Journal of Colloid and Interface Science 453 (2015) 159–168

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

In-situ deposition of silver�iron oxide nanoparticles on the surfaceof fly ash for water purification

http://dx.doi.org/10.1016/j.jcis.2015.04.0440021-9797/� 2015 Elsevier Inc. All rights reserved.

⇑ Corresponding authors at: Department of Bionanosystem Engineering, Gradu-ate School, Chonbuk National University, Jeonju 561-756, Republic of Korea (C.S.Kim). Fax: +82 63 270 2460.

E-mail addresses: khanjoo@jbnu.ac.kr (H.J. Kim), biochan@jbnu.ac.kr (C.H. Park),chskim@jbnu.ac.kr (C.S. Kim).

Mahesh Kumar Joshi a,b, Hem Raj Pant a,c, Nina Liao a, Jun Hee Kim a, Han Joo Kim a,⇑, Chan Hee Park a,⇑,Cheol Sang Kim a,d,e,⇑a Department of Bionanosystem Engineering, Graduate School, Chonbuk National University, Jeonju 561-756, Republic of Koreab Department of Chemistry, Tri-Chandra Multiple Campus, Tribhuvan University, Kathmandu, Nepalc Department of Engineering Science and Humanities, Institute of Engineering, Pulchowk Campus, Tribhuvan University, Kathmandu, Nepald Division of Mechanical Design Engineering, School of Engineering, Chonbuk National University, Jeonju 561-756, Republic of Koreae Eco-friendly Machine Parts Design Research Center, Chonbuk National University, Jeonju 561-756, Republic of Korea

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 31 January 2015Accepted 21 April 2015Available online 5 May 2015

Keywords:Fly ashComposite particleLead(II) ionsAntibacterial activity

a b s t r a c t

In this study, a fly ash based composite, Ag–iron oxide/fly ash, was synthesized via a facile one-pothydrothermal process using fly ash, ferrous chloride, and silver nitrate as precursors. Field emission scan-ning electron microscopy (FE-SEM), EDX, transmission electron microscopy (TEM), X-ray diffraction(XRD), Fourier transform infra-red spectroscopy (FTIR), Photoluminescence (PL) and Brunauer–Emmett–Teller (BET) surface area measurement confirmed the formation of composite particle. FA provided asuitable surface for the in-situ deposition of Fe3O4 and Ag NPs during hydrothermal treatment. As aresult, the particle size of Fe3O4 and Ag NPs was sufficiently decreased, and the surface area of the NPsas well as, a whole matrix was increased. The antimicrobial activity of the composite was accessed byEscherichia coli inhibition assay. Lead(II) ion adsorption efficiency of the composite was analyzed froma series of batch adsorption experiments (the effects of concentration, contact time, pH and adsorbentdose on the adsorption of Pb(II) ion from aqueous solution). Results indicated that as-synthesized com-posite has high antibacterial capacity, and the metal ions uptake efficiency compared to fly ash particle.Furthermore, incorporation Fe3O4 NPs onto the fly ash make it easily separable from a reaction system

160 M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168

using an external magnet. The composite synthesis protocol is a simple method that utilizes a readilyavailable industrial byproduct to produce a unique composite for environmental remediation.

� 2015 Elsevier Inc. All rights reserved.

1. Introduction

In recent years, environmental contamination caused by toxicmetal ions has become a worldwide environmental problem dueto the bioaccumulation tendency of these toxic materials [1–3].Removal of heavy metal from aqueous solution have been a focusarea of research over recent years [4]. Various methods have beenused to remove heavy metal ions from aqueous environments suchas chemical precipitation [5,6], solvent extraction [7], ion exchange[8], reverse osmosis and adsorption [4]. Among these processes,adsorption with a selection of suitable adsorbent can be an effec-tive technique for the removal of heavy metal ions from wastewa-ter. Activated carbon, alumina, silica, fly ash, and ferric oxide arecommonly used adsorbents that have high metal adsorption capac-ity. However, most of them are expensive and difficult to separatefrom wastewater after water treatment. Therefore, a low cost flyash based magnetic composite could be an alternative materialthat could be mechanically removed by magnetic separation.

Fly ash (FA) is a waste byproduct generated from coal-industrialplants and consists of fine, powdery particles with a sphericalshape, lower density and substantial granular components of sili-con dioxide (SiO2), alumina (Al2O3), calcium oxide (CaO), and mag-nesium oxide (MgO) as explained in our previous report [9]. Everyyear large quantity of FA is produced as a byproduct from thermalplant, worldwide. However, only a small amount (10–20%) of thiswaste fly ash is reused [10]. In recent years, many studies haveinvestigated effective ways to modify FA for valuable applications[11]. It has been shown that FA can be used as an additive in thecement industry, in the building industry for brick making, forglass, lightweight materials, ceramic tableware and art ware, com-posite materials production, materials recovery, as an adsorbentfor waste management and, waste stabilization, as well as otheruses [12]. Furthermore, FA is a potential substrate to incorporatemetal and metal oxide NPs, such as Ag, TiO2, ZnO for various appli-cations [13–15].

Silver nanoparticle based nanocomposites have attractedresearch interest due to a variety of applications including asantimicrobial, catalysts, optoelectronics, sensing, biomedicine,and as an adsorbent. Many studies have established thechemisorption of heavy metal ions on nanoparticles [16], and theinteraction of silver nanoparticles (NPs) with mercuric and leadions have been reported [16–19]. However, the use of such com-posites is limited due to difficulties associated with separationafter their use. Such a drawback could be solved by doping mag-netic NPs onto the surface of FA particles that could be separatedusing the magnetic field. This would be beneficial in a situationwhere filtration is not a simple option. This process would alleviatedifficulties associated with separating nanoparticles dispersed inmedia directly for purification purposes [20]. Furthermore, flyash is a potential support for deposition of different inorganicNPs on its surface [13,14,21,22]. Therefore, the simultaneousgrowth of silver and magnetic iron oxide NPs on the surface offly ash could provide an effectively bonded composite. The fixedattachment of iron oxide and Ag NPs on the surface of high aspectratio FA, could be useful not only for separation but also act as anantibacterial agent, and assists in the adsorption of aqueous pollu-tants during water treatment. Therefore, the deposition of Ag andFe3O4 NPs on fly ash could be an efficient way to construct a com-posite with improved properties required for water treatment. Iron

oxide nanoparticles deposited on the fly ash surface not onlyimproves the separation performance but also effective for heavymetal ion adsorption. Kumari et al. [23] synthesized the magneti-cally separable magnetite nanospheres using template freesolvothermal method for the effective removal of chromium andlead from aqueous solution. Recently, Abdullah and coworkers[24] reported that the Ag and Fe NPs have excellent efficiency forthe adsorption/removal of Cr(II) and Pb(II) ions from the aqueoussolutions. The removal/uptake mechanism involves the interactionbetween the metal ions and the oxide/hydroxyl layer around thespherical metallic core of nanoparticle in water medium [24].

