J O U RN A L OF E N V I RO N M EN TA L S CI EN CE S 32 (2 0 1 5 ) 2 0 7–2 1 6

Available online at www.sciencedirect.com

ScienceDirect www.journals.elsevier.com/journal-of-environmental-sciences

Enhancement of elemental mercury adsorption by silver supported material Rattabal Khunphonoi1 , Pummarin Khamdahsag2 , Siriluk Chiarakorn3 , Nurak Grisdanurak1,⁎, Adjana Paerungruang3 , Somrudee Predapitakkun4 1. Department of Chemical Engineering, Thammasat University, Pathumthani 12120, Thailand. E-mail: [email protected] 2. Environmental Research Institute, Chulalongkorn University, Pathumwan, Bangkok 10330, Thailand 3. Division of Environmental Technology, School of Energy, Environment and Materials, King Mongkut's University of Technology Thonburi, Bangkok 10140, Thailand 4. Petroleum Authority of Thailand (PTT) Research and Technology Institute, Wong-Noi, Ayutthaya 13170, Thailand

AR TIC LE I N FO

ABS TR ACT

Article history:

Mercury, generally found in natural gas, is extremely hazardous. Although average mercury

Received 21 August 2014

levels are relatively low, they are further reduced to comply with future mercury regulations,

Revised 19 January 2015

which are stringent in order to avoid releasing to the environment. Herein, vapor mercury

Accepted 20 January 2015

adsorption was therefore investigated using two kinds of supports, granular activated carbon

Available online 22 April 2015

(GAC) and titanium dioxide (TiO2). Both supports were impregnated by silver (5 and 15 wt.%), before testing against a commercial adsorbent (sulfur-impregnated activated carbon, SAC). The

Keywords:

adsorption isotherm, kinetics, and its thermodynamics of mercury adsorption were reported.

Adsorption

The results revealed that Langmuir isotherm provided a better fit to the experimental data.

Mercury

Pseudo second-order was applicable to describe adsorption kinetics. The higher uniform Ag

Silver

dispersion was a key factor for the higher mercury uptake. TiO2 supported silver adsorbent

Natural gas

showed higher mercury adsorption than the commercial one by approximately 2 times.

Amalgam

Chemisorption of mercury onto silver active sites was confirmed by an amalgam formation found in the spent adsorbents. © 2015 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Published by Elsevier B.V.

Introduction In natural gas production, several undesirable compounds may contaminate in natural gas as received. One of the major concerns is a contamination of heavy metals such as Hg, Cr, Cd, etc. Among those, mercury (Hg) compounds are heavily regulated. Mercury could exist in three oxidation states (Hg(ΙΙ), Hg(Ι), and Hg0) and is predominantly found as Hg0 in the gas phase. Mercury is considered to be a toxic element to human health. The Environmental Protection Agency (EPA)

has disclosed the associated mercury, with possible health effects and threshold amount of Hg, and should be below a level of whole blood 5.8 ppb (for human being) (Slotnick, 2012). Elemental Hg in natural gas received from the Southeast Asia region is found in high levels of concentration up to 2000 μg/Nm3 (ca. 2400 ppbw) (Eckersley, 2010). Not only elemental Hg, but also mercury compounds such as, mercuric chloride (HgCl2), methyl mercuric chloride (CH3HgCl), dimethyl mercury (CH3HgCH3) and diethyl mercury (C2H5HgC2H5) were found (Wilhelm and Bloom,

⁎ Corresponding author. E-mail: [email protected] (Nurak Grisdanurak).

http://dx.doi.org/10.1016/j.jes.2015.01.008 1001-0742/© 2015 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Published by Elsevier B.V.

