.51%), indicating higher conformity of the former isotherm. However, at a higher amount of Au, the adsorption was more similar to the Freundlich isotherm than the Langmuir isotherm, indicating that the binding of AueHg and the p-cation interaction of the AC affected the adsorptive medium at the same time. The trends of the Langmuir adsorption isotherm and Freundlich adsorption isotherm are displayed in diagrams in Figs. 8 and 9. Fig. 9. Freundlich adsorption isotherm of Hg removal by impregnated BAC.

3.6. Performance test of the Hg sorbent trap The performance of the BACClAu-filled sorbent trap was evaluated. The conditions of the experiment are listed in Table 5. Elemental Hg was created and injected using a Hg0 calibrator. The current Hg discharge concentration through the thermoelectric power plant stack was approximately 1e5 mg m3. Thus, the Hg0 concentration to evaluate the performance of the sorbent trap was set at 1, 5, and 10 mg m3. The temperature was set at 80, 100, and 120  C to provide similar conditions to those of thermoelectric power plant stacks. Two sections of 0.3 g of BACClAu were filled into the sorbent trap for the experiment. The reactions were conducted for a total of 60 min under flow rates of 0.4, 0.6, and 1 lpm. The concentration of adsorbed Hg was measured by the CVAA Hg analyzer to verify that the amount of BACClAu, which finished Table 4 Langmuir and Freundlich isotherm parameters of Hg adsorption by BACAu. Gold (wt%)

Langmuir isotherm Qm

0 5 10 20

(mg g1)

82.644 81.967 109.890 174.413

Freundlich isotherm

Kl

R2

(%)

0.0052 0.0163 0.0176 0.0065

84.29 98.89 98.54 89.99

Kf

n

R2

(%)

1.7566 5.7161 11.3884 5.7013

2.7700 1.9282 1.7012 2.3110

74.73 89.01 99.49 94.83

Table 5 Performance evaluation condition of the sorbent trap. Sector

Range

Inlet Hg Concentration Temperature Flow Rate Time Adsorbent Amount

1, 5, 10 mg m3 80, 100, 120  C 0.4, 0.6, 1.0 lpm 60 min 0.3 g

adsorbing Hg0 from the gas, was accurate after preprocessing following the ASTM D6414 (ASTM D6414, 2001) standard test (Table 6). The results of the analysis of the BACClAu used in the experiments after the completion of the Hg adsorption experiment in the gas phase using the gas sorbent trap are detailed below. In general, the Hg adsorption rate was highest at 80  C; the adsorption rate was smaller for a larger flow rate and higher temperature. When analyzing the changing Hg inlet concentration (1, 5, and 10 mg m3), a higher adsorption rate of more than 95% of the average Hg concentration was observed. According to the US EPA, the measured value of the Hg content in the sampling section (Section 1) is not valid if it is greater than 20%

476

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

Table 6 Amounts of analysis for the sorbent trap (1, 5, and 10 mg m3). 

Inlet Hg conc. (mg m3)

Flow rate (lpm)

Temperature ( C)

Inlet Hg (mg)

Section

1

0.4

80

0.024

Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section

100 120 0.6

80

0.036

100 120 1.0

80

0.060

100 120 5

0.4

80

0.120

100 120 0.6

80

0.180

100 120 1.0

80

0.300

100 120 10

0.4

80

0.240

100 120 0.6

80

0.360

100 120 1.0

80

0.600

100 120

of that in the back-up section (Section 2) (US EPA, 2007). The results in Table 6 indicate that all of the measured Hg values were valid. 4. Conclusions In this study, AC (BACClAu) was impregnated with a halogen, Cl, and transition metal, Au, to improve its Hg impregnation capability for use in sorbent traps. Impregnating with 5% aqua regia and a 5 wt % Au standard solution exhibited the highest Hg spiking efficiency. However, the maximum amount of Hg adsorption increased as the Au content increased, indicating that Hg adsorption in the BACClAu solution was affected more by the surface chemistry than the texture. When applying pseudo-1st-order and pseudo-2nd-order models to estimate the adsorption kinetics, the adsorption

1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2

Hg (mg)

Total Hg (mg)

Adsorption (%)

0.0225 0.0008 0.0214 0.0013 0.0217 0.0008 0.0330 0.0012 0.0365 0.0011 0.0311 0.0070 0.0596 0.0012 0.0536 0.0012 0.0493 0.0020 0.1106 0.0055 0.0993 0.0036 0.1020 0.0028 0.1625 0.0058 0.1489 0.0045 0.1392 0.0040 0.3037 0.0097 0.2928 0.0076 0.2605 0.1106 0.2278 0.0020 0.2101 0.0029 0.2019 0.0012 0.3243 0.0008 0.2954 0.0012 0.3123 0.0012 0.5955 0.0154 0.6202 0.0038 0.5493 0.2278

0.0233

93.6 3.4 89.2 5.3 90.3 3.4 91.7 3.2 101.3 3.1 86.4 1.8 99.3 2.0 89.4 1.9 82.1 3.3 92.2 4.6 82.8 3.0 85.0 2.3 90.3 3.2 82.7 2.5 77.3 2.2 101.2 3.2 97.6 2.5 86.8 92.2 94.9 0.8 87.5 1.2 84.1 0.5 90.1 0.2 82.0 0.3 86.8 0.3 99.3 2.6 103.4 0.6 91.5 94.9

