RESEARCH ARTICLE

A sulfur-resistant CuS-modified active coke for mercury removalfrom municipal solid waste incineration flue gas

Wei Liu1& Yongxian Zhou2

& Yinfeng Hua3 & Bin Peng4& Mei Deng2

& Naiqiang Yan1& Zan Qu1

Received: 16 January 2019 /Accepted: 30 May 2019 /Published online: 25 June 2019# Springer-Verlag GmbH Germany, part of Springer Nature 2019

AbstractAdsorption is a typical method for air pollutant removal from flue gas. A CuS-modified active coke (CuS/AC) sorbent wasdeveloped to improve the elemental mercury removal efficiency from municipal solid waste incineration (MSWI) flue gas. Theinfluences of the loading amount of CuS, reaction temperature, and flue gas components including O2, SO2, H2O, and HCl onHg0 removal efficiency were investigated, respectively. The results showed that the mercury adsorption capacity of CuS/AC(20%)

sorbent was about 7.17mg/gwith 50% breakthrough threshold, which is much higher than that of virgin active coke. The analysisof XPS indicated that HgS was the main species of mercury on spent CuS/AC, which implied that adsorption and oxidation wereboth included in Hg0 removal. S2

2− played a vital role in the oxidation of physically adsorbed Hg0. Meanwhile, the commoncomponents of MSWI flue gas exhibited no significant inhibition effect on Hg0 removal by CuS/AC sorbent. CuS/AC sorbent isa promising sorbent for the mercury removal from MSWI flue gas.

Keywords Mercury removal . Active coke . Sulfur resistance .MSWI flue gas

Introduction

Mercury (Hg) is a global pollution for its high toxicity, bioac-cumulation, and long-range transport potential. In order toreduce mercury pollution, the Minamata Convention onMercury, the first international convention on mercury restric-tion, has come into force in 2017. It is well known that coalcombustion and non-ferrous metal smelting are the major an-thropogenic mercury emission sources (Pacyna et al. 2010).However, according to global mercury assessment in 2018,

municipal solid waste incineration (MSWI) contributes toabout 7.27% of the global total anthropogenic mercury emis-sions (UNEP 2018). Therefore, it is also one of the five majormercury emission sources in the Minamata Convention(Zhang et al. 2015). In China, about 32.5% of municipal solidwaste was treated by incineration (Hong et al. 2017). With therapid expansion of theMSWI industry, it should be consideredan important source in mercury control strategy.

The MSWI temperature is about 850–1200 °C. Almost allof the mercury releases into the flue gas in the form of Hg0

during the incineration process. Then, part of the Hg0 convertsto oxidized mercury (Hg2+) and particulate-bounded mercury(Hgp) (Cheng & Hu 2010). Generally, the typical pollutants ofMSWI flue gas, such as NOx, acidic gases, dioxins, and par-ticles are removed by SNCR system, dry/semidry scrubbingsystem, activated carbon injection (ACI) system, fabric bagfilter, and activated carbon absorption tank, respectively (Liet al. 2017). These air pollutant control devices could alsoremove over 60% mercury from MSWI flue gas (Cheng andHu, 2012, Zhang et al. 2008). Besides, wet electrostatic pre-cipitators (WESP) are installed downstream of the wet flue gasdesulfurization (FGD) to control the emissions of ultrafineparticulates and gypsum rain in some factories. In this unit,sorbents can be injected into the flue gas and spent sorbentscan be collected by WESP. Among them, ACI technology is

Responsible editor: Philippe Garrigues

Electronic supplementary material The online version of this article(https://doi.org/10.1007/s11356-019-05645-6) contains supplementarymaterial, which is available to authorized users.

* Zan Ququzan@sjtu.edu.cn

1 School of Environmental Science and Engineering, Shanghai JiaoTong University, Shanghai 200240, China

2 CSSC Nanjing Luzhou Environmental Protection Co., Ltd,Nanjing 210039, China

3 Shanghai Liming Resource Reuse Co., Ltd, Shanghai 201209, China4 Pudong DistrictWasteManagement Center, Shanghai 200135, China

Environmental Science and Pollution Research (2019) 26:24831–24839https://doi.org/10.1007/s11356-019-05645-6

an effective method for mercury removal although it is mainlyused to capture dioxins from flue gas (Li et al. 2018a).Nevertheless, some disadvantages of ACI technology, suchas poor capacity and high cost, had limited its application(Tan et al. 2012). Moreover, the MSWI flue gas is complicat-ed. It always contains a high concentration of SO2, which mayinhibit the mercury adsorption performance of activated car-bon (Liu 2008). Thus, improving the sulfur-resistant and mer-cury adsorption ability of activated carbon is significant formercury removal from MSWI flue gas.

