Bioresource Technology 102 (2011) 5278–5282

Contents lists available at ScienceDirect

Bioresource Technology

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

Sorption of phosphate onto giant reed based adsorbent: FTIR, Raman spectrumanalysis and dynamic sorption/desorption properties in filter bed

Xing Xu, Baoyu Gao ⇑, Qinyan Yue, Qianqian ZhongKey Laboratory of Water Pollution Control and Recycling (Shandong), School of Environmental Science and Engineering, Shandong University, Jinan 250100, PR China

a r t i c l e i n f o a b s t r a c t

Article history:Received 2 September 2010Received in revised form 23 October 2010Accepted 27 October 2010Available online 3 November 2010

Keywords:PhosphateAdsorbentGiant reedDesorptionRaman spectrum

0960-8524/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.biortech.2010.10.130

⇑ Corresponding author. Tel.: +86 531 88364832; faE-mail address: bygao@sdu.edu.cn (B. Gao).

A sorption process for the removal of phosphate was evaluated under various conditions using a filter bedpacked with giant reed (GR) based adsorbent. FTIR spectrum measurement validated the existence ofgrafted amine groups in the adsorbent and Raman spectrum displayed the characteristic peaks of differ-ent forms of phosphate. The column sorption capacity of the adsorbent for phosphate was 54.67 mg g�1

in comparison with the raw GR of 0.863 mg g�1. Influent pH demonstrated an essential effect on the per-formance of the filter bed as compared to other influent conditions (flow rates and influent concentra-tions) and the optimal pH was selected at 5.0–10.0. Eluents of HCl, NaOH and NaCl solutions withconcentrations of 0.01–0.1 mol l�1 showed the excellent capacities for desorption of phosphate fromthe adsorbent, and their elution processes could be finished in 90 min.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

The presence of phosphate in wastewaters provides an addi-tional nutrient in the aquatic environments. Though phosphate isan essential nutrient for growth of microorganisms in aquatic envi-ronments, concentration in an excess of the desired limit is a majorcause for eutrophication of lakes, rivers and sea thereby posingserious concern (Özacar, 2003). Phosphate makes its way into theaquatic environments through natural processes such as rocksweathering as well as human activities such as industrial, agricul-tural and domestic uses (Riahi et al., 2009). In China, the typicalraw domestic waste water has a total phosphorus concentrationof approximately 8–10 mg P/l with orthophosphate as theprincipal form of phosphate (5 mg P/l) in addition to other formsof phosphate. In order to meet effluent quality standards(0.5–1 mg P/l), development of treatment methods that facilitatethe removal of phosphate from wastewaters prior to discharge intonatural waters is required.

In wastewater-treatment technology, various chemical andbiological techniques have been successfully applied, such asreverse osmosis (Jeppesen et al., 2009), biochemical technology(Obaja et al., 2003), electro-dialysis (Mohamed, 2002), adsorption(Xu et al., 2009) and so forth. Of all these, adsorption is well recog-nized as one of the simplest and safest methods used for theremoval of pollutants from wastewater. The adsorbents could be

ll rights reserved.

x: +86 531 88364513.

prepared by chemical interaction between reactive entities to formstable covalent bonds between a sorption material and a functionalgroup possessing positive charges (Panthapulakkal et al., 2006). Inrecent years, numerous studies have been made by differentresearchers by using natural biomass such as spruce rice hull (Or-lando et al., 2002), pine bark (Orlando et al., 2002), pea stems(Wartelle and Marshall, 2006), and wheat straw (Panthapulakkalet al., 2006) as sorption materials.

