Journal of Hazardous Materials 284 (2014) 1–9

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Performance of magnetic activated carbon composite as peroxymonosulfate activator and regenerable adsorbent via sulfate radical-mediated oxidation processes Wen-Da Oh a,b , Shun-Kuang Lua a,c , Zhili Dong a,c , Teik-Thye Lim a,b,∗ a Nanyang Environment and Water Research Institute (NEWRI), Interdisciplinary Graduate School, Nanyang Technological University, 1 Cleantech Loop, CleanTech One Singapore, 637141, Singapore b Division of Environmental and Water Resources Engineering, School of Civil and Environmental Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore c School of Materials Science and Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

• Magnetic activated carbon composite (MACC) was synthesized.

• MACC was used as peroxymonosulfate (PMS) activator and regenerable adsorbent. • Regeneration efficiency of MACC was quantified. • Kinetic model for regeneration process was proposed. • MACC surface chemistry change during reaction with PMS was studied.

a r t i c l e

i n f o

Article history: Received 21 August 2014 Received in revised form 22 October 2014 Accepted 29 October 2014 Keywords: Magnetic activated carbon composite Activated carbon regeneration Sulfate radical Peroxymonosulfate Surface chemistry

a b s t r a c t Magnetic activated carbon composite (CuFe2 O4 /AC, MACC) was prepared by a co-precipitation– calcination method. The MACC consisted of porous micro-particle morphology with homogeneously distributed CuFe2 O4 and possessed high magnetic saturation moment (8.1 emu g−1 ). The performance of MACC was evaluated as catalyst and regenerable adsorbent via peroxymonosulfate (PMS, Oxone® ) activation for methylene blue (MB) removal. Optimum CuFe2 O4 /AC w/w ratio was 1:1.5 giving excellent performance and can be reused for at least 3 cycles. The presence of common inorganic ions, namely Cl− and NO3 − did not exert significant influence on MB degradation but humic acid decreased the MB degradation rate. As a regenerable adsorbent, negligible difference in regeneration efficiency was observed when a higher Oxone® dosage was employed but a better efficiency was obtained at a lower MACC loading. The factors hindering complete MACC regeneration are MB adsorption irreversibility and AC surface modification by PMS making it less favorable for subsequent MB adsorption. With an additional mild heat treatment (150 ◦ C) after regeneration, 82% of the active sites were successfully regenerated. A kinetic model incorporating simultaneous first-order desorption, second-order adsorption and pseudo-first order degradation processes was numerically-solved to describe the rate of regeneration. The regeneration rate increased linearly with increasing Oxone® :MACC ratio. The MACC could potentially serve as a catalyst for PMS activation and regenerable adsorbent. © 2014 Elsevier B.V. All rights reserved.

∗ Corresponding author at: Division of Environmental and Water Resources Engineering, School of Civil and Environmental Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore. Tel.: +65 6790 6933; fax: +65 6791 0676. E-mail address: [email protected] (T.-T. Lim). http://dx.doi.org/10.1016/j.jhazmat.2014.10.042 0304-3894/© 2014 Elsevier B.V. All rights reserved.