The aim of the present work is to investigate an efficient,low-cost and easily separable composite for the simultaneousremoval of the Pb(II) ions and microbial bodies from aqueous solu-tion. This simple approach highlights the possibility of using acomposite for waste-water purification, where one byproductmaterial, fly ash, was used to control other pollutants in a scalableand inexpensive process.

2. Experimental

2.1. Materials

Fly ash obtained from Won Engineering Company Ltd. (Gunsan,Korea) was ball-milled for 12 h before use. Ferrous chloridetetrahydrate (FeCl2�4H2O) from Samchun Chemicals, ammoniumhydroxide (NH4OH, 28% NH3 in water) and silver nitrate (AgNO3)from Showa chemicals, poly (vinylpyrrolidone) (PVP, MW-5800)from Alfa Aesar, and lead(II) nitrate (Pb(NO3)2) from SigmaAldrich were used as received.

2.2. Preparation of the Ag–iron oxide/Fly ash composite

Ag–iron oxide/fly ash, a fly ash based composite (FAC) particleswere synthesized using a one-step hydrothermal approach basedon our previous reports [25]. In brief, 400 mg of ball-milled FAPs(using 3 mm zirconia balls for 12 h and sieved) was washed andsonicated for 30 min in 20 ml distilled water. 40 mg of PVP wasadded to a 20 ml aqueous solution containing 200 mg ofFeCl2�4H2O, 1 ml of ammonium hydroxide (28%) and sonicatedfly ash suspension was added to this mixture followed by shakingat a rate of 200 rpm for 45 min. After shaking, 10 ml of4.07 � 10�2 M AgNO3 was added, and the mixture was transferredto an autoclave for hydrothermal treatment at 120 �C for 3 h. Theobtained composite was washed several times with distilled waterand ethanol, and was dried at 80 �C for 12 h.

2.3. Characterization

The surface morphologies of fly ash and the as-synthesized par-ticles were studied by field-emission scanning electron microscopy(FE-SEM, S-7400, Hitachi, Japan) and transmission electron micro-scopy (TEM-2010, Orius SC10002). EDX was also performed usingFE-SEM. Information about the phase and crystallinity of the mate-rial was obtained using Rigaku X ray diffractometer (XRD, Rigaku,Japan) with Cu Ka (k = 1.540 Å) radiation over Bragg angles rangingfrom 10� to 80�. Nitrogen adsorption/desorption isotherms wereobtained at the liquid nitrogen temperature of 77 K using aQuantachrome Nova 2200e automated gas adsorption system.

M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168 161

The specific surface area was determined using multipoint BETanalysis and pore size was measured by the BJH method of adsorp-tion using ASAP 2010 V5.02H software. Fourier-transform infrared(FTIR) spectra (Paragon 1000 spectrometer, Perkin–Elmer) and thephotoluminescence (PL) spectrum (Perkin–Elmer instruments,LS55) were measured at room temperature. UV–visible spectrawere recorded using UV–visible spectrometer (Lamda 900,Perkin–Elmer) and the particle size was determined using ImageJprogram.

2.4. Antibacterial property

The antibacterial activity of different composites was analyzedusing Escherichia coli as a model microorganism. The antibacterialexperiments were conducted in a sterilized 100 ml glass beakercontaining an E. coli suspension (50 ml) and 0.4 g/l of differentsamples following the as-fabricated matrix with magnetic stirring.The initial E. coli concentration was cultivated at approximately107 CFU/ml, which was incubated at 37 �C for 12 h in the darkand the measurement continuously proceeded at room tempera-ture for 12 h. The experiment equipment was set up at a UV lightintensity of 1.1 W/cm2 (only 10%) using a mercury vapor lamp(OmniCure, EXFO), and the different samples (3 � 3 cm) in a beaker(solution) were 5 cm from the tip of a light-guide (a mercury vaporlamp). At specific time intervals, 1 ml of the solution was extractedand immediately spread on Tryptic Soy agar plates. Petrifilm (3 M,USA) and agar plates were prepared to count bacterial colonies.After 24 h of incubation at 37 �C, the number of viable E. coli colo-nies was counted three times using a colony counter.

2.5. Sorption experiments

Batch adsorption experiments were conducted to test the effectof initial concentration of Pb(II) solution, adsorbent dose, pH, andcontact time on adsorption of metal ions. The stock solution of1000 mg/L of Pb(II) was prepared by dissolving an exact amountof lead nitrate in deionized water. To observe the effect of concen-tration of metal ion solution on the adsorption efficiency of adsor-bent, solutions of desired concentrations were prepared by serialdilutions of the stock solution. Each experiment was carried in aseparate conical flask (250 ml) containing 100 ml of Pb(II) ion solu-tion of different concentration, and stirred at 150 rpm with con-stant amount adsorbent for 210 min at 25 �C. To study the effectof pH on adsorption of Pb(II), each flask containing 100 ml ofPb(II) solution of different pH values was studied with constantamount of FA and FAC (0.02 g) for 210 min at 25 �C. Similarly, batchadsorption experiments were conducted for dose of the adsorbent.The residual Pb(II) concentration in solutions were determined byan inductively coupled plasma mass spectrometer (ICP-MS; Agilent7500a, USA). The percentage removal of Pb(II) ion was calculatedfrom the following equation:

R ¼ Co � Ct

Co� 100

where R is the removal efficiency of Pb(II), Co is the initial concen-tration, and Ct is the final concentration of Pb(II) in mg/L and at timet. The adsorption capacity of adsorbent at equilibrium was calcu-lated by the following equation:

qe ¼ðCo � CtÞV

W

where qe is the equilibrium adsorption capacity of adsorbent inmg/g, Co is the initial concentration in mg/L, Ce is the equilibriumconcentration of Pb(II) in mg/L, V is the volume of Pb(II) solutionin liters and W is the weight of adsorbent in gram. Adsorption iso-therms were obtained with different initial concentrations of Pb(II)

solution maintaining a constant dose of FA and FAC and otherparameters. Freundlich and Langmuir models are the most commonisotherms for determining adsorption phenomena. Equilibrium datafor Pb(II) adsorption on adsorbents were applied to the Freundlichand Langmuir equations.

