Bioresource Technology 148 (2013) 221–227

Contents lists available at ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Preparation of novel alginate based anion exchanger from Ulva japonicaand its application for the removal of trace concentrations of fluoridefrom water

0960-8524/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.biortech.2013.08.116

⇑ Corresponding author. Tel.: +81 952 28 8671; fax: +81 952 28 8669.E-mail address: inoue@elechem.chem.saga-u.ac.jp (K. Inoue).

Hari Paudyal a, Bimala Pangeni a, Katsutoshi Inoue a,⇑, Hidetaka Kawakita a, Keisuke Ohto a,Kedar Nath Ghimire b, Shafiq Alam c

a Department of Applied Chemistry, Saga University, Honjo 1, Saga 840-8502, Japanb Central Department of Chemistry, Tribhuvan University, Kirtipur, Kathmandu, Nepalc Faculty of Engineering and Applied Science, Memorial University, St. John’s, NL A1B 3X5, Canada

h i g h l i g h t s

� Novel alginate based anion exchangerwas prepared from seaweed Ulvajapaonica.� Fluoride ion from plating solution

was successfully clear using smallamount of M-CSW.� M-CSW showed high sorption

capacity, stability and regenerated byNaOH solution.� Mechanism of adsorption/elution is

inferred to be ion exchange rexn

between OH and F.

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

OH

OOH

O

OO

O

O

O O

OH

O

OH

Ca

calcium alginate in CSW

OH

OOH

O

OO

O

O

O O

OH

O

OH

M

metal loaded CSW (M-CSW)

OH

Metal solution

Ca(II)

F- adsorption mechanism

Metal ions = Zr(IV) and La(III)

F- OH-Metal loading via Ca(II) exchange

OH

OOH

O

OO

O

O

O O

OH

O

OH

M

F

OH- F-

F- desorption mechanism

fluoride loaded M-CSW

a r t i c l e i n f o

Article history:Received 4 July 2013Received in revised form 15 August 2013Accepted 19 August 2013Available online 3 September 2013

Keywords:Waste seaweedAlginateMetal loadingFluoride removalZirconium(IV) ion

a b s t r a c t

A green seaweed, Ulva japonica, was modified by loading multivalent metal ions such as Zr(IV) and La(III)after CaCl2 cross-linking to produce metal loaded cross-linked seaweed (M-CSW) adsorbents, which werecharacterized by elemental analysis, functional groups identification, and metal content determination.Maximum sorption potential for fluoride was drastically increased after La(III) and Zr(IV) loading, whichwere evaluated as 0.58 and 0.95 mmol/g, respectively. Loaded fluoride was quantitatively desorbed byusing dilute alkaline solution for its regeneration. Mechanism of fluoride adsorption was inferred in termsof ligand exchange reaction between hydroxyl ion on co-ordination sphere of the loaded metal ions of M-CSW and fluoride ion in aqueous solution. Application of M-CSW for the treatment of actual waste platingsolution exhibited successful removal of fluoride to clear the effluent and environmental standards inJapan, suggesting high possibility of its application for the treatment of fluoride rich waste water.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Fluoride is easily reacted with hydroxyapatite, a main compo-nent of teeth and skeletal tissue, whereby it replaces the hydroxideion to form fluoroapatite (Sundaran et al., 2008; Aoba, 1997; Mous-ny et al., 2008). As such, pollution of water resources with fluoride

is now a worldwide health concern, which may directly linked tothe discharge of waste water containing fluoride ion from variousindustries including coal mining, semiconductor production, metalplating, etc. (Turner et al., 2005 Samatya et al., 2010). Such wastewater contains ten to several thousand mg/l of fluoride, concentra-tions far greater than both the WHO (World Health Organization)standard (1.5 mg/l) and the Japanese environmental standard(0.8 mg/l) for drinking water (Chen et al., 2011; Paudyal et al.,2012a). Therefore, excessive fluoride must be removed from waste

222 H. Paudyal et al. / Bioresource Technology 148 (2013) 221–227

water before discharging into environment. With this in regard,several techniques including precipitation, ion exchange, mem-brane separation and adsorption have been developed for the re-moval of fluoride from water (Wioeniewski and Rozanska, 2007;Hu et al., 2005; Liao and Shi, 2005; Sinha et al., 2003).

The combined precipitation of fluoride by Ca2+ followed by Al3+

is a widely used method. Here, fluoride ions are first precipitatedwith lime or calcium salt thereby lowering the fluoride concentra-tion to approx. 10–20 mg/l, as a result of the solubility limit ofCaF2. Further purification of trace concentrations of fluoride tothe acceptable standard (0.8 mg/l in Japan), requires additionaltreatment with alum or aluminium salts (Wajima et al., 2009).However, this method generates large amounts of sludge contain-ing high levels of retained water, necessitating additional costlytreatment for dewatering and drying prior to disposal. Moreover,the increase in the concentration of Al3+ ions in the treated water(>0.2 mg/l) may cause dementia, a dangerous disease (Meenakshiand Maheshwari, 2006). Membrane separation and ion exchangetechnologies are generally expensive and unsuitable for the treat-ment of waste water as it typically contains very fine solid particlesthat can easily plug the micropores of membranes and resins. In re-sponse to these shortcomings, novel technology for fluoride re-moval using inexpensive biomasses containing carboxylic acidand amine functional group have attracted attention and havebeen extensively studied in recent years (Wang et al., 2013; Paud-yal et al., 2011; Viswanathan et al., 2009). The adsorption of fluo-ride by Ulva japaonica biomass has never been studied yethowever; U. japaonica, known to be one of the most importantsources of natural alginate containing carboxylic acid functionalgroups (Ghimire et al., 2007).