208

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S 32 (2 0 1 5 ) 2 07–2 1 6

2000). Therefore, several treatments have been incorporated to reduce the Hg concentration contaminated in natural gas to the threshold level. To reduce the vapor Hg dispersed in the air after combustion, the USEPA sets a standard Hg level in air to be less than 0.1 ppb. Consequently, refineries and gas separation plants generally reduce the levels of Hg in gas processing down to 0.001 ppb (Mokhatab and Poe, 2012). Treatments of mercury in liquid state have been extensively investigated (Ghassabzadeh et al., 2010; Ma et al., 2009; Idris et al., 2011). The removal of Hg from aqueous phase is performed using the Outokumpu process, the Bolkem process, selenium filter process, the Boliden–Norzink process and sulfide precipitation (Louie, 2008). They have been applied to commercial scales. The processes are carried out in scrubbers using acid solution (H2SO4 and H2SeO3) to scrub elemental mercury (Hg0) in the gas into a sulfate and mercury selenide for further removal. These methods require chemicals and chemical treatments, in which, a large amount of chemical waste is produced. Unlike mercury treatment in liquid phase, the treatment in vapor phase has not been widely studied. The removal of mercury in vapor phase is mainly performed by adsorption technique. The process has been investigated by various adsorbents, including calcium-based one (Ghorishi and Sedman, 1998), coal (Dίaz-Somoano et al., 2007), zeolites (Morency, 2002), iodine-modified rice husk ash (Zhao et al., 2010), activated carbon (Sasmaz et al., 2012; Padak and Wilcox, 2009) and carbonaceous material derived from sewage sludge (Fang et al., 2010). However, the uptake of mercury onto adsorbents was extremely inefficient. Based on the property of mercury, the physical adsorption between Hg0 and active adsorbents may not be attractive. Active sites on the adsorbent surface should be modified for chemical attraction. It has been noted that metals, such as palladium, platinum, rhodium, gold, zinc, aluminum, copper and silver, are ready to form amalgam with Hg0 (Granite et al., 2006; Wilcox et al., 2012). Besides, the solubility of these metal mercury amalgams is relatively low, and consequently very little mercury releases after the uptake (Henderson et al., 2001). Among those chemicals, silver was reported for having lower solubility, therefore, it was selected to modify adsorbent support and create more active sites. In this research, the suitable loading of Ag on two kinds of adsorbent supports, TiO2 and granular activated carbon (GAC), was investigated. Selected adsorbents, based on high adsorption performance, were reported adsorption isotherms, temperature effect on the adsorption and adsorption kinetics.

1. Materials and methods 1.1. Preparation Two types of materials, TiO2 (P25) and GAC, were used as adsorbent supports. Silver was loaded onto the materials via an impregnation technique. AgNO3 corresponding to 5% and 15% by weight of Ag over the supports, was dissolved in deionized (DI) water under a vigorous stirring condition until the solution was homogeneous. The solutions were gradually dropped onto the supports, TiO2 or GAC particles. The samples were sonicated and kept for 3 hr, then dried at 80°C

for 6 hr and ground with an agate mortar. The obtained adsorbents in powder form were calcined in air at 480°C with the heating rate of 10°C/min and kept for 3 hr. Adsorbent powder was pelletized to cylindrical form and sieved to 2 mm in diameter before testing. Commercial adsorbent used in this study was sulfurimpregnated activated carbon (SAC) supplied by Carbokarn, Thailand.

1.2. Characterization The physico-chemical properties of adsorbents were examined by various techniques such as X-ray diffraction (XRD), nitrogen adsorption isotherms, and field emission scanning electron microscopy (FESEM). The XRD patterns were obtained using CuKα radiation on Bruker AXS diffractometer (Bruker-AXS D8-A25 Advance, Karlsruhe, Germany). The samples were scanned from 10 to 80° (2θ) in steps of 0.02° per second. Nitrogen adsorption isotherms of samples were measured using an Autosorb-1 analyzer (Quantachorme Instruments, Boynton Beach, FL, USA) for calculating the Brunauer–Emmett–Teller (BET) surface area. Sample micrograph was observed under field emission scanning electron microscope (FESEM, JSM-6301F, JEOL, Ltd., Tokyo, Japan). Powder sample was scattered on an adhesive tape on a brass bar. The sample was then coated with carbon and transferred into the sample chamber. The microscope was equipped with energy dispersive X-ray spectrometry (EDS, INCA 350, Oxford Instruments, High Wycombe, UK), for elemental mapping images.

1.3. Mercury adsorption The adsorption vessel for static test was made of three parts: holder, Pyrex glass-stand and cap. Hg0 was dropped into the holder. A Pyrex glass-stand was placed over 0.5 g of Hg0. Before the study, a blank test was carried out for the adsorption baseline. In the test, a certain amount of prepared adsorbent was placed on the Pyrex glass-stand (Fig. 1). The holder was closed tightly with a cap and wrapped with parafilm. The vessel was placed inside the heating bath throughout the test. Temperatures were set at 40 and 60°C for the tests, which could minimize mercury stick on the surface of experimental vessel. The adsorption was carried out for 40–60 days. Mercury was determined by a mercury analyzer (Lumex-RA915+, Ohio Lumex, Twinsburg, OH, USA). The mercury amount in spent adsorbent (after mercury uptake) was measured by the mercury amount according to US EPA 7473. The adsorbent was thermally decomposed and amalgamation with gold and detected by cold vapor atomic fluorescent spectrometry (CV-AFS). The detection limit was 0.5 μg/kg. To prevent erroneous interpretation of instrument readings, sample was diluted with the same support material. Mercury vapor concentration inside the vessel was also analyzed in the same instrument with a frequency of 1 Hz at concentrations ranging from 0.01 to 22 μg/m3 using a long-path analytical cell. To investigate the effect of temperature on the adsorption, the experiment was conducted at four different temperatures of 30, 40, 50 and 60°C. In order to prevent the multilayer adsorption, a certain amount of 0.2 g of Hg0 was used.