3. Results and discussion

3.1. Characterization of the composite particles

Fly ash is a heterogeneous material consisting of small spheresof irregular, porous, coke-like particles [9]. Surface morphologies ofball-milled FA and FAC were investigated by FE-SEM analysis(Fig. 1). FA particles were found to be agglomerates of irregularshapes and different sizes. Most of these particles had cracks andridges, and non-shaped aggregates heterogeneously dispersed overthe whole matrix as shown in Fig. 1a. FE-SEM images of FAC(Fig. 1c) showed that Ag–iron oxide NPs (reasonably sphericaland uniform in shape and size) were uniformly deposited on thesurface of FA particles. Nevertheless, EDX analysis of FAC(Fig. 1d) showed the presence of silver and iron along with the con-stituent elements of FA, confirming the formation of the compositeparticles. The TEM micrographs of Ag–Fe3O4, FA, and FAC areshown in Fig. 2. Results showed that Ag–Fe3O4 NPs (Fig. 2a) wereextremely small in the range of 10–40 nm with average size of�20 nm in diameter. But FA particles (Fig. 2b) exhibited a hetero-geneous structure containing a number of precipitates with vari-ous morphologies. Furthermore, FAC (Fig. 2c) showed that Ag–iron oxide nanoparticles are homogeneously grown on the surfaceof FA during hydrothermal treatment. The functional site presenton the FA surface may provide sufficient nucleation site for metaland metal oxide NPs growth. Therefore, numerous small size NPs(about 11 nm) were grown on the FA surface (Fig. 2c). It is clearfrom this figure that NPs were eager to accumulate homogeneouslythroughout the FA surface.

The surface areas and pore size distributions of FA and FAC werecharacterized with nitrogen physisorption measurements. Theiradsorption–desorption isotherms for both adsorbents exhibits atype IV isotherm at a relative pressure between 0.2 and 0.9(Fig. 3a and b), characteristics for mesopores with different poresize [26]. Moreover, FA and FAC exhibits H4 and H3 type hysteresisloops, respectively, according to the IUPAC classification. The H3type hysteresis loop for FAC (Fig. 3b) can be attributed to the meso-porous structure formed by the aggregation of Ag–iron oxide NPSon the surface of FA particles. The morphological characterizationof different sample is summarized in Table 1. BET specific surfaceareas of 48.57 m2/g and 25.65 m2/g and pore volumes of0.260 cm3/g and 0.067 cm3/g were calculated for FAC and FA,respectively. The deposition of extremely small NPs (having highsurface area) on the surfaces of FA particle may have increasedthe surface area of the whole matrix.

XRD analysis was performed to determine the crystal structureof the synthesized composite. Fig. 4 shows the XRD pattern of FA,iron oxide/fly ash and FAC. The pristine FA (Fig. 4a) showed a mildhump ranging from 20� to 25� indicating the existence of the amor-phous glass phase alumina and silica [27]. Moreover, the majormineral constituents of fly ash were characterized as SiO2,3Al2O3�2SiO, Ca3Si2O7, Ca9Si6O21�H2O, Ca2SiO4�0.35H2O, and Ti4O7

[28,29]. In addition, FAC (Fig. 4c) showed the presence of Fe3O4

with new peaks at 2 theta = 30.2(220), 35.8(311), 43.5(422),53.7(511) and 62.7(440) [30] and embedded Ag NPs with peaksat 2 theta = 38.2(111), 44.3(200), 64.5(220) and 77.4(311) thatare characteristics of face centered cubic Ag NPs [31].

Additionally, the molecular structures of different samples wereinvestigated by FT-IR spectroscopy. Fig. 5 shows the FT-IR spectra

Fig. 1. FE-SEM images and EDX mapping of FA (a and b), and FAC (c and d).

Fig. 2. TEM images of Ag/iron oxide (a), FA (b), and FAC (c). Histogram shows the size distribution of Ag/Fe3O4 NPs.

162 M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168

of FA, Fe3O4/FA and Ag–Fe3O4/FA particles. The IR bands of pristineFe3O4 (Fig. 5a) were similar to our previous study [31]. The broadpeak near 3400 cm�1 of FA (Fig. 5b) was assigned to the Si–OHbond vibration [32]. The broadening of the IR band at 3400 cm�1

for the FAC (Fig. 5d) is due to the Si–OH bond vibration as wellas moisture absorbed by the composite. Furthermore, theright-shift of the peaks at 3400 cm�1 in the composite was

attributed to interactions between Ag, Fe3O4 and FA throughhydrogen or dative bond as explained by Yeole et al. [13]. The peakintensity of the composite is essentially due to silica and the alu-mina framework of FA at 874, 1005, 1181, 1439, 1617 and2926 cm�1 (Si–O–Si bond, Al–OH bending vibration absorption)relatively decreased compared to pristine FA, suggesting thehomogeneous deposition of Fe3O4 and Ag NPs on the FA surface

Fig. 3. Nitrogen adsorption–desorption isotherms for FA (a) and FAC (b) (inset shows the BJH pore size distribution plot from desorption branch).

Table 1Morphology of different particles.