2

W.-D. Oh et al. / Journal of Hazardous Materials 284 (2014) 1–9

1. Introduction Water pollution due to dyes is inevitable as dyes are extensively used in textile, leather, food and pharmaceutical industries [1]. Many dyes are persistent in the environment (e.g.: reactive blue 19, t1/2 = 46 y) and toxic (LD50 value of greater than 2 × 103 mg kg−1 ) [2,3]. In some cases, the degradation products of dye are more toxic than the parent compound. Various treatment methods were employed to treat dye-contaminated water such as adsorption, biological treatment and membrane separation process [3–5]. Some of the disadvantages of these processes are (i) production of secondary waste, (ii) requirement of acclimated sludge and (iii) membrane fouling and management of membrane reject. Sulfate radical-based advanced oxidation process (SR-AOP) emerged as an eco-friendly and effective method for treating dyes in water. An efficient way to generate SO4 •− is by transition metal activation of peroxymonosulfate (PMS, HSO5 − ). To date, various heterogeneous catalysts were prepared for PMS activation which can be generally grouped into metal oxide (e.g.: Co3 O4 , CuFe2 O4 , Fe2 O3 etc.) and supported metal oxide (e.g.: Co3 O4 /MnO2 , MnFe2 O4 /graphene) catalysts [6–12]. Among the metal oxide catalysts, CuFe2 O4 is an exceptional PMS activator with magnetic property [13]. The performance of CuFe2 O4 can be enhanced by supporting it with activated carbon (AC). AC is regarded as a good support for metal oxide due to its remarkable properties with abundant functional groups. AC is a quintessential adsorbent widely used for removing a myriad of organic pollutants including dyes. AC has high surface area and when deposited with catalyst, it is able to enhance pollutant removal and degradation through synergistic adsorption–catalysis processes. To date, limited work was conducted to investigate the role of AC as metal oxide support for PMS activation. In a previous study, Shukla et al. impregnated cobalt ions on an activated carbon for phenol removal via PMS activation but the composite still suffered from cobalt leaching [14]. The AC supported CuFe2 O4 composite has the potential to be utilized as regenerable adsorbent. One of the greatest challenges of employing adsorption process for pollutant removal is the difficulty in regeneration of the adsorbent. Various regeneration methods are available such as thermal, biological and chemical regeneration methods. Thermal regeneration employs heat treatment to decompose the adsorbed pollutant but it is not eco-sustainable due to its exceptionally high carbon footprint [15]. Biological regeneration by microorganisms is an eco-friendly approach to replenish the active sites of AC but with the disadvantages of (i) requiring acclimatized biomass and optimum environment (e.g.: pH, nutrient, temperature), (ii) deactivation of active sites due to biofilm formation and (iii) relatively slower than other regeneration methods [16–18]. Chemical regeneration method employing AOPs (i.e. regeneration mediated by • OH and SO4 •− ) has the advantages of being eco-friendly and faster. Previously, • OH has been employed for regenerating AC through electro-Fenton, Fenton and photocatalysis processes [19–21]. The disadvantages of • OH-mediated regeneration include the short lifetime of • OH and limited light penetration into the pores in photocatalysis process. Compared to the hydroxyl radical (• OH), SO4 •− is longer-lived and can diffuse further to utilize the reactive radicals for oxidation and regeneration processes. To the best of our knowledge, there is no reported study on the regeneration of AC using sulfate radicals generated by transition metal activation of PMS. Herein, magnetic activated carbon composites (CuFe2 O4 /AC, MACCs) were prepared using a facile co-precipitation method followed by calcination. The MACC is magnetic and can be easily separated by using a simple magnetic separation technology. The performance of MACCs were evaluated as PMS activator for pollutant removal from water and as regenerable adsorbent via SO4 •− -mediated oxidation processes using methylene blue (MB)