209

J O U RN A L OF E N V I RO N M EN TA L S CI EN CE S 32 (2 0 1 5 ) 2 0 7–2 1 6

Adsorbent Mercury

Fig. 1 – Adsorption vessel for static test.

2. Results and discussion 2.1. Characterization Fig. 2 shows the XRD patterns of adsorbents based on GAC and TiO2, respectively. As-received materials and materials doped with Ag with the loading of 5 and 15 wt.% are presented in Fig. 2a and b, respectively. After the Ag modification, four sharp peaks appearing at 37.7, 43.8, 64.2, and 77.1° were observed, corresponding to the (111), (200), (220), and (311) diffraction planes, respectively, of cubic silver (JCPDS No. 4-0783). Considering GAC doped Ag, XRD peaks of silver clearly presented for both loadings. The higher amount of silver doping shows the stronger and sharper diffraction peaks. Typically, TiO2 crystals have two typical phases: anatase and rutile phases, which are shown in Fig. 2b. 5% Ag/TiO2 spectra did not indicate significant presence of Ag metallic features. The peaks of Ag clearly appeared in the XRD pattern for 15% Ag/TiO2. The effects of Ag doping on the specific surface area were determined by BET surface area. The specific surface area of the pure GAC and TiO2 were ca. 1300 and 50 m2/g, respectively. Both of these values decreased upon doping, as shown in Table 1, due to micropore filling/blockage by dispersed silver metals on the surface.

a

FESEM images presented in Fig. 3 show surface morphology of GAC samples at 100, 1000 and 10,000 magnitudes. The FESEM images of pure GAC, as shown in Fig. 3(a–c), illustrate the availability of pores, which are in large range of pore sizes. It was also apparently that surface of GAC was clean without any depositions, and had high porosity as well as absence of cracks on the surface. Upon doping Ag onto GAC, significant changes in the porosity morphology of GAC were observed. Fig. 3(d–f) shows FESEM images of 5% Ag/GAC showing layered structure of particles. A reduction of the porosity of the GAC was observed. The size of Ag particles on GAC was approximately 1–2 μm. Fig. 3(g–i) illustrates the morphology of 15% Ag/GAC. The size of Ag particles on GAC was approximately 0.5–1 μm. The particles also exhibited a platelet structure. The FESEM images showed the presence of bright metallic features on the surfaces of the samples. Fig. 4 presents FESEM images for pure TiO2, 5% Ag/TiO2 and 15% Ag/TiO2 at magnitudes 10,000, 50,000, and 70,000. Though there were no significant differences in surface morphology, FESEM revealed that the three samples were unique in both particle size and shape. It was found that the particle size ranged from around 25 to 50 nm. The presence of Ag on the surfaces of TiO2 was also observed, and Ag dispersion over this sample was higher than that on the surface of GAC. Ag particles adhered to the TiO2 surfaces as can be clearly observed from Fig. 4(d–g). It is

A: Anatase R: Rutile

b

Ag

A Ag

Ag Ag

15%Ag/TiO2

A A R ARAg

RA R

Intensity (a. u.)

Intensity (a. u.)

15%Ag/GAC

R

Ag

5%Ag/GAC

R Ag A A R Ag

5%Ag/TiO2

GAC

TiO2 20

30

40

50 2θ (degree)

60

70

80

20

30

40

50 2θ (degree)

60

70

80

Fig. 2 – X-ray diffraction (XRD) patterns of with and without Ag loading on granular activated carbon (GAC) (a) and TiO2 (b).

210

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S 32 (2 0 1 5 ) 2 07–2 1 6

Table 1 – Specific surface area of studied adsorbents. 2

Adsorbents

Surface area (m /g)

GAC 5% Ag/GAC 15% Ag/GAC TiO2 5% Ag/TiO2 15% Ag/TiO2

1360 907 520 50 34 31

GAC: granular activated carbon.

indicated that TiO2 possesses high chemical surface heterogeneity compared to GAC. This was in accordance with the observation reported by Lei et al. (2014); Pant et al. (2013); and Ortiz-Ibarra et al. (2007).