Sample Size of Ag–Fe3O4

particle (nm)BET surfacearea (m2/g)

Shape of theparticles

Ag/Fe3O4 20 ± 5.98 148.3 SphericalFA n/a 25.6 IrregularFAC 11 ± 3.75 48.5 n/a

M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168 163

throughout the matrix. Nevertheless, no shifts of the FA character-istic bands were observed after loading with Fe3O4 and Ag NPs,indicating the formation of the composite via van der Waals inter-actions. Such interactions are also reported in TiO2-deposited FAmaterials [33]. The major constituents in the IR patterns of thecomposite are essentially due to the silica and alumina framework.Therefore, the vibration of bonds corresponding to other phases(such as Fe3O4 and Ag NPs) overlaps and renders the peak decon-volutions and interpretation of the patterns more difficult in termsof their appearance and/or significance in the resultant compositematrix. However, the typical IR absorption band at 570 cm�1 andanother at 688 cm�1 are assigned to c (Fe–O) stretching vibrationin tetrahedral site and c (Fe–O) torsional vibrational mode of Fe

Fig. 4. XRD patterns of different composites particles: d. Ag NPs, e. Iron oxide, a. SiO2,

in octahedral site of magnetite (Fig. 5, inset) [31,34] is present inFe3O4/FA, and FAC (Fig. 5c and d) but absent in FA (Fig. 5a). Thisresult confirms the presence of Fe3O4 NPs in composite particle.Furthermore, the peak intensity at 688 cm�1 is increased inFe3O4/FA and FAC compared to FA due to the synergetic effect ofiron oxide NPs in the FAC. FA is composed of silica and other semi-conductor materials and doping of Ag can enhance the opticalproperties of the composite particle. The decrease in intensity ofthe NBE emission peak in FAC compared to FA (Fig. S1) indicatesthat the rate of recombination of photogenerated e–h might be lesswhich is indication of the high photocatalytic efficiency [14]. Thisresult suggests that the Ag NPs are homogeneously imprisonedthroughout the matrix enhancing the optical properties of compos-ite particle. To further study the optical properties of compositeparticle, UV–visible absorption spectra of different samples wererecorded and displayed in Fig. 6. The spectrum of pristine FA dis-plays a broad intense absorption in the UV region below 400 nm.But Fe3O4 deposited sample shows a red shift of absorption edgeand significant enhancement of light absorption at 400–800 nm,revealing that the deposition of iron oxide generated morephoto-generated charge compared to pristine FA. As evidencedfrom absorption spectrum of FAC, the incorporation of Ag NPs in

b. CaCl2, c. Ca2SiO4�0.35H2O, d. Ca3Si2O7, e. Ca9Si6O21�H2O, f. 3Al2O3�2SiO2, g. Ti4O7.

Fig. 5. FTIR spectra of different particles.

164 M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168

a composite led to higher absorption in visible range compared topristine FA and Fe3O4/FA composite, which is assumed due to thesurface Plasmon absorption of spatially confined electrons in theAg NPs [35]. The extension of light absorption into the visibleregion is assigned to the synergistic effects of a transition metalion and noble metal NPs [36]. This result revealed that the deposi-tion of Ag and Fe3O4 NPs on the fly ash surface greatly improvedthe optical properties of the fly ash particle.

3.2. Formation mechanism of FAC

As described in experimental section, FAC composite particleswere synthesized using FA particles as support, FeCl2�4H2O asFe2+ ion source and AgNO3 as Ag+ ion source through one pothydrothermal method. Many papers report on the synthesis of fineFe3O4 nanoparticles with 2–10 nm by co-precipitation reaction;generally treating stoichiometric amount of ferrous and ferric saltswith base solution to produces Fe3O4 NPs [37]. However, using aferrous salt alone, the absence of ferric ions calls for different reac-tion mechanism for the formation of Fe3O4 NPs. The formation ofan intermediate species on mixing two aqueous solutions is likelyto be responsible for the formation of Fe3O4 NPs. Refait and Olowereported the following reaction mechanism for the formation ofFe3O4 NPs [38,39].

Fe2þ þ 2OH� ! FeðOHÞ2 ðaÞ3FeðOHÞ2 þ 1=2O2 ! FeðOHÞ2 þ 2FeOOHþH2O ðbÞFeðOHÞ2 þ 2FeOOH! Fe3O4 þ 2H2O ðcÞ

Thus, it is expected that in the synthesis with ferrous ions alone,Fe3O4 NPs are formed as a result of dehydration reaction of ferroushydroxide and ferric oxyhydroxide as shown in Eq. (c), in which

Fig. 6. UV–vis spectra of different particles.

ferric oxyhydroxide is produced by partial oxidation of ferroushydroxide by O2 in dissolved air, according to Eq (b). The ferroushydroxide particles are expected grow in size via hydroxylation(Eq. (a)). Sugimoto and coworkers reported the formation ofFe3O4 NPs with diameter less than 37 nm by oxidation of ferroushydroxide as a precursor [40]. In the present study, it is likely thatthe FA particles present in reaction medium provided a suitablesurface for the deposition of NPs, preventing their agglomerationsand the NPs less than 20 nm are deposited on the FA surface. Onthe other hand, Silver ions form diamine complex [Ag(NH3)2]+ asshown in Eq (d) with ammonia solution. PVP reduced silver ionto silver nanoparticle and also worked as stabilizing agent [41,42].

Agþ þ 2NH4OH! AgNH3ð Þ2� �þ þ 2H2O ðdÞ

AgNH3ð Þ2� �þ !PVP

Ago ðeÞ

Stabilizing agents are required to obtain the stable mono dis-persed nanoparticles. They protect the particles from the aggrega-tion, nanoparticles collisions and coalescence due to the reactionbetween functional groups of the stabilizers and nanoparticle. Anumber of reagents are used as protecting agent to control theshape and size of particles in nanoparticles synthesis. The shapeand size of Ag nanocrystal could be selectively controlled and sta-bilized using capping agents such as sodium citrate [43], sodiumdodecyl sulfate [44] and cetyl trimethyl ammonium bromide[45]. However, the polymers such as gelatin [46], D-sorbitol [47],poly (vinyl pyrrolidone) (PVP) [48,49], and PVA polyvinyl alcohol(PVA) [50] are commonly used as stabilizers and protective agentsin nanoparticle synthesis. Among them, PVP is considered as versa-tile capping agent and stabilizer because of the favorable protect-ing properties owing to its unique structure [48,49,51]. Duringthe NPs synthesis, PVP protects the silver NPs from growing andagglomerating by compounding with it [51]. PVP is a homopoly-mer with a polyvinyl backbone and its repeating unit containspolar amide group and non-polar methylene groups. The N and Oin the polar groups have a strong affinity for silver nanoparticles.Silver particles with diameter shorter than 50 nm are protectedby the coordination between silver and N in PVP, and for the biggerparticles both N and O coordinated with the silver [51]. The PVPmolecules also stabilizes Fe3O4 particles, and gather Ag and Fe3O4