As a carbon-based product, activated coke is comprisedmainly of macropores and mesopores instead of micropores,which reduces significantly the impact of internal diffusion onthe general rate of adsorption processes (Du et al. 2018).Accordingly, activated coke has attracted lots of attention forthe removal of SO2 from flue gas (Li et al. 2010, Yuan et al.2018, Zhang et al. 2017). Meanwhile, many active compo-nents have been used to modify activated coke for removingHg0 from flue gas, including sulfur, halogens, and metal ox-ides (Chen et al. 2018, Liu et al., 2000, Xie et al. 2015, Zenget al. 2004), while most of these modified sorbents were stillpoisoned by SO2. Recently, the fast reaction rate and highmercury adsorption performance of metal sulfide sorbent forthe removal Hg0 from coal-fired flue gas was reported (Liet al. 2016, Li et al., 2018a, b, Yang et al. 2018, Liu et al.2019). Moreover, these sulfide materials also showed goodmercury removal efficiency in the presence of SO2.Therefore, metal sulfide-modified activated coke maybe agood choice for mercury removal from MSWI flue gas.

In this work, a CuS-modified activated coke (CuS/AC)adsorbent was developed to remove Hg0 under SO2 at lowtemperature. Effects of reaction temperature, loading amount,and flue gas component were investigated. Besides, the pos-sible mechanisms of Hg0 adsorption and oxidation were iden-tified. The results show that CuS/AC is an effective and prom-ising sorbent for Hg0 capture from MSWI flue gas.

Materials and methods

Preparation of sorbents

All chemicals applied in this study were analytical grade. Thechemicals copper nitrate [Cu(NO)2·3H2O] and sodium sulfide[Na2S·9H2O] were obtained from Sinopharm ChemicalReagent Co., Ltd. Activated cokes used in this research werecommercial cokes from Inner Mongolia Kexing CarbonIndustry Co., Ltd. The samples were prepared through theprecipitation method as follows. Firstly, the ACwith a particlesize of 40–60 mesh was washed with deionized water anddried in an electric blast oven at 101 °C for 12 h. Secondly,1 g AC was impregnated with a certain amount of cupricnitrate solution for 12 h. Thirdly, equal molar ratio of sodium

sulfide was added to the mixed solution under vigorous stir-ring for 1 h. Fourthly, the mixture was transferred to a rotaryevaporation flask with water at 60 °C, and rotary evaporationwas performed for approximately 2 h. Finally, the sample(CuS/AC) was washed by deionized water and dried undervacuum at 65 °C overnight to get the powder sorbent.Different mass ratios of CuS/AC sorbents were prepared;mass ratios of CuS to AC were 2%, 5%, 10%, 20%, respec-tively. And these sorbents were marked as CuS/AC(2%), CuS/AC(5%), CuS/AC(10%), and CuS/AC(20%), respectively.

Characterization of adsorbents

The crystal-phase structures of sorbents were determined byX-ray diffraction (XRD, APLX-DUO, BRUKER, Germany)with Cu-Kα radiation. The XRD spectra were recorded from10 to 80° (2θ) at a scanning rate of 10°/min operating at 40 kVand 30 mA. To measure the real loading ratio of CuS, thesample was dissolved in concentrated HNO3 solution andwas analyzed by inductively coupled plasma mass spectrom-etry (ICP-MS) (Agilent, ICP-OES 5110).

The Brunauer-Emmett-Teller (BET) surface area of thesamples was measured using a N2 sorption apparatus(NOVA 2200e, Boynton Beach, FL). Prior to BET measure-ments, all the samples were degassed at 110 °C for 4 h. Thepore volume and pore diameter were calculated based on theBarrett-Joyner-Halenda (BJH) method.