2.2. Hg adsorption 2.2.1. Preliminary adsorption tests The tests were carried out at 40 and 60°C which meant that the vapor concentration had a vapor pressure of 8.5 × 10−4 and 3.5 × 10−3 kPa, respectively (Huber et al., 2006). Fig. 5a and b shows the adsorption amount at 40 and 60°C, respectively. Considering the adsorption results at 40°C (Fig. 5a), it was clear that pristine TiO2 showed very little mercury adsorption,

while GAC itself presented higher adsorption capacity but was still low. For the pristine one, it seemed that higher adsorption relied on higher surface area. To enhance the adsorption capacity of mercury, the Ag modification on both GAC and TiO2 was done and it was found that it definitely increased the adsorption much better. It was observed that 5% Ag based TiO2 showed higher mercury uptake (52 mgHg/gadsorbent) than 5% Ag based GAC (21 mgHg/gadsorbent). This was probably due to the fact that uniform dispersion of Ag onto the support TiO2 was much better than that onto GAC support. The investigation of Ag dispersion can be seen clearly by scanning electron microscopy (SEM) coupled to EDS of spent adsorbents in Section 2.3. The higher the Ag loading adsorbent was, the higher the uptake of mercury could be easily observed. However, the test was extended to 60 day-adsorption period, to ensure that the equilibrium was reached. The amounts of mercury adsorption over 15% Ag/TiO2 and 15% Ag/GAC are 105 and 79 mgHg/gadsorbent, respectively (Fig. 5a). The results were in the same capacity of mercury uptake using metal chlorides loaded on activated carbon reported by Shen et al. (2010). Besides, the mercury uptakes for both modified adsorbents (15% Ag) surpassed that for SAC adsorbent, which was of 58 mgHg/gadsorbent. The adsorption results at 60°C are shown in Fig. 4b. Overall, the results were found in the same patterns of the adsorption at 40°C, as previously described. The only difference was

a

b

c

d

e

f

g

h

i

Fig. 3 – Field emission scanning electron microscopy (FESEM) photographs of (a–c) fresh granular activated carbon (GAC), (d–f) 5% Ag/GAC and (g–i) 15% Ag/GAC samples with the magnitude of 100, 1000 and 10,000 (left–right).

211

J O U RN A L OF E N V I RO N M EN TA L S CI EN CE S 32 (2 0 1 5 ) 2 0 7–2 1 6

a

b

c

d

e

f

g

h

i

Fig. 4 – Field emission scanning electron microscopy (FESEM) photographs of fresh TiO2 (a–c), 5% Ag/TiO2 (d–f) and 15% Ag/TiO2 (g–i) samples at magnitudes of 10,000, 50,000 and 70,000 (left–right).

adsorption capacity between these two conditions. Higher adsorptions were observed in higher temperature conditions. 15% Ag/TiO2 showed the highest mercury adsorption for the whole experiment. However, it should be noted that the adsorption still increased along the experiment time, and was not able to reach the equilibrium. This was perhaps due to the

multilayer adsorption by the induction of adsorbed Hg molecules and Hg molecules in the vapor phase.

2.2.2. Adsorption isotherms Two general isotherms, Langmuir and Freundlich, are used to describe the experimental adsorption isotherm in this study.

120

120

a

15% Ag/TiO2

b

100

15% Ag/TiO2

100

60

5% Ag/TiO2

40

SAC 5% Ag/GAC

20

0

10

20

30 40 Time (day)

80 5% Ag/TiO2

60

SAC 40 5% Ag/GAC

20

GAC TiO2

0

15% Ag/GAC qt (µg Hg/gadsorbent)

qt (µg Hg/gadsorbent)

15% Ag/GAC 80

GAC TiO2

0 50

60

0

10

20

30 40 Time (day)

Fig. 5 – Adsorption capacities of adsorbents (mg/g) at 40°C (a) and 60°C (b).

50

60

212

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S 32 (2 0 1 5 ) 2 07–2 1 6

15% Ag/TiO2

15% Ag/GAC 4.8

0.6 Langmuir model

4.6

0.5

4.4 4.2 lnCe

0.4 Ce / qe

Freundlich model

0.3

4.0 3.8 3.6

0.2

3.4 3.2

0.1 0

5

10

15

20 Ce (ppm)

25

30

3.0 1.0

35

1.5

2.0

2.5 lnqe

3.0

3.5

Fig. 6 – Linear plots of mercury adsorption on 15% Ag/TiO2 and 15% Ag/GAC fitting two isotherm models.

Langmuir adsorption is based on the assumption that the molecules of the adsorbate form monolayer on the surface of the adsorbent. The adsorbed site does not induce any adsorbate for more adsorption. The Langmuir isotherm model is employed by the following equation: q CeK qe ¼ max CeK þ 1

ð1Þ

where, Ce (mg/L) is the equilibrium concentration of Hg, qe (mg/gadsorbent) is the amount of mercury per unit mass of adsorbent at equilibrium, qmax (mg/gadsorbent) is the maximum amount of mercury adsorbed per unit mass of adsorbent for the formation of complete monolayer on the surface of adsorbent, and K (L/mg) is Langmuir equilibrium constant related to energy of adsorption. The constants are