Accompanied by eutrophication in bays and inland seas close tolarge cities in Japan, U. japonica, a green seaweed, has recently seenextensive growth in the coastal regions of such seas. Largeamounts of these seaweed species have been commonly observeddecaying on the beaches of these coasts, which has becomeincreasingly undesirable due to their bad odor and unsightlyappearance. Consequently, the post treatment of such waste sea-weed had become a serious problem for the municipalities of thesecoast areas. Many cities on such coast area have regularly investedlarge sums of money every year to clean the coast by removing thewaste of the U. japonica. Therefore, the development of some effec-tive uses of this waste seaweed is now strongly desired in Japan.The Ulva species of seaweed contain high levels of polysaccharideswhich including alginate (15–65%), lignin (1–3%), protein (6–15%),lipid (1–4%), other amino acid, minerals and water soluble organics(Kraan, 2012). In the present work, we attempted to develop a no-vel removal technology for dilute concentrations of fluoride fromwater by effectively using the waste of the U. japonica, containinglarge amounts of alginate.

Alginate exists in seaweeds as calcium alginate which possesses anegg box structure. However, because calcium alginate itself exhibitsonly a weak affinity for anionic species including fluoride as will bementioned later, some modification is necessary for the purpose ofdeveloping active adsorption sites for fluoride ion on the polymermatrices of alginate in U. japonica. In previous work involving assess-ing fluoride removal using orange juice residue, it was reported thatorange pectic acid containing carboxylic acid functional group devel-ops active adsorption sites for fluoride ions once loaded with high va-lence metal ions thereby successfully removing fluoride via a ligandexchange mechanism (Paudyal et al., 2013).

In the present study, metal ions were loaded onto the polymermatrix of the U. japaonica alginate after cross-linking. It is expectedthat carboxyl groups of the alginic acid in the U. japaonica and thepyranose oxygen atom will form stable five membered chelates withloaded metal ions where active sites for fluoride were created. Theapplication of metal loaded cross-linked seaweed as novel adsorp-

tion gels to remove fluoride from synthetic solution as well as actualwaste plating solution, and underlying fluoride adsorption/desorp-tion mechanism has also been studied in the present paper.

2. Methods

2.1. Chemicals and analysis

The chemicals used in this study were of reagent grade andused directly without further purification. Zirconiumoxy-chlorideoctahydrate (ZrOCl2 � 8H2O) and lanthnium chloride heptahydrate(LaCl3 � 7H2O) were used for the modification of the waste sea-weed. Fluoride stock solution (1000 mg/l) was prepared by dissolv-ing 1.05 g of NaF in 500 ml of deionized water. All the chemicalsused were purchase from Wako Chemical Co. Ltd. Japan. Functionalgroup analysis of the adsorbents was carried out using a JASCOFTIR-410 Fourier Transform Infrared Spectrometer. The concentra-tion of metal ions were measured using an inductively coupledplasma atomic emission spectrometer (ICP-AES, Shimadzu modelICPS 8100). The elemental composition of U. japnonica adsorbentswere measured using a Shimadzu model EDX-800HS, an energydispersive X-ray spectrometer. An ion chromatography system(Dionex model ICS-1500) equipped with a self-regenerating sup-pressor with conductivity detector and anion exchange columnwas used for the quantitative analysis of anionic species in thetested sample.

2.2. Preparation of metal loaded cross linked seaweed

The sample of U. japnonica waste was collected from the Wajirocoast of Hakata bay in Fukuoka city, located in the Northern regionof Kyushu Island, Japan, in order to prepare the adsorption gel forfluoride ion adsorption. The sample was washed several times withtap water followed by distilled water to remove impurities fromseaweed surface. After drying in a convection oven at 70 �C for24 h, the dried seaweed was crushed and sieved to particles sizeof less than 100 lm. As will be mentioned in detail the latter sec-tion, adsorption sites in seaweeds such as U. japnonica are locatedin the alginic acid. Because alginic acid is partly water soluble, itshould be cross-linked to avoid the dissolution of such adsorptionsites in water in order to be effectively employed as a solid adsor-bent. In the present work, the seaweed was cross-linked using cal-cium chloride in the similar manner as applied in previous work,and as follows (Ghimire et al., 2008). Fifty gram of powdered sam-ple of U. japnonica was mixed with 500 ml of 1 M CaCl2 solutionand stirred for 24 h to enable the cross-linking reaction. The result-ing solid particles, the cross-linked product, was separated by fil-tration and washed several times with distilled water andsubsequently dried in a convection oven at 70 �C. The dark greenproduct (16.3 g) obtained by this method was termed as cross-linked seaweed and abbreviated as CSW hereafter. Since one ofthe main components in CSW is calcium alginate which functionsas a cation exchange type-material containing exchangeable cal-cium, some modification is required to create active adsorptionsites for anionic species such as fluoride. The active sites for fluo-ride ion removal were generated in CSW by metal loading reac-tions as follows. At first, 1000 mg/l of Zr(IV) and La(III) solutionwere prepared individually by dissolving 1.76 g of ZrOCl2 � 8H2Oand 1.34 g of LaCl3 � 7H2O in 500 ml water, respectively. Afterthat, 3 g of CSW was incorporated into a flask containing 500 mlof 1000 mg/l metal solution at optimal pH (pH = 2.2 for Zr(IV) load-ing and pH = 7.0 for La(III) loading). Then, the mixture was stirredat 40 �C for 24 h to complete the loading reaction of the metal ionsonto the CSW. Finally, the mixture was filtered and washed thor-oughly with distilled water until a neutral pH was obtained, and