as the model pollutant. A kinetic model with simultaneous firstorder desorption, second-order adsorption and pseudo-first order MB degradation was employed to describe the rate of MACC regeneration by monitoring the changes in the residual MB concentration in the aqueous phase. The effects of PMS on the surface chemistry of AC and MACC were also evaluated. To the best of our knowledge, no study was conducted on the possible AC surface modification by PMS after oxidation reaction which is imperative in view of high oxidation potential of PMS (E0 = 1.85 V). 2. Experimental 2.1. Chemicals The chemicals used in this study were Cu(NO3 )2 ·3H2 O (QrëcÔ), Fe(NO3 )3 ·9H2 O (Merck), activated carbon (Sigma–Aldrich, Darco® G-60, −100 mesh, powder), NaOH (Alfa Aesar), HCl (37%, Schedelco), Oxone® (2KHSO5 ·KHSO4 ·K2 SO4 , Alfa Aesar), KI (Fisons), TM KCl (Merck), NaCl (Qrëc ), NaNO3 (Sigma–Aldrich), humic acid (Aldrich) and methylene blue (J.T. Baker). All the chemicals are of analytical grade. All the experiments were conducted using deionized water (18.2 M cm). 2.2. Preparation of MACCs The MACCs were prepared via a co-precipitation method followed by calcination. In a typical preparation, 2 mmol of Cu(NO3 )2 ·3H2 O and 4 mmol of Fe(NO3 )3 ·9H2 O were dissolved in 50 mL of deionized water. AC (0.48, 0.72 and 0.96 g for CuFe2 O4 :AC w/w ratios of 1:1, 1:1.5 and 1:2, respectively) was added into the solution followed by rapid stirring with a magnetic stirrer for 1 h to homogenize the solution. Then, the pH of the resulting solution was adjusted to pH 10–11 using 2 M NaOH. The solution was gradually heated to 95–100 ◦ C and maintained for 4 h under vigorous stirring. The resultant product was filtered, washed with deionized water and dried in an oven at 60 ◦ C for 12 h. Finally, the dried product was calcined in a muffle furnace at 300 ◦ C for 1 h to enhance the crystallinity and stability of the impregnated CuFe2 O4 . The MACCs with CuFe2 O4 :AC w/w ratios of 1:1, 1:1.5 and 1:2 were denoted as 1-MACC, 1.5-MACC and 2-MACC, respectively. 2.3. Characterization of MACC The crystal structure of MACCs was determined using X-ray diffractometer (Bruker AXS D8 Advance) at the scan rate of 0.02◦ s−1 operating under 40 kV and 40 mA using Cu-K␣ ( = 1.5418 Å) as the X-ray source. The morphology and distribution of CuFe2 O4 on AC were investigated by obtaining the secondary electron images (SEIs), backscattered electron images (BSIs) and elemental maps using field emission scanning electron microscopy (FESEM, JEOL JSM-7600F) equipped with energy dispersive X-ray spectroscope (EDS, Oxford Xmax80 LN2 Free). The total acidity of AC and 1.5-MACC was determined using the indirect Boehm titration method (Section 2.5). The Brunauer–Emmett–Teller (B–E–T) surface area, pore volume and pore radius were determined using N2 adsorption–desorption isotherm analysis at 77 K (Quantachrome Autosorb-1 Analyzer). The surface functional groups of AC and MACC were characterized using Fourier transform infrared spectrometer (FTIR, PerkinElmer Spectrum GX). The magnetism of MACC was determined using vibrating sample magnetometer (VSM Lakeshore 7400). 2.4. Performance study Prior to the performance study, the adsorption capacities of AC and MACC were determined by agitating 100 mg of the adsorbent

W.-D. Oh et al. / Journal of Hazardous Materials 284 (2014) 1–9

with 500 mL of 500 mg L−1 of MB for 1 h at pH 5 using a magnetic stirrer with stirring speed of 600 rpm. Despite being magnetic, the stirring speed was sufficient to fully disperse all the MACC particles in the solution. The adsorption condition was optimum (pH and contact time) to establish the equilibrium and fully saturating the adsorbent. The MB concentration was determined from a calibration curve by using a UV–vis spectrophotometer (Shimadzu UV-1800) at  = 664 nm. The amount of MB adsorbed (QMB , mg g−1 ) was calculated as follows: QMB =