To identify the active groups and mercury species on sor-bents, X-ray photoelectron spectroscopy (XPS) was carriedout on an Ultra DLD (Shimadzu-Kratos, Japan) spectrometerwith Al Kα as the excitation source. The observed spectrawere adjusted with the C 1s binding energy (BE) value of284.6 eV. Morphology and structure of materials were ob-served using a scanning electron microscope (Nova NanoSEM450) at 10 kV.

Mercury temperature-programed desorption (Hg-TPD) ex-periment was carried out on a fixed-bed experimental system.The desorbed mercury was detected by a cold vapor atomicabsorption spectroscopy (CVASS) analyzer (SG921, JiangfenChina).

Hg0 adsorption experiments

As shown in Fig. 1, the Hg0 adsorption experiments werecarried out in a fixed-bed reactor system. The experimentalsystem was made up of gas supply system, Hg0 vapor gener-ating device, H2O vapor generating device, fixed bed reactor,mercury analyzer, and exhaust gas treatment device. Hg0 con-centration was maintained at about 1.4 mg/m3 by adjusting theN2 flow rate (100 mL/min) passing through the mercury per-meation tube (placed in a water bath with a temperature of 45°C). The sorbents were loaded in a 6-mm diameter quartztube. Fifty milligrams of sorbent was tested in each

24832 Environ Sci Pollut Res (2019) 26:24831–24839

experiment. The reaction temperature was controlled by a tu-bular furnace and the experiments were performed at 25–150°C. The flow rate of simulated flue gas was controlled bymassflow controllers (MFC) and total flow rate was maintained at800 mL/min. Hg0 concentration in inlet and outlet gases wasdetected and analyzed by a cold vapor atomic absorption spec-troscopy (CVASS) analyzer. The CVASS analyzer was cali-brated by a special mercury analyzer (Lumex RA915+,Russia). The mercury adsorption capacity (Q) was calculatedaccording to Eq. 1:

Q ¼ 1

m∫t0 Ci−Coð Þ � f � dt ð1Þ

where Ci and Co are the inlet and outlet Hg0 concentration(mg/m3) respectively, m is the mass of sample (g), f denotesthe flow rate of the influent, and t is the adsorption time (min).

Results and discussion

Characterization of adsorbents

As shown in Table 1, the ICP results indicated that the actual loadof the sample we synthesized was close to the theoretical load.The BET surface area, pore volume, and pore diameter of virginAC and CuS/AC are summarized in Table 2. As shown inTable 1, CuS loading obviously affected the physical propertiesof AC. The AC impregnated with CuS posesses higher BETsurface area, except for CuS/AC(20%). Sa smilar trend was found

in other literatures (Hua et al. 2010). In the process of impregna-tion, some new pores might be generated due to the reactionbetween AC and CuS components, which account for the higherBET surface area of impregnated AC compared with virgin AC.Nevertheless, the pore diameter and pore volume of impregnatedAC reduced with the increase of CuS loading amount because ofits deposition in some pores of AC (Serrano-Ruiz et al. 2008).

Figure 2 shows the XRD patterns of virgin AC and CuS/AC sorbents. Three strong diffraction peaks attributed to ACwere detected at 26.67°, 29.15°, and 43.61°, respectively (Xieet al. 2015). Nevertheless, the intensity of these peaks de-creased with the increase of CuS loading amount. Two strongpeaks attributed to CuS were detected at 12.51° and 26.12°when CuS content is over 5%. The SEM images of virgin ACand CuS/AC are shown in Fig. 3. It can be seen that thesurface characteristics of AC changed because CuS depositedon the surface of AC. The surface of AC was covered andpores of AC were blocked remarkably when the loadingamount of CuS exceeded 5%.

Hg0 adsorption tests

The mercury removal performances of different CuS/AC sor-bents were tested to investigate the influence of CuS loadingamount on mercury adsorption. As shown in Fig. 4, the Hg0

removal efficiency was only about 54.4% when virgin ACwas tested. Meanwhile, the virgin AC could hardly removeHg0 after 60 min of adsorption, while the Hg0 removal perfor-mance of AC increased greatly after it was modified by CuS.About 99% Hg0 could be removed by CuS/AC(2%) sorbent.Moreover, the high Hg0 removal efficiency could last about