NPs on the FA surface forming FAC composite similar to that forAg–Fe3O4 formation [52]

Therefore, in present study, post-treatment by means of ahydrothermal reaction produced an effectively bonded compositeparticle of Fe3O4, Ag and FA in which later particle provided astrong support for the deposition of former particles. In-situ depo-sition of Fe3O4 and Ag NP on the FA surface prevented the agglom-erations of NPs, and highly effective small NPs with high surfacearea were deposited on the FA surface.

3.3. Antibacterial activity

The antibacterial behavior the present composite was testedwith and without UV light against E. coli by carrying out a zoneinhibition as well as bacterial colony count method. The antibacte-rial efficiency of the FAC was evaluated using mild UV irradiationto the bacterial solution at ambient condition. Fig. 7 shows theantibacterial efficiency of different samples under UV light. Theefficiency of the FAC was considerably high among the samples.The highest antibacterial activity of the as-synthesized compositeunder UV light was attributed to the high photocatalytic behaviorof the composite as explained in Section 3.1. Inset of figure showsthat the fly ash (Fig. 7a) and iron oxide/FA (Fig. 7b) have diminutivezone of inhibition while a larger diameter inhibition zone for FAC(Fig. 7c) shows high efficiency for bacterial deactivation, which is

Fig. 7. Antibacterial activity of different particles and Inset shows the zones of inhibition tests for fly ash (a), Iron oxide/fly ash (b) and FAC (c) against gram-negative E. colibacteria.

M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168 165

attributed to the Ag NPs thoroughly distributed on the compositematrix. However, the exact mechanism of bacterial deactivationis not clear yet, some report have proposed the innate antibacterialactivity of Ag NPs [53]. It has been suggested that silver ions maycause bacteria to reach an active but nonculturable (ABNC) stateand eventually die. Ag NPs can directly damage bacterial cell mem-brane, Ag NPs appears to exert bactericidal activity through therelease of Ag ions which increases membrane permeability includ-ing leakage of cellular content and disruption of DNA replication[54]. These results confirmed the improved antibacterial activityof FAC composite.

3.4. Pb(II) metal ion adsorption

3.4.1. Mechanism for the adsorption of Pb(II) ion by compositeIn aqueous solution, the Pb(II) ions undergoes solvation and

hydrolysis as per the following reactions [55]:

Pb2þ þ nH2O$ PbðH2OÞ2þn

PbðH2OÞ2þn $ Pb H2Oð Þn�1 þHþ

Pb2þ þmH2O$ Pb H2Oð Þð2n�mÞm þmHþ

It has been reported that Pb(OH)2 is the dominant species atpH P 6.5 and Pb2+ and Pb(OH)+ at pH < 6.5 [55,56]. In the mean-time, the surface of the metal oxides in FA or FAC may undergoprotonation/deprotonation by following process [57]:

H2OþM—O� ��! ��Hþ

OH�M—OH ��! ��Hþ

OH�MOHþ2

where H+ and OH� refer to the potential determining ions. The pres-ence of charged surface promotes redistributions of the ions in thesolution, with counter ions preferentially attracted toward the sur-face of the adsorbent. Thus the adsorption of Pb(II) is mainly due tothe coulomb attraction between charged surface and Pb(II) ions. Atlow pH value, excess hydrogen ion competes with the Pb(II) ions forthe adsorption site resulting the lower adsorption of Pb(II). Athigher pH value, high electrostatic attraction exists between nega-tively charged surface of adsorbent and Pb(II) ions; as result theadsorption of Pb(II) ion increases.

3.4.2. Parameters affecting Pb(II) ions adsorptionAdsorbent dose, concentration of metal ion, pH, and contact

time affects the adsorption efficiency of adsorbents. The optimumdose for the adsorption of metal ion is found to be 0.2 g/L (Fig. S2a).

It has been found that the metal ion uptake increase from 25% to94.3%, and 27% to 89. 2% for FAC and FA, respectively, with increas-ing adsorbent dose from 0.05 to 0.3 g/L (Fig. S2a), adsorption sitesincrease with increasing the amount of adsorbent affecting theadsorption [58]. Results of experiments varying the initial Pb(II)ions concentrations (10–200 mg/L) with a 0.002 g per 100 mLadsorbent dose is illustrated in (Fig. S2b). At a low initial solutionconcentration, the surface area and availability of adsorption sitesare relatively high while for a higher concentration, the totaladsorption sites are limited, thus resulting in a decrease in theuptake of metal ion [59]. The adsorption of Pb(II) ion is greatlyaffected by the pH of the solution. The optimum pH for the adsorp-tion of Pb(II) ions is found to be P6.0 for both of adsorbents(Fig. S2c). The higher pH of the solution may result the precipita-tion of lead hydroxide [55], whereas at lower pH of the solution,the positively charged H3O+ ions, and Pb(II) ions may competefor available active sites including hydroxyl groups existing onthe surface of the adsorbents [60]. The adsorption rate was rapidin the initial stage (30 min) due to the availability of large numbersof active sites at the first stage of the adsorption process, while slo-wed down as the equilibrium state was reached (Fig. S2d) whichmight be due to the intra-particle diffusion of metal ions ontothe adsorbent [61]. Furthermore, it has been observed that, theequilibrium is attained faster for FA compared to FAC (Fig. S2d).The deposition of NPs on the surface of fly ash might lead theintra-particle diffusion and may have slowed down the rate ofadsorption for FAC. The high adsorption efficiency of FAC comparedto FA is attributed to the synergetic effect of FA and small sizediron oxide NPs deposited on its surface.