(Ci − Ce )V mA

where mA (g) is the mass of adsorbent, Ci (mg L−1 ) and Ce (mg L−1 ) are the initial and equilibrium concentrations of MB, respectively, and V (L) is the volume used. Batch study was conducted to study the performance of MACC as a catalyst for MB treatment via PMS activation. In a typical procedure, 200 mL of 20 mg L−1 MB solution was prepared. Then, 0.4 g of Oxone® , corresponding to Oxone® dosage of 2.0 g L−1 , was dissolved in the MB solution and the pH of the solution was adjusted to five using 1 M NaOH. Then, 40 mg of MACC was introduced into the solution under rapid stirring and the reaction was immediately commenced. The MB concentration was monitored at various time intervals. At t = 120 min, the dissolved organic carbon (DOC) of the sample was analyzed using a TOC analyzer (Shimadzu TOC analyzer) and the PMS concentration was quantified from a calibration curve using iodometric method with the aid of a spectrophotometer at  = 352 nm. ICP-OES (PerkinElmer, Elmer Optima 2000DV) was also employed to determine Cu2+ leaching from MACC during the catalytic reaction. The influence of matrix species was also investigated to evaluating the effects of organic and inorganic ions, namely Cl− (100 mg L−1 ), NO3 − (10 mg L−1 ) and humic acid (5 mg L−1 ) on MB degradation. For MACC regeneration study (as a regenerable adsorbent), the MACC with the loading of 1.0 g L−1 was initially saturated using the adsorption protocol as described above. It was then separated from the loading solution using a simple magnetic separation technique. A solution containing 2.0 g L−1 of Oxone® (corresponded to 2 g Oxone® g−1 MACC) was prepared and the pH was adjusted to pH 5 using 1 M NaOH. The MB-loaded MACC was immersed into the 100-mL Oxone® solution and then was agitated with a magnetic stirrer at 600 rpm. After 24 h, the regenerated MACC was recovered using magnetic separation technique and washed several times with deionized water. The adsorption capacity of the regenerated MACC was determined by carrying out the adsorption protocol as described above and the regeneration efficiency (RE) was calculated by comparing the adsorption capacity of fresh and regenerated MACC as follows: RE(%) =

QMBr × 100% QMBf

50 mL of 0.01 M NaOH and then sonicated for 15 min. The solution was left without stirring for 24 h at 25 ◦ C. Then, the solution was filtered using CA filter and 5 mL of the filtrate was measured and diluted to 50 mL. The diluted filtrate was titrated against 0.001 M HCl in 0.01 M KCl. The pH change during the titration was measured for every 1 mL of titrant added using a pH meter. The experiments were conducted in triplicates. The total acidity of AC/MACC was calculated as follows: Total acidity(

(1)

3

mol 10−5 mol V2 )=( ) × VN − V3 ) )×( ) × (( g W mL VN

(3)

where W is the weight of AC/MACC and V2 , VN and V3 are the volume of titrant required for the control NaOH, initial volume of NaOH and volume of titrant required for AC/MACC, respectively. All the experiments were conducted in triplicate. 3. Results and discussions 3.1. Characterization of MACC Fig. 1 shows the XRD patterns of pristine AC and MACCs. The XRD pattern of the pristine AC shows a typical amorphous micrographitic structure with some diffraction peaks from impurities phases [23]. In MACCs, two distinctive peaks close to 2 = 31◦ and 36◦ give evidence of the presence of CuFe2 O4 spinel with some impurity consisting of monoclinic CuO (2 = 36◦ , 39◦ and 48◦ ). The broad XRD peaks suggest that the CuFe2 O4 was of nanosize particles. A sharp peak at 2 = 26◦ for 1-MACC is due to impurities from AC. The calcination step has a profound impact on the stability and crystallinity of MACC. Without calcination step, the XRD peaks are of relatively lower intensity indicating lower crystallinity and possible existence of amorphous phase. The SEIs and elemental mappings for pristine AC and 1.5-MACC shown in Fig. 2(a) indicate that Cu and Fe were successfully impregnated homogeneously on the AC. The morphologies of AC consisting of micro-particles remained unaffected by the surface deposition of CuFe2 O4 . Fig. 2(b) shows the BSIs of 1.5-MACC. From the backscattered electron imaging (Fig. 2(c)) wherein regions containing heavier elements (i.e. Cu and Fe) backscattered electrons stronger than lighter elements (i.e. C and O), and therefore, will appear brighter, the homogeneous distribution of CuFe2 O4 on AC is

(2)

where QMBf and QMBr are the adsorption capacities of fresh and regenerated MACC, respectively. The experiments were repeated for 1.5-MACC by using (i) MACC with different w/w ratios of CuFe2 O4 :AC, (ii) different Oxone® dosages (4, 8 and 13 g Oxone® g−1 MACC) and (iii) different 1.5-MACC loadings. During the regeneration process, residual MB concentration was also monitored. An additional mild heat treatment of 150 ◦ C for 1 h was also employed for selected conditions after the PMS regeneration process. 2.5. Total acidity determination The total acidity of AC/MACC was quantified using indirect Boehm titration as described by Hanelt et al. with slight modification [22]. Briefly, exactly 100 mg of AC/MACC was immersed in

Fig. 1. XRD patterns of pristine AC, uncalcined 1.5-MACC, calcined 2-MACC, calcined 1.5-MACC and calcined 1-MACC. The (# ) indicates CuFe2 O4 while (*) indicates CuO.