Fig. 1 Schematic diagram of thefixed-bed experimental system

Environ Sci Pollut Res (2019) 26:24831–24839 24833

200 min when CuS/AC(20%) sorbent was tested. According tothe mercury adsorption amount of different sorbents(Figure S1), the Hg0 adsorption saturated amount of virginAC was only about 0.30 mg/g. And the Hg0 adsorptionamount increased from 1.62 to 8.78 mg/g when the CuS load-ing amount increased from 2 to 20% after 10 h of adsorption.Obviously, CuS/AC sorbent showed better Hg0 adsorptionrate and capacity than virgin AC. And CuS has a key role inthe Hg0 adsorption process. The mercury adsorption capacityof CuS/AC(20%) sorbent was about 7.17mg/gwith 50% break-through threshold, which was better than that of reportedmod-ified activated carbon (Table S1) (Liu et al., 2000, Xiao et al.2017, Shen et al. 2015, Hua et al. 2010). However, the utili-zation rate of CuS/AC(10%) sorbent was better than that ofCuS/AC(20%) sorbent if we compare with the ratio of theHg0 adsorption amount and CuS loading amount of both sor-bents. Thus, it is not necessary to increase the CuS loadingamount on AC unlimitedly. The proper CuS loading amounton AC was about 10%.

Influence of adsorption temperature

The influences of reaction temperature on Hg0 removal byCuS/AC(10%) sorbent was investigated over a temperaturerange from 25 to 150 °C. From Fig. 5, it could be seen thatCuS/AC(10%) sorbent exhibited an excellent Hg

0 removal per-formance at low temperatures (25–75 °C). When the temper-ature increased from 25 to 50 °C, the mercury adsorptionamount increased from 4.86 to 5.87 mg/g. However, the mer-cury adsorption amount decreased to 5.18 mg/g when thetemperature continuously increased to 75 °C. Then, the mer-cury adsorption amount decreased to 0.27 mg/g rapidly withthe reaction temperature continuing to increase to 150 °C. In a

typical MSWI flue gas cleaning systems, an activated carbonadsorption tank was installed after the wet scrubber devices.The flue gas temperature of activated carbon adsorption tankwas about 40–60 °C. It is a suitable place for mercury removalby CuS/AC sorbent.

Influences of flue gas components

Generally, the common components of flue gas, such as O2,SO2, and H2O will affect the mercury adsorption performanceof sorbent. In order to evaluate the mercury adsorption perfor-mance in real MSWI flue gas, the influnces of flue gas com-ponents on the removal of Hg0 were investigated, respectively.

Themercury adsorption amount of CuS/AC(10%) sorbent after10 h of adsorption under pure N2 was about 5.82 mg/g. It did notchange greatly when the concentration of O2 increased from 0 to10%O2. This indicated that oxygen has no obvious influence onmercury removal by CuS/AC(10%) sorbent. The influences ofSO2 on Hg0 removal over different adsorbents in the flue gaswere inconclusive, including promotional, inhibitive, and insig-nificant effects (Liu et al., 2000, Suresh Kumar Reddy et al.2013). As shown in Fig. 6, when 500 ppm and 1000 ppm ofSO2were added to the simulated flue gas, themercury adsorptionamount was 5.91 mg/g and 5.72 mg/g, respectively. Thus, SO2

exhibited almost no impact on Hg0 removal by CuS/AC sorbentin this study, while H2O (vapor) has a negative impact on

Table 1 CuS contents of CuS/AC composites determined by ICPmethods

Samples Measured CuS loading(wt.%)

Theoretical CuS loading(wt.%)

2% CuS/AC 1.9 2

5% CuS/AC 3.8 5

10% CuS/AC 9.4 10

20% CuS/AC 17.2 20

Table 2 The BET surface area,pore volume, and pore diameterof the sorbents

Materials BET surface area (m2/g) Total pore volume (cm3/g) Average pore diameter (nm)

Virgin AC 208.14 0.0437 5.792

CuS/AC(2%) 325.81 0.0519 5.806

CuS/AC(5%) 296.04 0.0487 5.785

CuS/AC(10%) 219.69 0.0416 5.728

CuS/AC(20%) 107.89 0.0314 5.739

10 20 30 40 50 60 70 80

CuS CuS/AC(20%)

CuS/AC(10%)

CuS/AC(5%)

CuS/AC(2%)

2 theta (degree)