H. Paudyal et al. / Bioresource Technology 148 (2013) 221–227 223

then dried in convection oven at 70 �C. The material obtained inthis way was termed as metal loaded cross-linked seaweed andabbreviated as M-CSW hereafter. Here, the Ca2+ ion of calcium algi-nate in CSW were replaced by loaded metal ions by cation ex-change mechanism to produce M-CSW adsorbent. It was furtherconfirmed from the energy dispersive X-ray spectroscopic analysisin Section 3.1.

2.3. Batch wise studies

2.3.1. Metal loading experimentsThe loading test of Zr(IV) and La(III) ions onto the CSW was car-

ried out by mixing 15 mg of CSW together with 10 ml of 0.5 mmol/lmetal solution at varying pH (1–7) and shaken for 24 h at 30 �Cusing thermostatic air bath incubator. The respective pH valueswere adjusted through the addition of a few drops of 0.5 M HNO3

or 0.5 M NaOH solutions. The loading capacities of CSW for Zr(IV)and La(III) ion were investigated at 30 �C using varying concentra-tions of metal ion (0.1–10 mmol/l). The percentage loading andquantity of loaded metal ions (q, mmol/g) were calculated accord-ing to the following mass balance equations:

%A ¼ Ci � Ce

Ci� 100 ð1Þ

q ¼ Ci � Ce

W� V ð2Þ

where, Ci and Ce are initial and equilibrium concentrations (mmol/l),respectively. W is the dry weight of the adsorbent (g) and V is thevolume of solution (l). For the determination of the total amountof loaded metal ions in M-CSW, 50 mg of the M-CSW was dissolvedin 10 ml of aqua regia, stirred for 24 h to ensure complete dissolu-tion and finally analyzed using ICP-AES.

2.3.2. Adsorption tests of fluoride on M-CSWBatch wise adsorption test of fluoride ions was conducted using

the CSW and the M-CSW at a solid–liquid ratio of 1.5 g/l using a0.5 mmol/l fluoride solution at different pH (2–12) in order toinvestigate the effect of pH on the adsorption behavior. The mix-tures were stirred at 30 �C for 24 h to ensure that adsorption equi-librium was attained. After this, the samples were filtered and theresidual concentration of fluoride ions in the filtrate was measuredvia ion chromatography. The percentage adsorption and theamount of fluoride adsorption onto the CSW and the M-CSW werealso calculated according to Eqs. (1) and (2). Adsorption isothermsfor fluoride were investigated for a solid liquid ratio of 1.5 g/l byvarying the concentration of fluoride ions (0.5–6 mmol/l) at pH 5.Adsorptive removal of fluoride from actual waste plating solutionwas tested using zirconium(IV) loaded CSW (Zr(IV)-CSW) and lan-thanum(III) loaded CSW (La(III)-CSW) at varying solid liquid ratio(0.25–8 g/l) using the actual sample solution containing traceamount of fluoride ions (0.64 mmol/l) at a pH of 6.7 kindly pro-vided by Federation of Electro Plating Industry Association, Japan.After the solid liquid mixture was shaken for 24 h, the slurry was

Table 1Comparison of infrared band of Ulva japonica before and after modification.

Functional group Infrared absorption frequencies (cm�1) by seawe

SW CSW

OH stretching 3421 3426CH2 vibration 2914 2918COO stretching 1627 1654C@O stretching – –C–O–C stretching – 1167OH bending 1022 1054

filtered and the residual concentration of fluoride in treated samplewas measured using ion chromatography.

2.3.3. Desorption of loaded fluoride and cycle stabilityThe desorption of adsorbed fluoride on the M-CSW was carried

out by using dilute alkali solution at 30 �C where 25 mg of fluorideloaded Zr(IV)-CSW was mixed with 10 ml of varying concentra-tions of NaOH solution (0.01–2 M) and stirred for 24 h. After filtra-tion, the fluoride concentration in the filtrate was measured by ionchromatography. Cycle test of fluoride adsorption followed bydesorption was carried out by using Zr(IV)-CSW for 6 repeated cy-cles. For each cycles, fluoride ions was first adsorbed onto Zr(IV)-CSW by using 16.3 mg/l fluoride solution at solid liquid ratio of2.5 g/l at pH 5 then adsorbed fluoride was eluted by using NaOHsolution also at 2.5 g/l.