4

W.-D. Oh et al. / Journal of Hazardous Materials 284 (2014) 1–9

Fig. 2. SEIs and elemental mappings of (a) pristine AC and (b) 1.5-MACC and (c) BSIs of 1.5-MACC.

10

5 Moment (emu/g)

further confirmed. The micro-analytical EDX results showed Cu:Fe compositional ratio of 1:1.8 which was very close to the theoretical value of 1:2. Slightly higher Cu:Fe ratio was due to the presence of CuO impurity. Indirect Boehm titration method was used to characterize the total acidity of pristine AC and 1.5-MACC. The results showed that the total acidity of 1.5-MACC (0.012 mmol g−1 ) was notably lower than the pristine AC (0.045 mmol g−1 ). A plausible explanation is, during the synthesis process, precursors consisting of Fe3+ and Cu2+ cations were electrostatically-attracted and adsorbed to the negatively-charged acidic functional groups which served as a site for nucleation of CuFe2 O4 . The acidic functional groups, namely carboxylic and lactones, are known to have promotional effect on metal cation adsorption [24]. This promoted homogeneous interstitial growth of CuFe2 O4 on AC as observed in the elemental mapping (Fig. 2(b)) at the expense of the acidic functional groups. The surface area (SB–E–T ) and total pore volume (Table 1) decreased from 786 to 429 m2 g−1 and 0.32 to 0.23 cm3 g−1 , respectively, corresponding to the decrease of CuFe2 O4 :AC w/w ratio but with only marginal increase in the average pore size. This indicated that the CuFe2 O4 was mainly deposited on the larger pores (i.e. meso- and macro-pores) and on the surface of AC. By increasing CuFe2 O4 :AC ratio, the CuFe2 O4 crystallite size increased while lowering the SB–E–T and total pore volume. The magnetic saturation moment per unit mass for 1.5-MACC was determined to be 8.1 emu g−1 (Fig. 3) compared with 0.1 emu g−1 for pristine AC attributed to the presence of well-crystalline magnetic CuFe2 O4 in 1.5-MACC. The presented magnetism value was higher the samples prepared by Zhang et al. (0.97–5.93 emu g−1 ) which were prepared without additional calcination step [25].

0 1.5-MACC saturaon magnezaon = 8.1 emu g-1

-5

-10 -15000

-10000

-5000

0 Field (G)

5000

10000

15000

Fig. 3. M-H hysteresis loop of 1.5-MACC measured at T = 300 K.

3.2. Performance as PMS activator The processes governing MB removal are simultaneous adsorption and SO4 •− -mediated oxidation reactions. The impregnated CuFe2 O4 in MACC acts as catalyst sites to generate highly-reactive • SO4 − from PMS for MB oxidation based on the following proposed mechanisms [12,13,26]: Cu(II) − OH + HSO5 − → −

Cu(II) (HO)OSO3 →

Cu(II) (HO)OSO3 − + OH−

Cu(III) − OH + SO4

Cu(III) − OH + HSO5 − →

Cu(II)

•−

• OOSO − 3

+ H2 O

(4) (5) (6)

W.-D. Oh et al. / Journal of Hazardous Materials 284 (2014) 1–9

5

Table 1 Textural characteristics and MB adsorption capacities of pristine AC and MACCs. Composites

Oxone® dosage (g L−1 )

SB–E–T (m2 g−1 )

Total pore volume(cm3 g−1 )

Average pore radius(Å)

Adsorption capacity(mg g−1 )

Pristine AC 2-MACC 1.5-MACC 1-MACC

– – – –

786 607 556 429

0.32 0.28 0.27 0.23

17.9 18.1 18.0 18.1

166 143 122 105

Pristine AC Pristine AC 1.5-MACC 1.5-MACC Used 1.5-MACC

0.5 2.0 0.5 2.0 2.0

774 760 555 556 327

0.32 0.31 0.27 0.25 0.22

17.9 18.0 18.0 18.0 17.9

– – – 102 ± 6 –

• OOSO − 3

+ 2H2 O

→ 2 Cu(II) − OH + O2 + 2SO4 •− + 2H+

(7)