3.4.3. Adsorption isothermsThe adsorption process of Pb(II) onto FAC adsorbent was tested

using Langmuir and Freundlich adsorption isotherms models, twocommonly used empirical adsorption models which correspondsto heterogeneous and homogeneous adsorbent surfaces, respec-tively. The linear form of the Freundlich model is expressed as fol-lows [62]:

log qe ¼1n

log Ce þ log K f

where Kf and 1/n are characteristic constants representing theadsorption capacity and adsorption intensity of the system, respec-tively. The linear plot between logqe versus logCe (Fig. S3a) gave aslope, which is equal to the value of 1/n with logKf intercept

Table 2Adsorption isotherm model parameters for adsorption of Pb(II) ion on FA and FAC at297 K.

Samples Langmuir Freundlich

qmax (mg/g) KL (L/mg) R2 Kf (mg/g) 1/n R2

FAC 526.5 344.83 0.999 335.4 0.092 0.933FA 416.6 138.88 0.9952 184.3 0.177 0.9668

166 M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168

(Table 2). The magnitude of 1/n < 1 indicates the favorability of theprocess of adsorption.

The Langmuir isotherm model assumes maximum adsorptionoccurs when the surface is covered by adsorbate. The isothermcan be represented as follows [63]:

Ce

qe¼ 1

kLþ aL

kLCe

where KL and aL are Langmuir equilibrium constants. TheseLangmuir parameters were obtained from the linear correlationbetween the values of Ce/qe and Ce. The slope and intercept of linearplot of Ce/qe versus Ce (Fig. S3b) gave the value of aL/KL and 1/KL,respectively. The theoretical maximum adsorption capacity (qmax)corresponding to Langmuir constant (Table 2) is numerically equalto KL/aL [64]. It has been observed that the maximum adsorptioncapacity of FAC (526.5 mg/g) was greatly enhanced compared toFA (416.6 mg/g) (Table 2) which is may be due synergetic adsorp-tion efficiency of iron oxide NPs and FA. Langmuir isotherm is usedto access the feasibility of adsorbent and can be expressed in termsof a dimensionless separation factor or equilibrium factor RL, whichis defined by:

RL ¼1

1þ aLCe

The RL values calculated from Langmuir isotherm were found to begreater than zero and less than one, indicating that the composite issuitable for adsorption of Pb(II) ions. Langmuir isotherm (Fig. S3b)generated a satisfactory fit to the experimental data for both theadsorbents FA and FAC compared to Freundlich isotherms. Similarresult is indicated by high regression coefficients shown inTable 2. The fact that the Langmuir isotherm fits the experimentaldata very well may be due to homogeneous distribution of activesites on the FA and FAC surface [65].

Fig. 8. Pseudo-first order adsorption (a), pseudo-second order adsorption (b), and intrsolution and 20 mg/100 mL of adsorbent).

3.4.4. Adsorption kineticsSeveral models have been used to express the mechanism of

solute sorption onto a sorbent. In order to investigate the kineticmechanism for adsorption of Pb(II) ions on FAC, pseudo-first order,pseudo-second order, and intra-particle diffusion were employedto interpret the experimental data [66], a good correlation withthe kinetic data could be used to explain the adsorption mecha-nism of metal ion onto the solid phase [64].

3.4.4.1. Pseudo-first order model. The Pseudo-first order kineticmodel for sorption analysis is generally expressed as follows [67]:

logðqe � qÞ ¼ log qe � ðk1=2:303Þt

where qe is the amount of metal ion adsorbed per unit weight ofadsorbent at equilibrium, i.e. adsorption capacity (mg/g) at any timet, and k1 is the rate constant of first order sorption. The value of k1

was calculated from the slope of the linear plot log (qe � q) verses t(Fig. 8a).

3.4.4.2. Pseudo-second order model. A pseudo-second order rateexpression based on sorption equilibrium capacity may be derived,and if pseudo-second order kinetics holds true, the rate law for thereaction is expressed as follows [67]:

tq¼ 1

k2q2e

� �þ 1

qe

� �t

where k2 is the rate constant of sorption (g/mg min). The value of k2

can be determined from the plot of t/q versus t (Fig. 8b).The kinetic parameters evaluated from the adsorption of Pb(II)

are given Table 3. On the basis of the obtained correlation coeffi-cient, adsorption of Pb(II) onto the FAC composite perfectly fitthe pseudo-second order model. In addition, the theoretical valuesof qe obtained from the pseudo-second order model are very close,confirming the validity of the model for the adsorption of Pb(II) iononto the composite.

3.4.4.3. Intra-particle diffusion model. In order to investigate themechanism and adsorption behavior of Pb(II) ion onto FAC, kineticsresults were also analyzed using the intra-particle diffusion model.During the batch mode of adsorption, there is the possibility oftransport of sorbate species in the pores of the sorbent, which isoften the rate-controlling step. This model is significant becauseit is the rate-determining step in the liquid adsorption systems.The widely applied intra-particle diffusion equation for such a sys-tem is given as [68]:

a-particle diffusion model(c) of Pb(II) ion adsorption by FAC (100 mg/L Pb(II) ion

Table 3Adsorption isotherm model parameter of adsorption of Pb(II) ion on FAC at 296 K.

Pseudo-first-order constants Pseudo-second-order constants Intra-particle diffusion

qe (mg/g) K1 (/min) R2 qe (mg/g) K2 (g/mg/min) R2 Ki1 Ki2 C R21 R22

535.4 0.02418 0.909 500 7.339 � 10�5 0.991 24.42 0.963 131.1 0.9955 0.9996

Fig. 9. Separation of particles from reaction system using external magnet, FA (a),and FAC (b).