In this work, a new kind of adsorbent was prepared from giantreed (GR) by amination reaction. The preparation of GR basedadsorbent was the chain reactions between cellulose/hemicellu-lose chain and side chains of different grafting chemical reagents(The preparation scheme was not presented in this paper, and sim-ilar result was shown in our previous research work for wheatstraw modification Xu et al., 2009). N,N-dimethylformamide wasused as an organic medium which enhanced the susceptibility ofthe epoxide ring in epichlorohydrin (Orlando et al., 2002). Ethy-lenediamine could be used as a crosslinking agent for some synthe-sis reactions between epichlorohydrin and amine (Liu et al., 2010;Ong et al., 2007). After the preparation synthesis, various charac-teristic measurements were conducted to determine the physico-chemical properties of the GR based adsorbent. Column sorptioncharacteristics of phosphate onto the adsorbent were studied byvarying the influent conditions in the continuous system such asinfluent concentrations, flow rates and pH. In addition, an attemptwas made by selecting various desorption eluents (HCl, NaCl andNaOH solutions) to evaluate the regeneration property of GR basedadsorbent in filter bed.

X. Xu et al. / Bioresource Technology 102 (2011) 5278–5282 5279

2. Methods

2.1. Preparation of GR based adsorbent

GR was obtained from Weishan Lake of Jinin, Shandong, China.The GR was washed with water, dried at 60 �C for 12 h and sievedinto particles with diameters from 100 to 250 lm.

GR was reacted with epichlorohydrin and N,N-dimethylform-amide (DMF) in a three-neck round bottom flask at 85 �C for60 min. Ethylenediamine was added and the mixture was stirredfor 45 min at 85 �C, followed by adding triethylamine (TEA) forgrafting and stirring for 120 min at 85 �C. The product was washedwith 500 ml of distilled water to remove the residual chemicals,dried at 60 �C for 12 h and sieved to obtain particles smaller than250 lm in diameter.

2.2. FTIR and Raman spectrum techniques

The functional groups presenting in GR, GR based adsorbent andadsorbent saturated with phosphate were investigated by usingthe FTIR technique (Perkin–Elmer ‘‘Spectrum BX’’ spectrometer).The spectrum was scanned from 400 to 4000 cm�1. The saturatedsample was prepared by mixing the adsorbent with solutioncontaining 1 mol l�1 of phosphate.

Raman spectroscopic analysis was performed to provide in-sights into the mechanisms of phosphate interactions with GRbased adsorbent. In the Raman analysis, 0.1 g GR based adsorbentwas placed in 50 ml of solution containing phosphate concentra-tion of 1 mol l�1. The wet solid samples were analyzed by Ramanspectroscopy (Nicolet Almega XR Dispersive Raman, Thermo Elec-tron Corporation, USA). The laser wavelength used in Ramanmeasurement was 1050 nm. The standard materials monopotas-sium phosphate (solid form) and 1 mol l�1 of solution monopo-tassium phosphate (liquid form) were also analyzed (Yoonet al., 2009).

2.3. Column sorption tests

An organic-glass column with 200 mm length and 12 mm diam-eter filled with 1 g of GR based adsorbent was fed with variousconcentrations of phosphate solutions (50, 100 and 200 mg P/l).The flow rates were controlled at about 3.3, 5, 10 and 15 ml min�1

and influent pH of the solutions were selected at 2.00, 3.68, 5.12,7.08, 8.55, 9.85 and 12.14. The effluent solutions were collected,and every 10 ml was selected as a sample to determine the concen-trations of phosphate in the effluent solutions. The flow to the col-umn was continued until the effluent phosphate concentration (Ct)approached the influent concentration (C0), Ct/C0 = 0.98. Phosphatesolutions were prepared by high purity water, and its normal pHvalue is about 5.12.

2.4. Desorption tests

Regeneration of the GR based adsorbent as well as recovery ofadsorbate material was achieved by eluting a suitable solventthrough the exhausted column. In the present studies, various con-centrations of dilute NaCl, HCl and NaOH solutions (0.1, 0.01 and0.001 mol l�1) were eluted through the column. After washed withdistilled water, the regenerated adsorbent was used again in thesubsequent experiments. In addition, if adsorption is by physicalbonding then the loosely bound ions can be easily desorbed withdistilled water, and therefore, a control test was carried out byusing distilled water as eluent.