3. Results and discussion

3.1. Characterization of M-CSW adsorbents

The elemental composition of the waste U. japnonica seaweed(SW), the feed material, and CSW before and after metal loadingwas qualitatively measured by using an energy dispersive X-ray(EDX) spectroscopy as shown in Supplementary Fig. 1. In the EDXspectra of SW (Supplementary Fig. 1a), the elemental peaks of C(0.24 keV), O (0.33 keV), Na(1.06 keV), Al (1.48 keV), Si (1.74 keV),P (2.01 keV), S (2.31 keV), K (3.32 and 3.59 keV), Ca (3.70 and4.02 keV), Ti(4.52), Mn (5.92 keV), and Fe (6.41 and 7.07 keV) wereobserved. On the other hand, in that of CSW (SupplementaryFig. 1b), the intensity of the peaks assigned to Ca at 3.70 and4.02 keV was found to have increased, suggesting that some of freealginic acid in the SW might also have been converted into calciumalginate during cross-linking with CaCl2. However, entirely newpeaks assigned to La were observed at 4.66, 5.06, 5.41, 5.80 and6.09 keV for the La loaded CSW (Supplementary Fig. 1c) whilenew peaks assigned to Zr appeared at around 2.08, 7.90, 9.04 and9.38 keV in the Zr(IV)-CSW (Supplementary Fig. 1d), suggestingthat these metal ions were effectively loaded onto the CSW. More-over, the intensity of the peaks assigned to Ca at 3.70 and 4.02 keVwere drastically decreased after metal loading in both M-CSW,suggesting that these metal ions were loaded according to the Casubstitution mechanism as stated earlier in Section 2.2.

Infrared spectra of different seaweed samples were recorded forthe identification of functional groups in the tested samples. Sinceactive species in the CSW are Ca-type alginic acid, free acid type(H-type) CSW (H-CSW) was prepared by washing CSW with 5 Mof HCl at the solid/liquid ratio of 2 g/l for comparison. The detailsof the FT-IR spectra of the SW, CSW, H-CSW, La(III)-CSW andZr(IV)-CSW are listed in Table 1. From this table, it was determinedthat the bands at 3419–3426, 2914–2928, 1627–1658, 1159–1167and 1022–1067 cm�1 were assigned to O–H stretching, CH2 vibra-tion, COO stretching, C–O–C stretching and O–H bending vibration,respectively. The C@O stretching vibration appearing at 1745 cm�1

in the H-CSW was absent in the CSW and the M-CSW while twonew peaks were observed at around 1627–1658 and 1511–

ed adsorbents

La(III)-CSW Zr(IV)-CSW H-CSW

3423 3419 34232920 2928 29161658 1656 1624– – 17451159 1164 –1067 1062 1057

Table 2Loading parameters of Zr(IV) and La(III) onto cross-linked seaweed and its comparison with leached concentrations of metal ions from the M-CSW using aqua-regia.

Dissolution of metal ions from M-CSW Loading test of metal ions onto CSW

Mn+ ions Mn+ leachedusing aqua-regia (mmol/g) pH (Mn+ loading) Langmuir parameters

qmax. (mmol/g) b (l/mmol) r2

Zr(IV) 1.16 2.2 1.18 11.46 0.99La(III) 0.48 7.0 0.52 6.35 0.99

224 H. Paudyal et al. / Bioresource Technology 148 (2013) 221–227

1539 cm�1, indicating the presence of metal alginate bonds in theCSW and the M-CSW.

3.2. Metal loading on CSW by Zr(IV) and La(III)

The loading efficiency of CSW for Zr(IV) and La(III) as a functionof equilibrium pH is presented in Supplementary Fig. 2. As shownin this figure, the loading efficiency increases with increasing equi-librium pH for both the metal ions. Nearly complete loading ofZr(IV) (99%) was achieved at pH as low as 2.23 while, in the caseof La(III), the loading efficiency was observed to have increasedfrom 13 to 98% with increasing pH from 2.12 to 7.00. Increasingloading efficiency with increasing pH of the solution for both themetal ions indicates that these metal ions were loaded onto CSWaccording to the cation exchange mechanism.

In order to evaluate the loading capacity of the CSW for Zr(IV)and La(III) ions, adsorption isotherms of these metal ions on theCSW were investigated by varying the metal ion concentration(0.1–10 mmol/l) at their optimal pH (Supplementary Fig. 3). TheLangmuir equation (Eq. (3)) was used to evaluated the maximumloading capacity (qmax, mmol/g) and the binding constant oradsorption equilibrium constant (b, l/mmol). The plots of Ce/qe vsCe were clustered on a straight line as expected from Eq. (3).