Table 2 – Isotherm parameters for vapor Hg adsorption at 40°C. Linearized form of Langmuir model: Ce Ce 1 q ¼q Kþq e

max

K (L/mg)

R2

81.97 111.11 50.76

0.0001 0.0002 0.0005

0.990 0.995 0.997

qe ¼ kC 1=n e

ð2Þ

where, k and n are Freundlich constants, k is a measure of amount of adsorption, while n represents the degree of linearity. Both values are evaluated from the slope and interception of the plot of ln(qe) versus ln(Ce), as shown in Fig. 6. The fitted parameters of Langmuir and Freundlich models are listed in Table 2. The correlation coefficient R2 of Langmuir model was around 0.99, higher than that of Freunlich model (R2 around 0.85–0.92). It indicates that the Langmuir models suitably predict describing the mercury adsorption system. According to the maximum adsorption capacities (qmax) presented in Table 2, 15% Ag/TiO2 confirms the highest value (ca. 111 mg Hg/gadsorbent) compared to other adsorbents. It was

max

15% Ag/GAC 15% Ag/TiO2 SAC Linearized form of Freundlich model: ln ðqe Þ ¼ ln k þ n1 ln ðC e Þ 15% Ag/GAC 15% Ag/TiO2 SAC

qmax (mg/gadsorbent)

determined from slope and interception of linear plot of Ce/ qe versus Ce, as shown in Fig. 6. Freundlich isotherm involves formation of multilayers and could be expressed as Eq. (2).

k

1/n

0.015 0.027 0.025

0.442 0.382 0.203

R2

0.878 0.880 0.727

GAC: granular activated carbon; SAC: Sulfur-impregnated activated carbon. Ce: the equilibrium concentration of Hg; qe: the amount of mercury per unit mass of adsorbent at equilibrium; qmax: the maximum amount of mercury adsorbed per unit mass of adsorbent for the formation of complete monolayer on the surface of adsorbent; k, n: Freundlich constants; K: Langmuir equilibrium constant related to energy of adsorption.

Table 3 – Kinetic parameters for mercury adsorption at 40°C onto adsorbents. Model

15% Ag/TiO2

15% Ag/GAC

Pseudo  first-order model: ln qeq−qt ¼ −k1 t e Pseudo second-order model: t t 1 qt ¼ qe þ k2 q2e Intraparticle diffusion model: qt = kit 0.5 + A

−k1 = 0.0047 R2 = 0.949 k2 = 6.25 × 10−5 R2 = 0.973 ki = 3.028 A = 1.318 R2 = 0.982

−k1 = 0.0054 R2 = 0.914 k2 = 6.19 × 10−5 R2 = 0.968 ki = 2.265 A = −0.153 R2 = 0.989

GAC: granular activated carbon; qt: the amount of gas adsorbed on the adsorbent surface at any time; qe: adsorption capacity of adsorbent at equilibrium; k1 (hr−1): the rate constant of pseudo first-order; k2 (g/(mg·hr)): the rate constant of pseudo second-order; ki (g/(mg·hr1/2)): the rate constant of intraparticle diffusion model; A: intraparticle diffusion constant; t: time.

213

J O U RN A L OF E N V I RO N M EN TA L S CI EN CE S 32 (2 0 1 5 ) 2 0 7–2 1 6

15% Ag/TiO2

0.0

15% Ag/GAC Pseudo first-order model

ln(qe-(qt /qe))

-0.2

time of 40 hr. The experimental data were analyzed by three general types of kinetic adsorption models. These include pseudo first-order, pseudo second-order, and intraparticle diffusion models, which are expressed as Eqs. (3)–(5) below. Pseudo first-order model dqt ¼ k1 ðqe −qt Þ dt

-0.4

ð3Þ

Pseudo second-order model -0.6

-0.8

dqt ¼ k2 ðqe −qt Þ2 dt Intraparticle diffusion model 0

20

40

60 80 Time (hr)

100

120

qt ¼ ki t 0:5 þ A

20

15

t/qt

10

5

0

Pseudo second-order model 0

200

400

600

800 1000 Time (hr)

1200

1400

140 120

Intraparticle diffusion model

100

qt

80 60 40 20 0 -20

ð4Þ

0

10

20 Time0.5 (hr0.5)

30

40

Fig. 7 – Linear plots of mercury adsorption on 15% Ag/TiO2 and 15% Ag/GAC, evaluated for pseudo-first order, second-order and intraparticle diffusion model.