Virgin AC

AC

Fig. 2 XRD patterns of virgin AC and CuS/AC sorbents

24834 Environ Sci Pollut Res (2019) 26:24831–24839

mercury removal over metal oxide-based sorbents because of thecompetitive adsorption of H2O andHg0 in active adsorption sites(Li et al. 2011, Xie et al. 2015). As shown in Fig. 6, the mercuryadsorption of sorbent was close to 5.82 mg/g when the watervapor concentration in simulated flue gaswas increased from0 to10%, respectively. Therefore, H2O also has no influence on themercury adsorption byCuS/AC sorbent. But H2O could promotethe conversion of SO2 to generate H2SO4, which could cover thesurface of the sorbents and affect the adsorption of Hg0. Hence,the mercury adsorption amount of CuS/AC(10%) sorbent wasdecreased to 5.35 mg/g under SFG (1000 ppm SO2 +10%H2O) condition.

HCl is a typical component in MSWI flue gas. The influ-ence of HCl on mercury removal by CuS/AC(10%) sorbent was

also investigated. As shown in Figure S2, the mercury remov-al efficiency and adsorption amount of CuS/AC(10%) sorbentwere both enhanced in the presence of HCl. Thus, the majorcomponents of MSWI flue gas, such as SO2, HCl, and O2 didnot inhibit the mercury adsorption by CuS/AC(10%) sorbent.

Hg0 adsorption mechanism

The XPS of O 1s, S 2p, and Hg 4f on the fresh and used CuS/AC sorbents are shown in Fig. 7. As shown in Fig. 7a, thepeak at 531.6 eV was assigned to chemisorbed oxygen andOH group (Xie et al. 2015). For S 2p in Fig. 7b, three types ofsulfur were observed on the sorbents, the peaks at 162.1 eV,164.1 eV, and 169.5 eV could be regarded as S2−, S2

2−and

Environ Sci Pollut Res (2019) 26:24831–24839 24835

Fig. 3 SEM images of virgin AC and CuS/AC. a Virgin AC. b CuS/AC(2%). c CuS/AC(5%). d CuS/AC(10%). e CuS/AC(20%)

SO42−, respectively (Liao et al. 2016). After mercury adsorp-

tion under O2 (5%) or SO2 (1000 ppm) at 50 °C, the used CuS/AC samples were analyzed by XPS. As shown in Fig. 7 c andf, no obvious difference was observed on the binding energiesof O 1s in two samples. It indicated that chemisorbed oxygenand OH group was not the active site for Hg0 adsorption. For S2p in Fig. 7d, three peaks at 162.3 eV, 163.8 eV, and 169.5 eVwere assigned to S2−, S2

2−, and SO42−respectively. As the

analysis in Table 3, the ratios of S22− were decreased to

44.67% and the ratios of S2− were increased to 9.88% aftermercury adsorption. The results could be inferred that part ofS2

2− was oxidized to S2−. However, the ratio of SO42− in-

creased from 45.45 to 61.05% in the presence of SO2. Onthe basis of the results and previous research (Yan et al.2013), the possible reaction can be described as follows:

SO2 in the flue gas is adsorbed on the surface over AC, andthen, SO2 is oxidized by active oxygen to form SO3, whichcan subsequently combine with H2O to form H2SO4 on thesurface of the AC.

SO2 þ C−O→C−SO3 ð2ÞC−SO3 þ H2O→Cþ H2SO4 ð3Þ

As shown in Fig. 7e, two peaks centered at 101.0 eV and105.1 eV were corresponding to HgS. Another bindingenergy peak at 102.4 eV was assigned to the Si 2p (Liaoet al. 2016, Xu et al. 2017). Mercury primarily existed asHgS on the surface of CuS/AC sorbents. To further iden-tify the mercury species, Hg-TPD experiment was con-ducted after Hg0 adsorption under 5% O2. The used sor-bents were heated from 25 to 700 °C with the heatingrate of 2 °C/min under the protection of N2. From thedesorption curve of used virgin AC (Fig. 8), it could beseen that mercury began to be released at 150 °C and twoobvious peaks appeared at around 200 and 280 °C, re-spectively. According to the desorption temperatures fordifferent mercury compounds and previous literatures, thetwo peaks were due to HgO (Wu et al. 2011). The mech-anisms of Hg0 removal over virgin AC was adsorbed Hg0

reacted with chemisorbed oxygen or OH groups to formHgO. However, the adsorption of mercury onto CuS/ACwas more stable than that onto virgin AC. The adsorbedmercury began to release from the used CuS/AC at ap-proximately 200 °C and had a maximum desorption rateat 260 °C. At such a narrow desorption temperature