Cu(II) + HSO5 − →

Cu(I) + SO5 •− + H+

(8)

Cu(I) + HSO5 − →

Cu(II) + SO4 •− + OH−

(9)

Cu(I) + Fe(III) →

Cu(II) + Fe(II)

−

Fe(III) + HSO5 →

Fe(II) + SO5

Fe(II) + HSO5 − →

Fe(III) + SO4

(10) +

•−

+H

•−

+ OH−

MB concentration (C/Co)

1.0

2 Cu(II)

(11)

The time-course of MB removal (Fig. 4(a)) indicated that the optimum CuFe2 O4 :AC ratio was 1:1.5 with no severe compromise to the AC adsorption capacity (∼25% decreased) and giving the fastest rate of removal with pseudo-first order rate constant (kapp ) of 0.0121 ± 0.006 min−1 . Poorer performance of 1-MACC was observed despite having higher CuFe2 O4 loading which can be attributed to its lower surface area for adsorption and accumulation of large CuFe2 O4 crystals in the pores making it inaccessible for catalysis reactions. At least 84% of DOC and 50% of PMS were consumed after 2 h of reaction with MACC indicating that the system is highly efficient in treating MB. As the main fraction of MB was mostly adsorbed, DOC removal was higher for PAC + Oxone® and 2-MACC + Oxone® but it comes with the lowest PMS consumption (Fig. 4). The 1.5-MACC gave the best performance among all the composites studied and it was selected for further scrutiny. In the pristine AC without Oxone® study, the amount adsorbed for the first two cycles were 93 mg g−1 and 32 mg g−1 , respectively, indicating saturation of active sites. The reusability study in Fig. 4(b) showed only infinitesimal decline in the performance of 1.5-MACC over three successive cycles of MB removal suggesting that the occupied active sites of 1.5-MACC can be partially regenerated by oxidation of the adsorbed MB and its intermediates via SO4 •− mediated reactions. The small decline in performance over three cycles could be attributed to the (i) partial deactivation of catalyst’s surface due to obstruction by MB and its intermediates and (ii) exhaustion of AC. The 1.5-MACC can be easily restored to its original performance after three cycles by moderate heat treatment at 150 ◦ C for 1 h. The Cu2+ leaching data was obtained for 1.5-MACC at pH 5 using ICP-OES. The results show low Cu2+ leaching (0.68 mg L−1 , ∼0.85%) which was below the US EPA limit of 1.3 mg L−1 for drinking water suggesting that the MACC is stable. Fig. 5 shows the effects of several common matrix species, namely Cl− , NO3 − and humic acid on MB degradation. The concentrations of the common matrix species were selected to simulate

MB concentration (C/Co)

(14)

3 2 2 3

k’app (min-1)

R2

0.081±0.002 0.098±0.006 0.121±0.006 0.091±0.003 0.092±0.002

0.79 0.98 0.99 0.91 0.88

PAC 1-MACC 1.5-MACC 2-MACC PAC-oxone

(b) Cycle

0.8

1 2 3 After heat treatment

0.6

DOC 2h (%) 85 83 76 83

PMS 2h (%) 54 47 43 45

k’app (min-1)

R2

0.121±0.006 0.086±0.002 0.064±0.002 0.103±0.006

0.99 0.98 0.96 0.99

Run 1 Run 2 Run 3 After heat treatment

0.4 0.2 0.0 0

10

20

30 Time, min

40

50

60

Fig. 4. Time courses of residual MB concentration with different CuFe2 O4 /AC ratios (a) and repeated runs using 1.5-MACC (b). Inset shows DOC 2 h, PMS 2 h and pseudofirst order rate constant (kapp ). Conditions: initial MB concentration = 20 mg L−1 , initial Oxone® dosage = 2.0 g L−1 , initial catalyst loading = 0.2 g L−1 and initial pH 5.