3. Results and discussion

3.1. FTIR spectrum

The FTIR spectrum exhibits the characteristic cellulose peak in thefinger print region of 1000–1200 cm�1, indicating that the main skel-eton of GR is the cellulose chains. Peak between 1650 and 1750 cm�1

are indication of free and esterified carboxyl groups which corre-spond to the identification of pectins present in GR. The broad bandsaround 3200 and 3600 cm�1 is indicative of the existence of hydroxylgroups of macromolecular association in cellulose, hemicellulose,pectin, etc. (Huang et al., 2009; Wahab et al., 2010).

Compared to the FTIR spectrum of GR, some significant changesare observed after the FTIR analysis of GR based adsorbent. Appear-ance of peak around 1345 cm�1 is assigned to C–N stretchingvibration, which corresponds to the amine groups in the frame-work of the adsorbent (Tsai et al., 2006). It is apparent in FTIR spec-trum of GR based adsorbent that an intense vibration is observed atband around 1330–1380 cm�1, indicating the grafted amine groupsin GR based adsorbent.

The FTIR spectrum of the phosphate-loaded adsorbent demon-strates a significant shift of peak from 1345 cm�1 to a higher wavenumber (1356 cm�1) in comparison with other peaks. The changein the 1345 cm�1 band may be associated to the coordination ofthe phosphate anions with the amine groups in an R �Nþ(CH2CH3Þ3 � � �H2PO�4 complex. Similar shifting trends were ob-served in the adsorption tests for Cr(III), Cd(II) and Pb(II) removalby amine functionalized silica gel (Huang et al., 2008).

3.2. Raman spectrum

In the Raman spectrum test, appearance of peaks in GR basedadsorbent at 2975.3, 1459.9 and 1600.5 cm�1 is the characteristicpeaks of cellulose chains. Characteristic peaks of phosphate varyaccording to the existence of different forms of phosphate. Phos-phate in a 1 mol l�1 of KH2PO4 solution has a characteristic peakat 899 cm�1. Solid sodium phosphate exhibits a peak at 914 cm�1,which is a large shift from the peak position of aqueous phosphate.This indicates the bonding environment of phosphate in the solid isquite different from that of free phosphate ion in the solution. Thepeak of the adsorbed phosphate on the adsorbent is at 895 cm�1. Itis observed that the peak in the spectra of adsorbed phosphateagrees with the characteristic peak of the free phosphate ion insolution, indicating that there is no strong chemical interaction be-tween the adsorbed phosphate and the adsorbent (Yoon et al.,2009; Hu et al., 2005). It suggests that phosphate is adsorbed ontothe adsorbent surface through ion exchange by displacing the chlo-ride ions as the phosphate ions in the positively charged aminegroups. Therefore, the mechanism of phosphate species sorptionby the adsorbent may be proposed as in Eq. (1).

R � NþðCH2CH3Þ3 � � � Cl� þH2PO�4! R � NþðCH2CH3Þ3 � � � þH2PO�4 þ Cl� ð1Þ

Ion exchange is a fast sorption process for the removal of vari-ous ions. A fast sorption of phosphate by the GR based adsorbentsuggests that column sorption test will be more practical for thepotential application. Therefore, further researches on the phos-phate sorption and desorption properties in filter bed tests wereexhibited as follows.

3.3. Comparison of GR and GR based adsorbent in the column sorptiontest

When the sorption zone moves up and the upper edge of thiszone reaches the bottom of the column, the effluent concentration

5280 X. Xu et al. / Bioresource Technology 102 (2011) 5278–5282

starts to rise rapidly. This is called the breakthrough point. Thedesired breakthrough point was determined to be 0.1 Ct/C0. Thepoint where the effluent phosphate concentration reached 95%(Ct/C0 = 0.95) of its influent value is called the point of columnexhaustion (Suksabye et al., 2008).