0

1

2

3

4

5

6

SFG (1

0% H

2O

)

10%

H2O

5% H

2O

5% O

2+10

00pp

m S

O 2

5% O

2+50

0ppm

SO 2

10%

O2

5% O

2

)g/

gm(

ytica

pac

noit

pros

da

yrucre

M

N 2

Fig. 6 Influences of flue gas components on Hg0 removal over CuS/AC.Reaction conditions: 50 mg CuS/AC(10%) sorbent, ~ 1.4 mg/m3 Hg0, N2

as balance. GHSV = 670,000 h−1, T = 50 °C. The adsorption reaction timewas 10 h

0

1

2

3

4

5

6

1501251007550

)g/

gm(

tn

uo

mA

noit

pros

dA

yrucre

M

Reaction Temperature (oC)

25

Fig. 5 Influences of reaction temperature on mercury adsorption amountby CuS/AC sorbent. Reaction conditions: 50 mg CuS/AC(10%) sorbent,5% O2, ~ 1.4 mg/m3 Hg0, N2 as balance. GHSV = 670,000 h−1, T = 25–150 °C. The adsorption reaction time was 10 h

0 100 200 300 400 500 6000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4g

H0

noitartnecnoc(

m/gm

3 )

Time (min)

AC

CuS/AC(2%)

CuS/AC(5%)

CuS/AC(10%)

CuS/AC(20%)

50% Breakthrough

Fig. 4 Mercury adsorption curve of virgin AC and CuS/AC sorbents.Reaction conditions: 50 mg sorbent, 5% O2, ~ 1.4 mg/m3 Hg0, N2 asbalance. GHSV = 670,000 h−1, T = 50 °C. The adsorption reaction timewas 10 h

24836 Environ Sci Pollut Res (2019) 26:24831–24839

range, mercury mainly existed in the HgS, which wascorresponding to the results of XPS analysis (Rumayoret al. 2015).

Based on above discussion and analysis, Hg0 removalmechanism can be described as follows: at first, gaseousHg0 was physically adsorbed on the surface of CuS/AC.

545 540 535 530 525 520

531.6

Blinding energy (eV)

O 1s

fresh CuS/AC

180 175 170 165 160

Blinding energy (eV)

162.1169.5

164.4 S 2p

fresh CuS/AC

545 540 535 530 525 520

531.4

Blinding energy, eV

O 1S

after Hg0 adsorption

180 175 170 165 160

Blinding energy (eV)

162.3

169.5

163.8 S 2p

after Hg0 adsorption

110 105 100 95

Blinding energy, eV

102.4

105.1

101.0

Hg 4f

after Hg0 adsorption

545 540 535 530 525 520

531.5

Blinding energy, eV

O 1S

after Hg0 adsorption

under SO2

a b

c d

e f

180 175 170 165 160

Blinding energy (eV)

164.4

168.8

162.1

g S 2p

after Hg0 adsorption

under SO2

Fig. 7 XPS spectra of fresh CuS/AC and used CuS/AC

Environ Sci Pollut Res (2019) 26:24831–24839 24837

Then, the adsorbed Hg0 reacted with the surface S22−and

formed HgS on its surface. The reaction can be described asfollows:

Hg0g þ AC→Hg0ad−AC ð4ÞHg0ad þ S2

2−→HgSþ S2− ð5Þ

Conclusion

In this work, CuS supported on commercial activated cokeswere prepared and applied to remove Hg0 from MSWI fluegas. CuS/AC sorbents showed better mercury removal effi-ciency and adsorption capacity than virgin AC. The optimalreaction temperature was 50 °C. Meanwhile, there was noobvious inhibitory impact on Hg0 removal in the presence ofthe common MSWI flue gas components, such as SO2, H2O,O2, and HCl. In summary, CuS/AC sorbent is a promisingsorbent for the Hg0 removal from MSWI flue gas.

Supporting information

Influences of CuS loading amount on Hg0 adsorption amountby CuS/AC sorbents are shown in Figure S1. Influences ofHCl on Hg0 adsorption efficiency and amount by CuS/AC

sorbents are presented in Figure S2. Comparative summaryof mercury sorbents is listed in Table S1.