1.0 MB concentration (C/Co)

AC + HSO5 − → AC − H + SO5 •−

PMS 2h (%) 55 54 50 36

0.2

1.0

(13)

DOC 2h (%) 84 85 86 86

0.4

AC is also an electron transfer catalyst and contains functional • groups that can activate PMS producing SO4 − as follows [4,14,27]: AC − OH + HSO5 → AC + SO4

PAC 1-MACC 1.5-MACC 2-MACC PAC-Oxone

0.6

0.0

•−

Composite

0.8

(12)

−

(a)

± ± ± ±

Species Deionized water ClNO3Humic acid All

0.8 0.6

k’app (min-1) 0.121±0.006 0.114±0.005 0.122±0.005 0.103±0.001 0.098±0.001

R2 0.99 0.96 0.98 0.98 0.99

100 mg/L chloride 10 mg/L nitrate 5 mg/L humic acid Deionized water Nitrate, chloride and humic acid

0.4 0.2 0.0 0

10

20

30 40 Time (min)

50

60

Fig. 5. Effects of Cl− , NO3 − and humic acid on MB degradation at various time intervals. Conditions: initial MB concentration = 20 mg L−1 , initial Oxone® dosage = 2.0 g L−1 , initial catalyst loading = 0.2 g L−1 and initial pH 5.

6

W.-D. Oh et al. / Journal of Hazardous Materials 284 (2014) 1–9

12

(a)

(b)

14 Total acid groups, 105 mol g-1

11 10

pH

9 8

PAC PAC 0.5Ox PAC 2Ox 1.5-MACC 1.5-MACC 0.5Ox 1.5-MACC 2Ox 1.5-MACC used NaOH

7 6 5

12.7

12 10.7

10 8 5.8

6

5.2

4.5

3.7

4 2

1.2

0

4 0

10

20

30 40 mL HCl

50

60

Fig. 6. (a) Indirect Boehm titration curves and (b) the calculated total acidity of pristine AC and 1.5-ACC before and after contacted with Oxone® at various dosages.

the condition in real water which could affect the MB degradation. The results showed comparable removal efficiency for all cases in 60 min indicating that the presence of these matrix species has negligible effect on the MB degradation. It was previously reported that Cl− is a strong SO4 •− scavenger with pseudo-second order rate constant, ks = 2.8 × 108 L mol−1 s−1 [28]. However, in this study, no significant Cl− scavenging effect was observed attributed to the synergistic adsorption–catalysis reaction which could minimize the influence of the radical scavengers. The MACC could concentrate the MB but not inorganic scavengers on the surface of the catalyst. In this regard, the MB can react more readily with the generated SO4 •− without significant influence by the radical scavengers. The remarkable role of AC in abating the detrimental effect of radical scavengers was also highlighted by Yap et al. [29]. The pseudo-first order rate constant also showed comparable value to that obtained in deionized water except for a small decrease (∼17%) when humic acid is present. This is anticipated as humic acid can compete with MB for adsorption and catalysis reaction [30]. 3.3. Influence of PMS on surface chemistry of AC/MACC Due to the high oxidation potential of PMS (E0 = 1.85 V), it can induce surface modification of AC. The surface chemistry of AC is crucial for the adsorption process. The effect of PMS on the surface chemistry of AC and 1.5-MACC was evaluated from the changes in total acidity and surface functionalities of AC and 1.5-MACC before and after being contacted with various concentrations of Oxone® (0–2.0 g L−1 ) for 24 h. The AC surface constitutes many surface oxygen-containing functional groups such as carboxyl, carbonyl, quinones, lactones and phenols [31]. In the presence of PMS these surface functional groups can be oxidized to mainly carboxylic acid, increasing the total acidity and oxygen content of AC. The textural characteristics (Table 1) of the AC/MACC remained unchanged after contact with 0.5 and 2.0 g L−1 of oxone. Anions such as PMS are hardly adsorbed by activated carbon due to their poor surface interaction with activated carbon, and therefore, they exert negligible influence on the pore system. Fig. 6 shows the amount of total acidity of AC/MACC determined using the indirect Boehm method before and after 24 h of contact with Oxone® at various dosages. When contacted with oxone, the total acidity of AC/MACC increased proportionally with the initial Oxone® concentration. The total acidities of oxone-treated MACCs