Solutions with concentration of 200 mg P l�1 were infused intotwo filter beds fed with 1 g of GR and GR based adsorbent, respec-tively. The flow rate was controlled at 5 ml min�1. The break-through curve, Ct/C0 versus volume is shown in Fig. 1.

It is observed from Fig. 1 that the raw GR in filter bed almosthave no effect on the sorption of phosphate, and the column sorp-tion capacity (qed) of GR for phosphate is only 0.863 mg g�1. The GRbased adsorbent was prepared from GR after the amination reac-tion, and it is apparent that its phosphate sorption capacity hasbeen greatly enhanced in comparison with that of GR. As is shownin Fig. 1, the breakthrough point occurs at the breakthrough vol-ume of 190 ml. As the phosphate solution continues to flow intothe column, the GR based adsorbent gradually becomes saturatedwith phosphate ions and becomes less effective for further adsorp-tion. The point on the S-shaped curve at which the phosphate con-centration approaches its exhaustion value is about 450 ml. Whenthe effluent phosphate concentration reaches the influent concen-tration, the calculated qed at this point is 54.67 mg g�1.

3.4. Effect of flow rate on the breakthrough curves

The effect of flow rates on the breakthrough curves was inves-tigated from 3.3 to 15 ml min�1 with constant influent phosphateconcentration of 200 mg l�1. The breakthrough curve Ct/C0 versusbreakthrough volume treated with various flow rates is shown inFig. 2a.

The breakthrough points for the various flow rates (3.3, 5, 10,15 ml min�1) are 243, 190, 164 and 125, respectively. The resultsindicate that a decrease in flow rates at constant influent concen-tration of 200 mg l�1 increases the breakthrough volume or break-through time due to an increase in empty bed contact time (EBCT).The lower the EBCT, the less effective the diffusion process be-comes, resulting in lower sorption capacity for phosphate (Sarinet al., 2006; Suksabye et al., 2008). Thus, the GR based adsorbentin filter bed needs more time to bond the phosphate ions effi-ciently. The results also show that the breakthrough process is sat-urated earlier at higher flow rate because the front of theadsorption zone quickly reached the bottom of column. In contrast,lower flow rates and longer contact time, resulted in a shallowadsorption zone.

100 200 300 400 500 6000.0

0.2

0.4

0.6

0.8

1.0

Ct/C

0

Breakthrough volume (ml)

GR GR based resin

Fig. 1. Breakthrough curves of GR and GR based adsorbent for the sorption ofphosphate (Flow rate: 5 ml min�1; influent concentration: 200 mg l�1; influent pH:5.12).

A decrease in the phosphate sorption capacity of the adsorbentfrom 55.24 to 45.29 mg g�1 is observed as the flow rate increasesfrom 3.3 to 15 ml min�1. This corresponds to the decrease in thebreakthrough volume from 243 to 125 ml. Based on the discussionon the effect of flow rate, it is suggested that lower flow rate orlonger contact time would be required for phosphate sorption inthe column tests.

3.5. Effect of influent concentrations on the breakthrough curves

The effect of phosphate concentrations (80, 120 and 200 mg l�1)on the breakthrough curves at a constant flow rate of 5 ml min�1

was investigated in Fig. 2b. As seen from Fig. 2b, the breakthroughpoints occur at 430, 270 and 190 ml for influent phosphate concen-trations of 80, 120 and 200 mg l�1, respectively. The breakthroughvolume decreases with the increase in influent phosphate concen-trations. However, it is observed that the sorption capacities of theadsorbent slightly increase from 52.59 to 54.98 mg g�1 in thatinfluent concentration range; this may be attributed to the concen-tration gradient which will enhance the sorption process at higherinfluent phosphate concentrations (Goel et al., 2005).

For the studies on the effect of influent phosphate concentra-tion, it can be concluded that the GR based adsorbent could be usedfor various influent concentration of phosphate, with little effecton its sorption capacity.