Funding information This study was financially supported by theNational Key Research and Development Program of China (No.2017YFC0806300) and the National Natural Science Foundation ofChina (No. 21677096).

Compliance with ethical standards

Competing interests The authors declare that they have no competinginterests.

References

Pacyna EG, Pacyna JM, Sundseth K, Munthe J, Kindbom K, Wilson S,Steenhuisen F, Maxson P (2010) Global emission of mercury to theatmosphere from anthropogenic sources in 2005 and projections to2020. Atmos Environ 44:2487–2499

UNEP (2018): global mercury assessment 2018Zhang L,Wang S,Wang L,WuY, Duan L,WuQ,Wang F, YangM,Yang

H, Hao J, Liu X (2015) Updated emission inventories for speciatedatmospheric mercury from anthropogenic sources in China. EnvironSci Technol 49:3185–3194

Hong J, Chen Y, Wang M, Ye L, Qi C, Yuan H, Zheng T, Li X (2017)Intensification of municipal solid waste disposal in China. RenewSustain Energy Rev 69:168–176

Cheng H, Hu Y (2010) China needs to control mercury emissions frommunicipal solid waste (MSW) incineration. Environ Sci Technol 44:7994–7995

Li G, Wu Q, Wang S, Li Z, Liang H, Tang Y, Zhao M, Chen L, Liu K,Wang F (2017) The influence of flue gas components and activatedcarbon injection on mercury capture of municipal solid waste incin-eration in China. Chem Eng J 326:561–569

Cheng H, Hu Y (2012) Mercury in municipal solid waste in China and itscontrol: a review. Environ Sci Technol 46:593–605

Zhang H, He PJ, Shao LM (2008) Fate of heavy metals during municipalsolid waste incineration in Shanghai. J Hazard Mater 156:365–373

Li G, Wu Q, Wang S, Duan Z, Su H, Zhang L, Li Z, Tang Y, Zhao M,Chen L (2018a) Improving flue gas mercury removal in waste in-cinerators by optimization of carbon injection rate. Environ SciTechnol 52

Tan Z, Sun L, Xiang J, Zeng H, Liu Z, Hu S, Qiu J (2012) Gas-phaseelemental mercury removal by novel carbon-based sorbents. Carbon50:362–371

Liu Y (2008) Impact of sulfur oxides on mercury capture by activatedcarbon. Environ Sci Technol 42:6579–6584

Du X, Li C, Zhao L, Zhang J, Gao L, Sheng J, Yi Y, Chen J, Zeng G(2018) Promotional removal of HCHO from simulated flue gas overMn-Fe oxides modified activated coke. Appl Catal B Environ 232:37–48

Li LT, Xie W, Liang DM, Sun ZC, Xu ZG (2010) Mechanism of removalof SO2 and NO on activated coke. Environ Sci Technol 33:79–83

Yuan J, Jiang X, ZouMJ, Yao L, Zhang CJ, JiangWJ (2018) Copper ore-modified activated coke: highly efficient and regenerable catalystsfor the removal of SO2. Ind Eng Chem Res 57:15731–15739

Zhang K, He Y, Wang ZH, Huang TH, Li Q, Kumar S, Cen KF (2017)Multi-stage semi-coke activation for the removal of SO2 and NO.Fuel 210:738–747

Chen Y, Guo X, Wu F, Huang Y, Yin Z (2018) Experimental and theo-retical studies for the mechanism of mercury oxidation over chlorine

0 100 200 300 400 500 600 700

lan

giS

yrucre

M

Temperature (oC)

Virgin AC(200 mg)

CuS/AC(10 mg)

Fig. 8 Hg-TPD spectrum of used virgin AC and CuS/AC sorbents

Table 3 Mole distribution of sulfur species in different sorbents

Samples SO42−(%) S2

2−(%) S2−(%)

Fresh CuS/AC 43.18 50.82 6.01

After Hg0 adsorption 45.45 44.67 9.88

After Hg0 adsorption under SO2 61.05 30.69 8.27

24838 Environ Sci Pollut Res (2019) 26:24831–24839

and cupric impregnated activated carbon. Appl Surf Sci 458:790–799

Liu W, Vidic RD, Brown TD (2000) Impact of flue gas conditions onmercury uptake by sulfur-impregnated activated carbon. EnvironSci Technol 34:154–159