0.0227 0.0225 0.0342 0.0376 0.0318 0.0608 0.0548 0.0513 0.1161 0.1029 0.1047 0.1683 0.1534 0.1432 0.3134 0.3004 0.2716 0.2297 0.2130 0.2031 0.3251 0.2965 0.3136 0.6109 0.6240 0.5542

97.0 94.5 93.6 95.0 104.4 88.2 101.3 91.3 85.4 96.7 85.8 87.3 93.5 85.2 79.6 104.5 100.1 90.5 95.7 88.7 84.6 90.3 82.4 87.1 101.8 104.0 92.4

equilibrium theory based on the 2nd-order model corresponded better to the experimental data. Hg adsorption by BACClAu progressed with a pseudo-2nd-order model. Furthermore, when substituting the adsorption equilibrium data as a function of the initial concentration into the isothermal adsorption model, the average value of the Langmuir isotherm was higher than the average value of the Freundlich isotherm. However, as the content of Au increased, the data tended to follow the Freundlich isotherm, indicating that the binding of AueHg and the p-cation interaction made by the AC affected the adsorption medium at the same time. In the Hg gas adsorption experiment using a sorbent trap, the Hg adsorption rate was highest at 80  C. Finally, a Hg adsorption rate of more than 95% was observed when varying the inlet Hg concentration (1, 5, and 10 mg m3).

Y.C. Song, T.G. Lee / Chemosphere 165 (2016) 470e477

References ASTM D6414-01, 2001. Standard Test Method for Total Mercury In Coal and Coal Combustion Residues by Acid Extraction or Wet Oxidation/Cold Vapor Atomic Absorption. ASTM D6784-02, 2008. Standard Test Method For Elemental, Oxidized, ParticleBond, and Total Mercury in Flue Gas Generated from Coal-Fired Stationary Sources (Ontario Hydro Method). Chingombe, P., Saha, B., Wakeman, R.J., 2005. Surface modification and characterisation of a coal-based activated carbon. Carbon 43, 3132e3143. Crockett, A.B., Kinnison, R.R., 1979. Mercury residues in soil around a large coal-fired power plant. Environ. Sci. Technol. 13, 712e715. Fu, X., Heimbürger, L.-.E., Sonke, J.E., 2014. Collection of atmospheric gaseous mercury for stable isotope analysis using iodine-and chlorine-impregnated activated carbon traps. J. Anal. At. Spectrom. 29, 841e852. Ghaedi, M., Ghaedi, A.M., Ansari, A., Mohammadi, F., Vafaei, A., 2014. Artificial neural network and particle swarm optimization for removal of methyl orange by gold nanoparticles loaded on activated carbon and Tamarisk. Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 132, 639e654.  mez-Gime nez, C., Ballestero, D., Juan, R., Rubio, B., Izquierdo, M.T., 2015. Mercury Go capture by a regenerable sorbent under oxycoal combustion conditions: effect of SO2 and O2 on capture efficiency. Chem. Eng. Sci. 122, 232e239. Granite, E.J., Pennline, H.W., Hargis, R.A., 1998. Sorbents for Mercury Removal from

477

Flue Gas. US DOE Federal Energy Technology Center, pp. 1e47. Granite, E.J., Pennline, H.W., Hargis, R.A., 2000. Novel sorbents for mercury removal from flue gas. Ind. Eng. Chem. Res. 39, 1020e1029. Gupta, R., Ganesan, V., 2015. Gold nanoparticles impregnated mesoporous silica spheres for simultaneous and selective determination of uric acid and ascorbic acid. Sens. Actuators B Chem. 219, 139e145. Ho, Y.S., McKay, G., 2000. The kinetics of sorption of divalent metal ions onto sphagnum moss peat. Water Res. 34, 735e742. Kobiela, T., Nowakowski, B., Dus, R., 2003. The influence of gas phase composition on the process of AueHg amalgam formation. Appl. Surf. Sci. 206, 78e89. Korpiel, J.A., Vidic, R.D., 1997. Effect of sulfur impregnation method on activated carbon uptake of gas-phase mercury. Environ. Sci. Technol. 31, 2319e2325. Portzer, J.W., Albritton, J.R., Allen, C.C., Gupta, R.P., 2004. Development of novel sorbents for mercury control at elevated temperatures in coal-derived syngas: results of initial screening of candidate materials. Fuel Process. Technol. 85, 621e630. UNEP, 2013. U.N.E., Minamata Convention on Mercurydtext and Annexes. UNEP, Geneva, Switzerland. US EPA, 2007. Method 30B-determination of Total Vapor Phase Mercury Emissions from Coal-fired Combustion Source Using Carbon Traps. Zhang, Z., Eom, Y., Lee, M.J., Lee, T.G., 2016. Application of a sorbent trap system to gas-phase elemental and oxidized mercury analysis. Chemosphere 154, 293e299.