ð5Þ

where, qt (mg/g) is the amount of gas adsorbed on the adsorbent surface at any time (hr), qe (mg/g) is the adsorption capacity of adsorbent at equilibrium, k1 (hr− 1) is the rate constant of pseudo first-order, k2 (g/(mg·hr)) is the rate constant of pseudo second-order, ki (g/(mg·hr1/2)) is the rate constant of intraparticle diffusion model, t is the time (hr) and A is the intraparticle diffusion constant. The kinetic parameters of each model were evaluated by the linearization of model as described in Table 3. The plots of linearized equations are shown in Fig. 7. The calculated values were further taken from the slopes and intercepts and given in the same table. The results showed that the correlation coefficients (R2) for the pseudo first-order model were 0.95 and 0.91 for 15% Ag/TiO2 and 15% Ag/GAC, respectively. On the other hand, the correlation coefficients for the pseudo second-order model were 0.97 and 0.97 for 15% Ag/TiO2 and 15% Ag/GAC, respectively, which are higher than those for the pseudo first-order model. The pseudo second-order model was, therefore, more appropriate to describe the kinetics for this adsorption. Besides, the mechanism rate might be dominated and limited by chemical adsorption process between mercury molecules and the active sites of silver dispersed on the support surface. Since the structure of both studied adsorbents contained micropores and mesopores, mass transport of adsorbate molecules (Hg vapor) within the pores of studied adsorbents, called “intraparticle diffusion”, was also tested for studying the limiting step of the adsorption. It was noted that the plot of linearized form, which passes through the origin, could refer to intraparticle diffusion as a limiting step of the adsorption (Vasiliu et al., 2011).

Table 4 – Effect of temperature on Langmuir model parameters of mercury adsorption on to 15% Ag/TiO2. further selected for the thermodynamic study of adsorption behavior.

2.2.3. Adsorption kinetics The rate of mercury uptake is an important characteristic to provide an understanding of adsorption mechanism. The study focused on kinetic data of 15% Ag/GAC and 15% Ag/TiO2 at 40°C, shown in Fig. 5. The adsorption of mercury increased during 20 hr and approached the equilibrium by an approximate

Temperature (°C)

qmax (mg/gadsorbent)

KL (L/mg)

30 40 50 60

115.5 111.1 103 86.7

0.42 0.17 0.08 0.03

qmax: the maximum amount of mercury adsorbed per unit mass of adsorbent for the formation of complete monolayer on the surface of adsorbent; KL: Langmuir equilibrium constant.

214

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S 32 (2 0 1 5 ) 2 07–2 1 6

Table 5 – Thermodynamic parameters adsorption onto 15% Ag/TiO2. ΔH (kJ/mol) Temperature (°C) ΔG (kJ/mol) ΔS (J/mol)

−61.259 30 −10.095 −168.77

40 −7.753 −170.86

for

mercury

50 −5.666 −172.03

60 −3.075 −174.65

ΔH: the enthalpy; ΔG: the Gibbs free energy; ΔS: the entropy.

In this case, the linearized plots did not pass through the origin strictly. It was possible that the adsorption could be simultaneously controlled by multi-diffusion steps which were film diffusion and intraparticle diffusion.

2.2.4. Temperature effect on the adsorption To evaluate their thermodynamic inherent energy values, temperature effect on the adsorption was investigated. The equilibrium adsorption capacities for different temperatures were performed, and analyzed based on Langmuir isotherm model. The results of adsorption capacities in temperature range of 30–60°C are tabulated in Table 4. In this part, a mass of Hg was controlled at 0.2 g. This was to prevent the induction of mercury remaining in vapor cloud inside the containment. As seen in Table 4, the monolayer adsorption capacity was found to decrease at higher temperatures. This suggested that the adsorption process of mercury onto studied adsorbents followed an exothermic process in nature. The adsorption of Hg vapor on Ag/TiO2 was decreased upon the increase in temperature. It referred that the rate of interparticle diffusion of the gas molecules might be increased. In other words, the higher temperature gas molecules tended to escape from the active sites of adsorbent to the bulk phase. As shown in Section 2.2.2, the Langmuir isotherm model was found to satisfactorily predict with the experimental data, and the Gibbs free energy of the reversible adsorption could be obtained from the equilibrium constant, as presented in Eq. (6). The enthalpy change was obtained via the Clausius–Clapeyron equation, and the integrated result could be presented in Eqs. (7)

a

ο

Δ+

ð6Þ ð7Þ ð8Þ

ΔG ¼ ΔH−TΔS

ð9Þ

where, ΔH is the enthalpy, ΔG is the Gibbs free energy, ΔS is the entropy, R is the gas constant, and T is the temperature. Thermodynamic parameters of mercury adsorption in the studied temperature of 30–60°C are summarized in Table 5. It was found that the heat of adsorption reaction was negative, confirming that adsorption reaction presented in an exothermic process. In general, the absolute number of ΔH lying in the range of 2–20 kJ/mol and 80–200 kJ/mol, represents adsorption type of physical and chemical adsorption, respectively. The value of ΔH in the present study was 61 kJ/mol, where chemisorption was expected to dominate the adsorption mechanism of mercury vapor onto 15% Ag/TiO2. The negative values of the Gibbs' free energy explained spontaneous behavior of mercury adsorption on the mentioned adsorbents. In addition, the spontaneous adsorption declined upon higher temperature. The negative value of entropy change implies a decrease in the organization (randomness) of Hg vapor between the adsorbate/adsorbent interfaces.