M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168 167

qt ¼ kit0:5 þ c

where ki is the intra-particle diffusion rate constant (mg/g min ½)and the intercept C, obtained by extrapolation of the linear portionof the plot of qt versus t0.5 back to the axis, is taken to be propor-tional to the boundary layer thickness. From the plots of qt versest0.5 multi-linearities were observed as shown in Fig. 8c, indicatinga two-step adsorption by FAC. The first sharper portion was attrib-uted to boundary layer diffusion of Pb(II) ion, and the second linearportion was attributed to the final equilibrium stage for whichintra-particle diffusion slowed due to the extremely low Pb(II) ionsconcentration remaining in solution. The initial portion of the plotwas attributed to the boundary layer effect or external mass trans-fer effect [64]. The slope and intercept of the first portion of the plotindicates the boundary layer diffusion characteristics of adsorption,while the second linear portion indicates pore diffusion. The valuesfor Ki1 and Ki2 were calculated from the slope of each plot (Table 3).The Ki1 values were approximately 25 times greater than Ki2 sug-gesting that boundary layer diffusion was rate controlling for theadsorption of Pb(II) by FAC.

3.5. Separation of FAC

Separation of the adsorbent dispersed in the media for adsorp-tion of heavy metal is an important aspect of a low cost materialfor water treatment. Fig. 9 shows the separation of particles fromthe reaction system using an external magnet. FAC particles(Fig. 9b) were separated using external magnet while fly ash parti-cles (Fig. 9a) could not be separated by this process. Magnetic ironoxide nanoparticle doped on the FA surface allowed easy separa-tion by an external magnetic field, which would alleviate the diffi-culties associated with removing adsorbent dispersed in media forwater purification purposes.

4. Conclusion

Simultaneous deposition of Ag and iron oxide NPs onto FA par-ticles was carried out to produce the magnetically separable com-posite with enhanced antibacterial capacity and lead(II) ionsadsorption performance. FA provided large surface area for

in situ deposition of NPs. The Pb(II) adsorption efficiency wasincreased from 416.6 to 526.5 mg/g, and also displayed highantibacterial activity due to the synergetic effect of Fe3O4 and AgNPs. The Langmuir model had a good fit with a high regressioncoefficient for adsorption process. The composite showed excellentseparation performance due to the magnetic property of the ironoxide decorated on the fly ash composite. Therefore, the obtainedcomposite from this facile one-pot synthesis utilizes a wastebyproduct of thermal plant to produce a composite particle thatexhibited great potential for environmental remediation.

Acknowledgments

This research was supported by a grant from the Basic ScienceResearch Program through the National Research Foundation ofKorea (NRF) by Ministry of Education, Science and Technology(Project Nos: 2013R1A2A2A04015484 and 2014R1A1A2009068).We would also like to thank KBSI Jeonju, Korea for BET surface areameasurement and Centre for University Research Facility (CURF),Chonbuk National University for spectroscopic measurements.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jcis.2015.04.044.

References

[1] A. Heidari, H. Younesi, Z. Mehraban, Chem. Eng. J. 153 (2009) 70–79.[2] A. Sarı, M. Tuzen, J. Hazard. Mater. 157 (2008) 448–454.[3] J. Song, H. Oh, H. Kong, J. Jang, J. Hazard. Mater. 187 (2011) 311–317.[4] S. Mohan, R. Gandhimathi, J. Hazard. Mater. 169 (2009) 351–359.[5] R.R. Navarro, S. Wada, K. Tatsumi, J. Hazard. Mater. 123 (2005) 203–209.[6] A. Maldonado-Reyes, C. Montero-Ocampo, O. Solorza-Feria, J. Environ. Monit. 9

(2007) 1241–1247.[7] N.K. Batchu, C.H. Sonu, M.S. Lee, Hydrometallurgy 144–145 (2014) 1–6.[8] P.E. Franco, M.T. Veit, C.E. Borba, G.d.C. Gonçalves, M.R. Fagundes-Klen, R.

Bergamasco, E.A. da Silva, P.Y.R. Suzaki, Chem. Eng. J. 221 (2013) 426–435.[9] H.J. Kim, H.R. Pant, N.J. Choi, C.S. Kim, Chem. Eng. J. 230 (2013) 244–250.

[10] B. Ersoy, T. Kavas, A. Evcin, S. Bas�pınar, A. Sarııs�ık, G. Önce, Fuel 87 (2008)2563–2571.

[11] P. Thongsanitgarn, W. Wongkeo, A. Chaipanich, C.S. Poon, Constr. Build. Mater.66 (2014) 410–417.

[12] R.S. Iyer, J.A. Scott, Resour. Conserv. Recycl. 31 (2001) 217–228.[13] K. Yeole, P. Kadam, S. Mhaske, J. Mater. Res. Technol. 3 (2014) 186–190.[14] H.R. Pant, B. Pant, R.K. Sharma, A. Amarjargal, H.J. Kim, C.H. Park, L.D. Tijing,

C.S. Kim, Ceram. Int. 39 (2013) 1503–1510.[15] G. Gupta, N. Torres, J. Hazard. Mater. 57 (1998) 243–248.[16] T. Morris, H. Copeland, E. McLinden, S. Wilson, G. Szulczewski, Langmuir 18

(2002) 7261–7264.[17] M. Annadhasan, T. Muthukumarasamyvel, V.R. Sankar Babu, N. Rajendiran,

ACS Sustain. Chem. Eng. 2 (2014) 887–896.[18] E. Sumesh, M.S. Bootharaju, Anshup, T. Pradeep, J. Hazard. Mater. 189 (2011)

450–457.[19] L. Deng, X. Ouyang, J. Jin, C. Ma, Y. Jiang, J. Zheng, J. Li, Y. Li, W. Tan, R. Yang,

Anal. Chem. 85 (2013) 8594–8600.[20] M. Biswal, K. Bhardwaj, P.K. Singh, P. Singh, P. Yadav, A. Prabhune, C. Rode, S.

Ogale, RSC Adv. 3 (2013) 2288–2295.[21] J. Zhang, H. Cui, B. Wang, C. Li, J. Zhai, Q. Li, Appl. Surf. Sci. 300 (2014) 51–57.[22] C. Li, B. Wang, H. Cui, J. Zhai, Q. Li, J. Mater. Sci. Technol. 29 (2013) 835–840.[23] M. Kumari, C.U. Pittman Jr, D. Mohan, J. Colloid Interface Sci. 442 (2015) 120–

132.[24] A. Alqudami, N. Alhemiary, S. Munassar, Environ. Sci. Pollut. Res. 19 (2012)

2832–2841.[25] H.J. Kim, M.K. Joshi, H.R. Pant, J.H. Kim, E. Lee, C.S. Kim, Colloids Surf., A 469

(2015) 256–262.[26] S. Yang, Y. Sun, L. Chen, Y. Hernandez, X. Feng, K. Müllen, Sci. Rep. 2 (2012).