Ce

qe¼ 1

qmaxbþ Ce

qmaxð3Þ

The maximum loading capacity (qmax) and binding constant (b)for both metal ions were calculated from the slope and intercept ofthe straight lines, respectively, as listed in Table 2, which showsthat the maximum adsorption capacity of the CSW for Zr(IV) wasgreater than that for the La(III), suggesting a higher affinity of the

0

20

40

60

80

100

0 2 4 6 8 10 12

Ads

orpt

ion

of f

luor

ide

ion

[%]

pHe [-]

CSW

La(III)-CSW

Zr(IV)-CSW

Fig. 1. Influence of equilibrium pH for the adsorption of fluoride onto CSW and M-CSW (fluoride concentration = 0.5 mmol/l, solid–liquid ratio = 1.5 g/l, tempera-ture = 30 �C, shaking = 24 h).

CSW for the Zr(IV) compared to La(III). The high value of the corre-lation regression coefficient for Langmuir isotherm model(r2 > 0.99) suggests that adsorption of metal ions onto the CSWlikely takes place according to the Langmuir’s monolayer adsorp-tion model.

3.3. Batch wise adsorption test of fluoride on M-CSW using syntheticsolutions

3.3.1. Effect of pH on the fluoride adsorption mechanismIn most of the cases, fluoride adsorption is highly pH dependent

because fluoride appears as different chemical species dependingon the solution pH. Hydrofluoric acid [HF, pKa = 3.18 (Tang et al.,2009)] is the predominant species in water at lower pH (<3) withtrace amount of HF2

�, whereas more than 90% of fluoride existsin the free anionic form (F�) at pH greater than 5 (Deng et al.,2011). Fig. 1 shows the% adsorption of fluoride on the CSW beforeand after metal loading as a function of equilibrium pH. It is clearfrom the result that, as the pH of the fluoride solution increasesfrom 2.1 to 3.2, the percentage adsorption of fluoride also increasesfrom 8.2% to 29.4% for the CSW and maximum adsorption of fluo-ride occurred at a pH of approx. 4–5 where fluoride ions interactedwith positively charged calcium ions contained in the alginate ofthe U. japonica by electrostatic interaction as shown in elementaryreaction a. In the case of La(III)-CSW, the% adsorption of fluorideincreased from 19.9% up to 79.7% with the increase of equilibriumpH from 2.1 to 3.4 whereas maximum adsorption of fluoride wasobserved at a pH of around 4–5. In the case of Zr(IV)-CSW, the %adsorption of fluoride was 97.7% at a pH of around 2 which in-creased with increasing pH reaching a maximum at a pH of around3–5, where 100% removal of fluoride was achieved. The drasticimprovement in adsorption efficiency of the CSW for fluoridewas observed after metal loading which is attributable to the highaffinity of fluoride ion with the loaded metal ions. On the otherhand, at pH values above 6, fluoride adsorption gradually de-creased; negligible adsorption of fluoride was observed at pH high-er than 9 for the CSW and both M-CSW tested. The decrease influoride adsorption at high pH region is attributable to competitiveadsorption between the fluoride and the hydroxide ions for the ac-tive sites, whereas the sharp decrease in fluoride adsorption at lowpH values is due to the existence of large amount of weakly ioniz-able hydrofluoric acid (HF) which may be hardly adsorbed. Suchphenomenon was also observed by Wang et al. (2013) during fluo-ride adsorption in carboxylated aerobic granules containing Ce(III).Consequently, the adsorption mechanism of fluoride on the M-CSW can be clearly interpreted by the following reactions.

PolymerBCadþ þ F� ! PolymerBCadþ � � � Fd� ðaÞ

PolymerBM� OHþHþ ! PolymerBM� OHþ2 ðbÞ

PolymerBM� OHþ2 þ F� ! PolymerBM� FþH2O ðcÞ

Net reaction : �PolymerBM� OHþ F�

! PolymerBM� Fþ OH� ðdÞ

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3 4 5 6 7

Flu

orid

e ad

sorp

tion

cap

acit

y [m

mol

/g]

Equilibrium concentration [mmol/l]

Zr(IV)-CSW

La(III)-CSW

CSW

Fig. 2. Adsorption isotherm of CSW and M-CSW for fluoride (pH = 5, solid–liquidratio = 1.5 g/l, temperature = 30 �C, shaking = 24 h).

H. Paudyal et al. / Bioresource Technology 148 (2013) 221–227 225

Hydroxide ions which are coordinated with the loaded metalions in the M-CSW are initially protonated to give protonated siteswhere fluoride ions were adsorbed with the loss of water moleculeas shown in the elementary reaction b and c. The overall reactionindicates that the hydroxyl ions on the M-CSW are substitutedby fluoride ions in aqueous solution as shown in reaction d; thatis, it is inferred that adsorption of fluoride onto the M-CSW alsotakes place by ligand exchange mechanism similar to that in La(III)loaded orange waste (Fang et al., 2003; Paudyal et al., 2012a).

3.3.2. Adsorption isotherms of fluorideAdsorption isotherms of fluoride on CSW and M-CSW are

shown in Fig. 2, which indicates that the fluoride adsorption capac-ity increases with increasing concentration within the low concen-tration region and approaches a nearly constant value in the higherconcentration region for each adsorbent tested and as observed formany other adsorption systems. The results shown in Fig. 2 wereanalyzed according to the Langmuir and Freundlich adsorption iso-therm equations. The Langmuir equation was used to evaluate themaximum adsorption capacity (qmax., mmol/g) and binding con-

Table 3Langmuir and Freundlich adsorption isotherm parameters for the adsorption of fluoride io

Langmuir parameters

Adsorbents pH qmax (mmol/g) b (l/mmol)

Zr(IV)-CSW 5 0.95 5.04La(III)-CSW 5 0.58 4.51CSW 5 0.23 2.03

Table 4Maximum adsorption capacities for fluoride on various adsorbents and their comparisonstudy.