2.3. Spent absorbent characterization Spent 15% Ag/GAC and 15% Ag/TiO2 were characterized by XRD, as shown in Fig. 8. It was found that XRD fingerprint peaks of chemisorbed Ag–Hg species appeared on both adsorbents. It was confirmed by the formation of Ag2Hg3 and Ag3Hg2 amalgam on GAC and TiO2 supports, respectively. Besides, Hg and HgO peaks were also observed over GAC, but

A

b

R

15%Ag/TiO2-Hg

ο

5%Ag/GAC-Hg



d ln ðC e Þ −ΔH ¼ dT RT 2 ΔH þA ln ðC e Þ ¼ RT

Intensity (a.u.)

Intensity (a.u.)

+

ΔG ¼ −RT ln q max;T K

+ Ag2Hg3 o Hg Δ HgO

+ + ο

15%Ag/GAC-Hg +

and (8), respectively. Finally, the entropy of adsorption is therefore readily obtained in Eq. (9).

+

+ Ag3Hg2 A: Anatase R: Rutile

A

R A AR +

A

RA R

R +

AA

R

+

5%Ag/TiO2-Hg

GAC-Hg

TiO2-Hg

20

40 2θ (degree)

60

80

20

40 2θ (degree)

60

Fig. 8 – X-ray diffraction (XRD) patterns of spent (a) GAC, Ag/GAC, and (b) TiO2, Ag/TiO2 adsorbents.

80

215

J O U RN A L OF E N V I RO N M EN TA L S CI EN CE S 32 (2 0 1 5 ) 2 0 7–2 1 6

Ag La1

Spent 15% Ag/GAC

Spent 15% Ag/TiO2

C Ka1_2 Ka1_1

Ag La1

Hg La1

Hg La1

Fig. 9 – Field emission scanning electron microscopy (FESEM) and elemental mappings of spent 15% Ag/GAC and 15% Ag/TiO2 in Hg adsorption for 40 days.

not over TiO2 support. This referred that some physical adsorption might take place over GAC surface, as well. Fig. 9 shows elemental mappings of spent adsorbents. The uniform dispersion of Ag on TiO2 was higher than that on GAC. We also observed more mercury on 15% Ag/TiO2 spent adsorbent, compared to 15% Ag/GAC. It was then suggested that support material established highly uniform silver distribution, which would enhance the mercury adsorption capability.

3. Conclusions The adsorption of mercury vapor on two different types of silver supported materials (TiO2 and GAC) was experimentally investigated. The adsorption isotherms, heats of adsorption and the adsorption kinetics and mechanisms within the temperature range of 30–60°C were discussed. Based on the analysis, silver supported on TiO2 had a higher adsorption capacity than that on GAC and a commercial adsorbent (SAC). This was mainly due to higher surface heterogeneity enhancing silver dispersed uniformly. The function of silver loading was directly related to the capacity of adsorption. The isotherm and kinetic adsorption experimental data fit well into Langmuir isotherm and the pseudo second-order model. Adsorption behaved spontaneously and exothermically. It was favorable at low temperature which is practical for a gas separation plant. Heat of adsorption was determined by the Clasius– Clapeyron equation and found to be ca. 60 kJ/mol, showing that chemisorption dominated the interaction of mercury vapor and active species.

Acknowledgments The authors would like to thank Petroleum Authority of Thailand (PTT) and Thailand Research Fund through the Royal Golden Jubilee Ph.D. Program (Grant no. PHD/0371/2552) for their financial supports.