Fig. 7. FTIR spectra of (a) pristine AC and (b) 1.5-MACC before and after contacted with Oxone® at various dosages.

were relatively higher compared with their AC counterparts which was probably due to higher degree of surface oxidation in the MACC by SO4 •− generated by PMS activation. The FTIR spectra of AC and 1.5-MACC before and after contact with various Oxone® dosages are presented in Fig. 7(a) and (b), respectively, and each of the absorption peaks are assigned to their corresponding functional groups in Table 2. In AC and 1.5-MACC, the evidence of increase

W.-D. Oh et al. / Journal of Hazardous Materials 284 (2014) 1–9

7

Table 2 Assignments of characteristic absorption bands of FTIR spectra of AC and 1.5-MACC. Absorption bands (cm−1 ) 3400 1500–1600 1400 1120–1200 600

Functional groups OH stretching C O stretching,C C stretching, OH binding OH angular deformation, COO stretching C O stretching, C O C stretching, C O stretching, Fe O

in carboxylic functionalities appears at ∼1400 cm−1 . The intensity of the absorbance bands increases proportional with increasing Oxone® dosage confirming the surface oxidation reactions and the indirect Boehm titration data. Concomitant with the increase in carboxylic moieties, the absorbance bands at 1550 cm−1 decrease slightly in intensity suggesting the oxidation of surface C C to epoxides and carboxylic acids [32,33]. The characteristic band at 3400 cm−1 , which is present in all the spectra, is indicative of the stretching vibration of OH bonds. The OH moiety on AC support is known to have positive effect on the PMS activation process [6]. The presence of Fe O bond of CuFe2 O4 in 1.5-MACC is vindicated by the presence of the 600 cm−1 band. The influence of surface modification of AC on MB adsorption was further investigated. It is known that two types of mechanisms are responsible for dye adsorption, namely electrostatic interaction involving charge attraction and dispersive interaction involving ␲electrons. Since MB is a cationic dye, the increase of total acidity (mainly from carboxylic acid) is anticipated to have a positive effect on adsorption due to the good anchoring ability of carboxylic group promoting electrostatic interactions [34,35]. However, this was not observed and the adsorption capacity of 1.5-MACC decreased from 122 ± 2 to 102 ± 6 mg g−1 when it was contacted with 2.0 g L−1 of Oxone® for 24 h. This could be because the increase of the acidic functional groups was at the expenses of other functional groups that have a high affinity for MB. For instance, the lost of aromaticity in AC as the result of oxidation by SO4 •− decreased the – dispersive interactions. This could be an important factor that hindered complete recovery of AC active sites. 3.4. Performance as regenerable adsorbent The feasibility of 1.5-MACC as MB adsorbent and its regeneration by SO4 •− -mediated oxidation reactions was evaluated. The regeneration of the active sites was conducted by loading the MACC to saturation followed by the regeneration step using oxone.

References [38–40] [39,41,42] [38,43,13] [38,41,44,45] [13]

The regeneration efficiency as defined by Eq. (2) was 32–37% for all MACCs (Table 3) under the same condition. Under different 1.5-MACC loadings, the regeneration efficiency decreased with increasing loading probably due to the limitation of oxone. Since for complete MB mineralization of 122 mg MB g−1 MACC, stoichiometric amount of PMS required is ∼13 g Oxone® g−1 MACC (mol ratio of MB:PMS = 1:108), various Oxone® dosages consisting of 4, 8 and 13 g Oxone® g−1 MACC were also employed. Almost similar efficiency of only 35–40% of the active sites of 1.5-MACC was regenerated for 4, 8 and 13 g Oxone® g−1 MACC. The factors contributed to the hindrance of complete regeneration are (i) adsorption hysteresis, pore constriction and pore blockage by MB and (ii) destruction of AC functional groups responsible for MB adsorption by PMS oxidation (Section 3.3). Pore blockage especially in the micropores (diameter =