Adsorbents pH Temp. (�C

Zr(IV)-CSW 5 30La(III)- CSW 5 30CSW 5 30Nano-hydroxyapatite/chitosan 7 30Commercial READF resin 3 30KMnO4 modified rice straw carbon 2 30Carboxylated crosslinked chitosan 7 30La(III) loaded 200CT resin 6 30Fe(III) loaded cotton cellulose 4 25Spirogyra species IO1 7 30

stant (b, l/mmol) of the CSW and the M-CSW for fluoride ions asshown in Table 3. It can be seen from this table that the adsorptioncapacity of the Zr(IV)-CSW (0.95 mmol/g) is higher than those ofLa(III)-CSW (0.58 mmol/g) and CSW (0.23 mmol/g) which suggestsincreasingly effective interaction between the fluoride ions and theloaded Zr(IV) ions relative to the La(III) and Ca(II) ions. The drasticincrease in fluoride adsorption capacity after loading with high va-lence metal ions onto the CSW is due to the development of newactive adsorption sites for fluoride by metal loading.

The Freundlich isotherm equation is expressed as:

log qe ¼ log Kf þ ð1=nÞ log Ce ð4Þ

The values of the Freundlich adsorption isotherm parameters, nand Kf, were calculated from the slope and the intercept of thestraight line for the plot of log qe vs. log Ce, respectively, and alsopresented in Table 3. The values of the correlation regression coef-ficient evaluated according to the Freundlich model are lower thanthose for the Langmuir model.

Table 4 shows the comparison of the fluoride adsorption capac-ities among various adsorbents including commercial resins andother chemically modified biomass adsorbents (Sundaran et al.,2008; Paudyal et al., 2012b; Daifullah et al., 2007; Viswanathanet al., 2009; Fang et al., 2003; Zhao et al., 2008; Mohan et al.,2007) together with the CSW and M-CSW investigated in thisstudy. It is evident from this table that the M-CSW investigatedin this study exhibits a much higher adsorption capacity for fluo-ride ions than the majority of the biosorbents reported in litera-tures. Although, the adsorption capacities of highly porouscommercial resins, READF-(PG), and La(III) loaded Amberlite200CT resin, also commercially available sulphonic acid type resin,are higher than those observed for the M-CSW, these commercialresins are considerably expensive. In addition, porous resins aresuffer from clogging of the micro-pores by very fine solid particlestypically present in the majority of waste water during water treat-ment, making the regeneration and reuse of these resins difficult.Moreover, the preparation of metal loaded CSW investigated in thisstudy is very simple, environmental friendly and economical be-cause the feed material is waste seaweed itself. The metal loadedseaweed U. japonica especially Zr(IV)-CSW showed high adsorptioncapacity and can be regenerated by simple method of alkali treat-ment which can be used for several cycles with insignificant reduc-tion of adsorption efficiency which will be described in detail in

ns on CSW and M-CSW.

Freundlich parameters

r2 Kf (mmol/g) n r2

0.99 0.84 4.46 0.930.99 0.69 5.66 0.970.99 0.45 5.97 0.90

with the gels of metal loaded cross-linked seaweed Ulva japonica investigated in this

) qmax (mmol/g) Refs.

0.95 This work0.58 This work0.23 This work0.11 Sundaran et al. (2008)2.10 Paudyal et al. (2012a,b)0.81 Daifullah et al. (2007)0.58 Viswanathan et al. (2009)1.34 Fang et al. (2003)0.97 Zhao et al. (2008)0.06 Mohan et al. (2007)

0

2

4

6

8

10

12

14

0 1 2 3 4 5 6 7 8

Rem

aini

ng c

once

ntra

tion

of

fluo

ride

[m

g/l]

S/L ratio [g/l]

La(III)-CSW

Zr(IV)-CSW

Industrial effluent standard (8 mg/l) in Japan

Drinking water standard (0.8

mg/l) in Japan

100 % removal

Fig. 3. Application of M-CSW adsorbent for the treatment of trace concentration offluoride from actual plating solution at different solid–liquid ratio (pH = 6.7,fluoride concentration = 12.2 mg/l, volume of solution = 10 ml, shaking time = 24 h,temperature = 30 �C).

226 H. Paudyal et al. / Bioresource Technology 148 (2013) 221–227

Section 3.5. Therefore, the utilization of waste of seaweed U. japon-ica for the preparation of a more sophisticated adsorbent for fluo-ride appears more promising for environmental remediation.