168 M.K. Joshi et al. / Journal of Colloid and Interface Science 453 (2015) 159–168

[27] Y.-F. Yang, G.-S. Gai, Z.-F. Cai, Q.-R. Chen, J. Hazard. Mater. 133 (2006) 276–282.[28] Z. Haiying, Z. Youcai, Q. Jingyu, Waste Manage. 31 (2011) 331–341.[29] M. Visa, L. Isac, A. Duta, Appl. Surf. Sci. 258 (2012) 6345–6352.[30] J. Pan, H. Yao, X. Li, B. Wang, P. Huo, W. Xu, H. Ou, Y. Yan, J. Hazard. Mater. 190

(2011) 276–284.[31] M.K. Joshi, H.R. Pant, H.J. Kim, J.H. Kim, C.S. Kim, Colloids Surf., A 446 (2014)

102–108.[32] A.N. Ökte, D. Karamanis, Appl. Catal. B 142–143 (2013) 538–552.[33] B. Wang, C. Li, J. Pang, X. Qing, J. Zhai, Q. Li, Appl. Surf. Sci. 258 (2012) 9989–

9996.[34] L. Slavov, M.V. Abrashev, T. Merodiiska, C. Gelev, R.E. Vandenberghe, I.

Markova-Deneva, I. Nedkov, J. Magn. Magn. Mater. 322 (2010) 1904–1911.[35] S. Saha, J.M. Wang, A. Pal, Sep. Purif. Technol. 89 (2012) 147–159.[36] Y. Wu, J. Zhang, L. Xiao, F. Chen, Appl. Catal. B 88 (2009) 525–532.[37] H. Iida, K. Takayanagi, T. Nakanishi, T. Osaka, J. Colloid Interface Sci. 314 (2007)

274–280.[38] P. Refait, J.M.R. Génin, Corros. Sci. 34 (1993) 797–819.[39] A.A. Olowe, J.M.R. Génin, Corros. Sci. 32 (1991) 965–984.[40] T. Sugimoto, E. Matijevic, J. Colloid Interface Sci. 74 (1980) 227–243.[41] M. Montazer, A. Shamei, F. Alimohammadi, Mater. Sci. Eng., C 38 (2014) 170–

176.[42] A. Amarjargal, L.D. Tijing, I.-T. Im, C.S. Kim, Chem. Eng. J. 226 (2013) 243–254.[43] J. Zeng, Y. Zheng, M. Rycenga, J. Tao, Z.-Y. Li, Q. Zhang, Y. Zhu, Y. Xia, J. Am.

Chem. Soc. 132 (2010) 8552–8553.[44] A. López-Miranda, A. López-Valdivieso, G. Viramontes-Gamboa, J. Nanopart.

Res. 14 (2012) 1–11.[45] Z. Khan, J.I. Hussain, A.A. Hashmi, Colloids Surf., B 95 (2012) 229–234.[46] J.-J. Zhang, M.-M. Gu, T.-T. Zheng, J.-J. Zhu, Anal. Chem. 81 (2009) 6641–6648.[47] G. Xie, S.N. Timasheff, Protein Sci. 6 (1997) 211–221.

[48] S.H.Y. Lo, Y.-Y. Wang, C.-C. Wan, J. Colloid Interface Sci. 310 (2007) 190–195.[49] C.E. Hoppe, M. Lazzari, I. Pardiñas-Blanco, M.A. López-Quintela, Langmuir 22

(2006) 7027–7034.[50] K.M. Rosenblatt, H. Bunjes, Mol. Pharm. 6 (2009) 105–120.[51] H. Wang, X. Qiao, J. Chen, X. Wang, S. Ding, Mater. Chem. Phys. 94 (2005) 449–

453.[52] D.-H. Zhang, G.-D. Li, J.-X. Li, J.-S. Chen, Chem. Commun. (2008) 3414–3416.[53] A.V. Rupa, D. Manikandan, D. Divakar, T. Sivakumar, J. Hazard. Mater. 147

(2007) 906–913.[54] C. Marambio-Jones, E.V. Hoek, J. Nanopart. Res. 12 (2010) 1531–1551.[55] N.N. Nassar, J. Hazard. Mater. 184 (2010) 538–546.[56] T.K. Naiya, A.K. Bhattacharya, S.K. Das, Environ. Prog. 27 (2008) 313–328.[57] S.B. Johnson, G.V. Franks, P.J. Scales, D.V. Boger, T.W. Healy, Int. J. Miner.

Process. 58 (2000) 267–304.[58] D. Xu, X.L. Tan, C.L. Chen, X.K. Wang, Appl. Clay Sci. 41 (2008) 37–46.[59] Z.-H. Huang, X. Zheng, W. Lv, M. Wang, Q.-H. Yang, F. Kang, Langmuir 27

(2011) 7558–7562.[60] S. Rahimi, R.M. Moattari, L. Rajabi, A.A. Derakhshan, M. Keyhani, J. Ind. Eng.

Chem. 23 (2015) 33–43.[61] T. Akar, Z. Kaynak, S. Ulusoy, D. Yuvaci, G. Ozsari, S.T. Akar, J. Hazard. Mater.

163 (2009) 1134–1141.[62] M.J. Baniamerian, S.E. Moradi, A. Noori, H. Salahi, Appl. Surf. Sci. 256 (2009)

1347–1354.[63] M. Özacar, _I.A. S�engil, Bioresour. Technol. 96 (2005) 791–795.[64] T. Sheela, Y.A. Nayaka, Chem. Eng. J. 191 (2012) 123–131.[65] M. Özacar, Chemosphere 51 (2003) 321–327.[66] L. Zhou, Z. Liu, J. Liu, Q. Huang, Desalination 258 (2010) 41–47.[67] Y.S. Ho, G. McKay, Process Biochem. 34 (1999) 451–465.[68] A.E. Ofomaja, Bioresour. Technol. 101 (2010) 5868–5876.