REFERENCES

Dίaz-Somoano, M., Unterberger, S., Hein, R.G.K., 2007. Mercury emission control in coal-fired plants. Fuel Process. Technol. 88 (3), 259–263. Eckersley, N., 2010. Advance Mercury Removal Technologies. Available at, http://business.highbeam.com/436407/article-1G1218509957/advanced-mercury-removal-technologies (Date accessed: 01 January 2010). Fang, P., Cen, C.P., Chen, D.S., Tang, Z.X., 2010. Carbonaceous adsorbents prepared from sewage sludge and its application for Hg0 adsorption in simulated flue gas. Chin. J. Chem. Eng. 18 (2), 231–238. Ghassabzadeh, H., Mohadespour, A., Torab, M., Zaheri, P., Maragheh, M.G., Taheri, H., 2010. Adsorption of Ag, Cu, Hg from aqueous solutions using expanded perlite. J. Hazard. Mater. 177 (1-3), 950–955. Ghorishi, S., Sedman, C.B., 1998. Low concentration mercury sorption mechanisms and control by calcium-based sorbents: application in coal-fired processes. J. Air Waste Manag. Assoc. 48 (12), 1191–1198. Granite, E.J., Myers, C.R., King, W.O., Stanko, D.C., Pennline, H.W., 2006. Sorbents for mercury capture from fuel gas with application systems. Ind. Eng. Chem. Res. 45 (13), 4844–4848 . Henderson, D.C., Cilfford, R., Young, D.M., 2001. Mercury-reactive lymphocytes in peripheral in blood are not a maker foe dental amalgam associate disease. J. Dent. 29 (7), 469–474. Huber, M.L., Laesecke, A., Friend, D.G., 2006. The vapor pressure of mercury. NISTIR 6643. Available at http://www.boulder.nist. gov/div838/publications.html. Idris, A.S., Harvey, R.S., Gibson, T.L., 2011. Selective extraction of mercury(II) from water samples using mercapto functionalised-MCM-41 and regeneration of the sorbent using microwave digestion. J. Hazard. Mater. 193, 171–176. Lei, X.F., Xue, X.X., Yang, H., 2014. Preparation and characterization of Ag-doped TiO2 nanomaterials and their photocatalytic reduction of Cr(VI) under visible light. Appl. Surf. Sci. 321, 396–403. Louie, K.D., 2008. Handbook of Sulphuric Acid Manufacturing. 2nd ed. DKL Engineering Inc. Publications, Ontario, Canada. Ma, F., Qu, R.J., Sun, C.M., Wang, C.H., Ji, C.N., Zhang, Y., et al., 2009. Adsorption behaviors of Hg(II) on chitosan functionalized by amino-terminated hyperbranched polyamidoamine polymers. J. Hazard. Mater. 172 (2-3), 792–801.

216

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S 32 (2 0 1 5 ) 2 07–2 1 6

Mokhatab, S., Poe, W.A., 2012. Handbook of Natural Gas Transmission and Processing. 2nd ed. Elsevier Inc. Publications, Oxford, United Kingdom. Morency, J., 2002. Zeolite sorbent that effectively removes mercury from flue gases. Filtr. Sep. 39 (7), 24–26. Ortiz-Ibarra, H., Casillas, N., Soto, V., Barcena-Soto, M., Torres-Vitela, R., de la Cruz, W., et al., 2007. Surface characterization of electrodeposited silver on activated carbon for bactericidal purposes. J. Colloid Interface Sci. 314 (2), 562–571. Padak, B., Wilcox, J., 2009. Understanding mercury binding on activated carbon. Carbon 47 (12), 2855–2864. Pant, H.R., Pant, B., Sharma, R.K., Amarjargal, A., Kim, H.J., Park, C.H., et al., 2013. Antibacterial and photocatalytic properties of Ag/TiO2/ZnO nano-flowers prepared by facile one-pot hydrothermal process. Ceram. Int. 39 (2), 1503–1510. Sasmaz, E., Kishofer, A., Jew, A.D., Caha, A., Abram, D., Jaramillo, T.F., et al., 2012. Mercury chemistry on brominated activated carbon. Fuel 99 (9), 188–196.

Shen, Z.M., Ma, J., Mei, Z.J., Zhang, J.D., 2010. Metal chlorides loaded on activated carbon to capture elemental mercury. J. Environ. Sci. 22 (11), 1814–1819. Slotnick, S., 2012. Mercury: Human Exposure. Available at. http:// www.epa.gov/hg/exposure.htm. Vasiliu, S., Bunia, I., Racovita, S., Neagu, V., 2011. Adsorption of cefotaxime sodium salt on polymer coated ion-exchange resin microparticles: kinetics, equilibrium and thermodynamic studies. Carbohydr. Polym. 85 (2), 376–387. Wilcox, J., Rupp, E., Ying, S.C., Lim, D.-H., Negreira, A.S., Kirchofer, A., et al., 2012. Mercury adsorption and oxidation in coal combustion and gasification processes. Int. J. Coal Geol. 90–91, 4–20. Wilhelm, S.M., Bloom, N., 2000. Mercury in petroleum. Fuel Process. Technol. 63 (1), 1–27. Zhao, P.F., Guo, X., Zheng, C.G., 2010. Removal of elemental mercury by iodine-modified rice husk ash sorbents. J. Environ. Sci. 22 (10), 1629–1636.

Enhancement of elemental mercury adsorption by silver supported material.

Mercury, generally found in natural gas, is extremely hazardous. Although average mercury levels are relatively low, they are further reduced to compl...
1MB Sizes 2 Downloads 11 Views