3.4. Application of M-CSW for the removal of fluoride from actualwaste plating solutions

As it was found that the M-CSW was effective for the adsorptionof fluoride from the synthetic solution, the removal of fluoridefrom actual waste plating solution was tested at different solid–li-quid ratio at its native pH (pH = 6.7) using Zr(IV)- and La(III)-CSWas shown in Fig. 3 where the residual fluoride concentration afteradsorption is plotted against the solid/liquid ratio, the ratio ofthe dry weight of the added adsorbent to the volume of the testsolution. This figure shows that the residual fluoride concentrationwas drastically decreased with increasing solid liquid ratio for boththe adsorbents. Although the waste plating solution contains sev-eral co-existing cations (Al: 13.8, Ca: 10.5, Fe: 36.4, Cu: 9.8 andZn: 1.9 mg/l) as well as some anionic species (SO4

2�: 754, F�:12.2 mg/l), the small amount of Zr(IV)-CSW (0.5 g/l) displayed thatit could successfully lower the fluoride concentration down to orbelow the effluent standard in Japan (8 mg/l) in comparison tothe La(III)-CSW (2 g/l), which may be due to the high affinity ofloaded Zr(IV) as compared to that of La(III) for fluoride ions asmentioned earlier. The fluoride concentration in plating solutioncan be successfully lowered down to drinking water standard(0.8 mg/l) by using more than 3 g/l of Zr(IV)-CSW. Moreover, theaddition of Zr(IV)-CSW in concentrations higher than 5 g/l can suc-cessfully achieved 100% removal of fluoride from this actual wasteplating solution.

3.5. Desorption of fluoride and cycle satiability

From the result of the adsorption test (Fig. 1), maximum fluo-ride adsorption onto the M-CSW was achieved for pH values from3 to 5, whereas negligible adsorption occurred in alkaline condi-tions (pH > 9), suggesting that selection of dilute alkali solutionfor desorption of loaded fluoride from the M-CSW is a more suit-able option. Consequently, desorption test of fluoride from fluorideloaded Zr(IV)-CSW was carried out at varying concentrations of

NaOH. It is clear from the results of desorption studies (Supple-mentary Fig. 4) that, the desorption of loaded fluoride increasedfrom 4.7% to 98.1% with increasing NaOH concentration from0.01 M to 0.1 M, respectively, which did not exceed 98.6% withincreasing NaOH concentration up to 2 M. Hence, 0.1 M NaOHsolution was concluded to be the most suitable desorbing agentfor repeated regeneration cycles. Adsorbed fluoride ions on theZr(IV)-CSW are inferred to be replaced by hydroxide ions providedfrom eluent (NaOH) solution by ligand substitution mechanism asshown in the elementary reaction e.

PolymerBM� Fþ OH� ! PolymerBM� OHþ F� ðeÞ

The adsorption of fluoride onto Zr(IV)-CSW followed by desorp-tion using 0.1 M NaOH solution up to 6 repeated cycles (Supple-mentary Fig. 5) shows that, the % adsorption of fluoride onZr(IV)-CSW was negligibly influence until 3 repeated cycles but itwas gradually decreased from �98% to �91% with increasingadsorption/elution cycles from 3 to 6. Nearly the quantitativedesorption of loaded fluoride from fluoride loaded Zr(IV)-CSWwas achieved over the entire cycle studied.

4. Conclusions

In present study, novel and highly efficient adsorbents for fluo-ride have been developed by loading high valence metal ions ontowaste seaweed U. japnonica. The Zr(IV)-CSW exhibited superiorfluoride adsorption than CSW and La(III)-CSW because the loadingreaction of the Zr(IV) ion onto the U. japonica alginate created moreactive sites for fluoride adsorption. The drastic improvement influoride adsorption potential of CSW by Zr(IV) loading and 100% re-moval of fluoride from actual waste plating solution using smallamounts of Zr(IV)-CSW provides considerable evidence thatZr(IV)-CSW could be a potential candidate for the effective removalof fluoride from waste water.

Acknowledgement

The authors are indebted to the Federation of Electro PlatingIndustry Association, Japan for the kind supply of the sample of ac-tual waste plating solution.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2013.08.116.

References

Aoba, T., 1997. The effect of fluoride on apatite structure and growth. Crit. Rev. OralBiol. Med. 8, 136–153.

Chen, N., Zhang, Z., Feng, C., Li, M., Zhu, D., Sugiura, N., 2011. Studies on fluorideadsorption of iron impregnated granular ceramics from aqueous solution.Mater. Chem. Phys. 125, 293–298.

Daifullah, A.A.M., Yakout, S.M., Elreefy, S.A., 2007. Adsorption of fluoride in aqueoussolution using KMnO4 modified activated carbon derived from steam pyrolysisof rice straw. J. Hazard. Mater. 147, 633–643.

Deng, Y., Nordstrom, D.K., McCleskey, R.B., 2011. Fluoride geochemistry of thermalwaters in Yellowstone national park: aqueous fluoride speciation. Geochem. etCosmochim. Acta 75, 4476–4489.

Fang, L., Ghimire, K.N., Kuriyama, M., Inoue, K., Makino, K., 2003. Removal offluoride using some lanthanum loaded adsorbent with different functionalgroups and polymer matrices. Chem. Technol. Biotechnol. 78, 1038–1047.