Xie Y, Li C, Zhao L, Zhang J, Zeng G, Zhang X, ZhangW, Tao S (2015)Experimental study on Hg0 removal from flue gas over columnarMnOx-CeO2/activated coke. Appl Surf Sci 333:59–67

Zeng H, Jin F, Guo J (2004) Removal of elemental mercury from coalcombustion flue gas by chloride-impregnated activated carbon. Fuel83:143–146

Li H, Zhu L, Wang J, Li L, Shih K (2016) Development of nano-sulfidesorbent for efficient removal of elemental mercury from coal com-bustion fuel gas. Environ Sci Technol 50:9551–9557

Li H, ZhuW,Yang J, ZhangM, Zhao J, QuW (2018b) Sulfur abundant S/FeS 2 for efficient removal of mercury from coal-fired power plants.Fuel 232:476–484

Yang Z, Li H, Feng S, Li P, Liao C, Liu X, Zhao J, Yang J, Lee PH, Shih K(2018) Multiform sulfur adsorption centers and copper-terminatedactive sites of Nano-CuS for efficient elemental mercury capturefrom coal combustion flue gas. Langmuir 34:8739–8749

Liu H, You Z, Yang S, Liu C, Xie X, Xiang K, Wang X, Yan X (2019)High-efficient adsorption and removal of elemental mercury fromsmelting flue gas by cobalt sulfide. Environ Sci Pollut Res Int

Hua XY, Zhou JS, Li Q, Luo ZY, K-f C (2010) Gas-phase elementalmercury removal by CeO2 impregnated activated coke. EnergyFuel 24:5426–5431

Serrano-Ruiz JC, Ramos-Fernández EV, Silvestre-Albero J, Sepúlveda-Escribano A, Rodríguez-Reinoso F (2008) Preparation and charac-terization of CeO2 highly dispersed on activated carbon. Mater ResBull 43:1850–1857

Shen B, Li G, Wang F, Wang Y, He C, Zhang M, Singh S (2015)Elemental mercury removal by the modified bio-char from medici-nal residues. Chem Eng J 272:28–37

Xiao Y, Pudasainee D, Gupta R, Xu Z, Diao Y (2017) Bromination ofpetroleum coke for elemental mercury capture. J Hazard Mater 336:232–239

Suresh Kumar Reddy K, Al Shoaibi A, Srinivasakannan C (2013) ele-mental mercury adsorption on sulfur-impregnated porous carbon – areview. Environ Technol 35:18–26

Li H, Wu CY, Li Y, Zhang J (2011) CeO2-TiO2 catalysts for catalyticoxidation of elemental mercury in low-rank coal combustion fluegas. Environ Sci Technol 45:7394–7400

Yan Z, Liu L, Zhang Y, Liang J, Wang J, Zhang Z, Wang X (2013)Activated semi-coke in SO2 removal from flue gas: selection ofactivation methodology and desulfurization mechanism study.Energy Fuel 27:3080–3089

LiaoY, Chen D, Zou S, Xiong S, XiaoX,DangH, Chen T, Yang S (2016)Recyclable naturally derived magnetic pyrrhotite for elemental mer-cury recovery from flue gas. Environ Sci Technol 50:10562–10569

XuH, Yuan Y, Liao Y, Xie J, Qu Z, ShangguanW, Yan N (2017): [MoS4)Cluster bridges in Co-Fe layered double hydroxides for mercuryuptake from S-Hg mixed flue gas. Environ Sci Technol 51(2):10109–10116

Wu S, Uddin MA, Nagano S, Ozaki M, Sasaoka E (2011) Fundamentalstudy on decomposition characteristics of mercury compounds oversolid powder by temperature-programmed decomposition desorp-tion mass spectrometry. Energy Fuel 25:144–153

Rumayor M, Fernandez-Miranda N, Lopez-Anton MA, Diaz-SomoanoM, Martinez-Tarazona MR (2015) Application of mercury temper-ature programmed desorption (HgTPD) to ascertain mercury/charinteractions. Fuel Process Technol 132:9–14

Publisher’s note Springer Nature remains neutral with regard tojurisdictional claims in published maps and institutional affiliations.

Environ Sci Pollut Res (2019) 26:24831–24839 24839