Ghimire, K.N., Inoue, K., Ohto, K., Hayashida, T., 2007. Adsorptive separation ofmetallic pollutants onto waste seaweed Porphyra yezonsis and Ulva japonica.Sep. Sci. Technol. 42, 2003–2018.

Ghimire, K.N., Inoue, K., Ohto, K., Hayashida, T., 2008. Adsorption study of metalions onto cross-linked seaweed Laminaria japonica. Bioresour. Technol. 99, 32–37.

H. Paudyal et al. / Bioresource Technology 148 (2013) 221–227 227

Hu, C.Y., Lo, S.L., Kaun, W.H., Lee, Y.D., 2005. Removal of fluoride fromsemiconductor wastewater by electrocoagulation flotation. Water Res. 39,895–901.

Kraan, S., 2012. Algal Polysaccharides, Novel Application and Outlook. Oceanharvest Technology Ltd., N17 Business Park, Milltown, Co. Galway, Ireland, pp.489–532.

Liao, X.P., Shi, B., 2005. Adsorption of fluoride on Zirconium(IV)-impregnatedcollagen fiber. Environ. Sci. Technol. 39, 4628–4632.

Meenakshi, S., Maheshwari, R.C., 2006. Fluoride in drinking water and its removal. J.Hazard. Mater. B137, 456–463.

Mohan, V.S., Ramanaiah, S.V., Rajkumar, B., Sharma, P.N., 2007. Bio-sorption offluoride from aqueous solution onto algal spirogyra IO1 and evaluation ofadsorption kinetics. Bioresour. Technol. 98, 1006–1011.

Mousny, M., Omelon, S., Wise, L., Everett, E.T., Dumitriu, M., Holmyard, D.P., Banse,X., Devogelaer, J.P., Grynpas, M.D., 2008. Fluoride effects on bone formation andmineralization are influenced by genetics. Bone 43, 1067–1074.

Paudyal, H., Pangeni, B., Ghimire, K.N., Inoue, K., Kawakita, H., Ohto, K., Alam, S.,2012a. Adsorption behavior of orange waste gel for some rare earth ions and itsapplication to the removal of fluoride from water. Chem. Eng. J. 195, 28–296.

Paudyal, H., Pangeni, B., Inoue, K., Kawakita, H., Ohto, K., Alam, S., 2012b. Removal offluoride from aqueous solution by using porous resin containing hydrated oxideof cerium(IV) or zirconium(IV). J. Chem. Eng. Jpn. 45, 331–336.

Paudyal, H., Pangeni, B., Inoue, K., Kawakita, H., Ohto, K., Alam, S., 2013. Adsorptiveremoval of fluoride from aqueous medium using a fixed bed column packedwith Zr(IV) loaded dried orange juice residue. Biores. Technol. 146, 713–720.

Paudyal, H., Pangeni, B., Inoue, K., Kawakita, H., Ohto, K., Harada, H., Alam, S., 2011.Adsorptive removal of fluoride from aqueous solution using orange wasteloaded with multi-valent metal ions. J. Hazard. Mater. 192, 676–682.

Samatya, S., Mijuki, H., Ito, Y., Kawakita, H., Uezu, K., 2010. The effect of polystyreneas a progen on the fluoride ion sorption of Zr4+-surface immobilized resin.React. Funct. Polym. 70, 63–68.

Sinha, S., Pandey, K., Mohan, D., Singh, K.P., 2003. Removal of fluoride from aqueoussolution by Eichhornia crassipes biomass and its carbonized form. Ind. Eng.Chem. Res. 42, 6911–6918.

Sundaran, C.S., Viswanathan, N., Meenakshi, S., 2008. Uptake of fluoride bynanohydroxyapatite/chitosan:- bioinorganic composite. Bioresour. Technol.99, 8226–8230.

Tang, Y., Gaun, T., Gao, N., Wang, J., 2009. Fluoride adsorption on activated alumina:modeling the effect of pH and competing ions. Physicochem. Eng. Aspects 337,33–38.

Turner, B.D., Binning, P., Stipp, S.L.S., 2005. Fluoride removal by calcite: evidence forfluorite precipitation and surface adsorption. Environ. Sci. Technol. 39, 9561–9568.

Viswanathan, N., Sundaram, C.S., Meenakshi, S., 2009. Sorption behavior of fluorideon carboxylated cross-linked chitosan. Colloids Surf. B 68, 48–54.

Wajima, T., Umeta, Y., Narita, S., Sugawara, K., 2009. Adsorption behavior of fluorideion using a titanium hydroxide derived adsorbent. Desalination 249, 323–330.

Wang, X.-H., Song, R.-H., Yang, H-C., Shi, Y-J., Dang, G-B., Yang, S., Zhao, Y., 2013.Fluoride adsorption on carboxylated aerobic granules containing Ce(III).Bioresour. Technol. 127, 106–111.

Wioeniewski, J., Rozanska, A., 2007. Donnan dialysis for hardness removal fromwater before electrodialytic desalination. Desalination 212, 251–260.

Zhao, Y., Li, X., Liu, L., Chen, F., 2008. Fluoride removal by Fe(III)-loaded ligandexchange cotton cellulose adsorbent from drinking water. Carbohydr. Polym.72, 144–150.