3.6. Effect of influent pH on the breakthrough curves

The effect of influent pH (2.00, 3.68, 5.12, 7.08, 8.55, 9.85 and12.14) on the breakthrough curves is shown in Fig. 3. As indicatedin Fig. 3a, the breakthrough points for the various influent pH (2.00,3.68, 5.12, 7.08, 8.55, 9.85 and 12.14) are 30, 80, 190, 200, 190, 160and 50 ml, respectively. Fig. 3b shows the column sorption capac-ities versus influent pH. As seen from Fig. 3b, the phosphate sorp-tion capacities of the adsorbent are 12.7, 28.54, 54.56, 53.65, 53.76,49.23 and 17.90 mg g�1 for the influent pH of 2.00, 3.68, 5.12, 7.08,8.55, 9.85 and 12.14, respectively. Based on the data of break-through points and column sorption capacities, it is apparent thatthe phosphate sorption capacities are much lower when the influ-ent pH is at strong acid and strong base conditions. When the pH islower than 3.0, the species of phosphate ions mainly exist in theform of H3PO4 and H2PO4

� (Gisbert et al., 2010). The increasingconcentration of phosphoric acid will interfere with the sorptionof phosphate onto the exchange sites in the adsorbent. When thepH increases beyond 11.0, the OH� will increase significantly,and the exchange sites for phosphate sorption will decrease onthe outer surface of the adsorbent due to the presence of excessOH� ions competing with phosphate ions for sorption sites and aresult of the sorption capacity decreases (Gisbert et al., 2010;Chand et al., 2009).

The effluent pH data of the column were also investigated bythe constant assay of the effluent solution. The pH of column efflu-ent is gradually decreased from the influent pH of 3.68, 5.12, 7.08,8.55, 9.85 and 12.14 to the final pH of 3.66, 4.70, 5.50, 6.13, 7.70and 11.20, respectively. This result may be attributed to the speciesof phosphate ions in the aqueous solution (i.e. H3PO4, H2PO4

�,HPO4

2� and PO43�), which forms a buffer solution in the aqueous

solution. When phosphate ions are sorbed onto the adsorbent,the H+ is released from the solution, which results in a decreaseof pH in the solution (Gisbert et al., 2010). In addition, weaklyacidic carboxyl inherently in the GR based adsorbent would alsodecrease the pH of solutions in neutral and mild alkali conditions.This result was generally in agreement with literature report forthe adsorption of basic and reactive dyes by ethylenediamine mod-ified rice hull (Ong et al., 2007).

100 200 300 400 500 6000.0

0.2

0.4

0.6

0.8

1.0

Ct/

C0

Breakthrough volume (ml)

3.3 ml min -1

5 ml min -1

10 ml min -1

15 ml min -1

(a) flow rate

200 400 600 8000.0

0.2

0.4

0.6

0.8

1.0

Ct/C

0

Breakthrough volume (ml)

80 mg L-1

120 mg L-1

200 mg L-1

(b) influent concentration

Fig. 2. Effect of flow rate and influent concentration on the breakthrough curves (a, influent phosphate concentration: 200 mg l�1, influent pH: 5.12; b, flow rate: 5 ml min�1,influent pH: 5.12).

0 100 200 300 400 500 6000.0

0.2

0.4

0.6

0.8

1.0

Breakthrough volume (ml)

Ct/C

0

pH 2.00 pH 3.68 pH 5.12 pH 7.08 pH 8.55 pH 9.85 pH 12.14

(a)

2 4 6 8 10 120

10

20

30

40

50

60

0

2

4

6

8

10

12

Eff

luen

t pH

q e (m

g g-1

)

Influent pH

qe (mg g-1)

Influent pH Effluent pH (b)

Fig. 3. Effect of influent pH on the breakthrough curves (a) and column sorption capacities (b) (Influent phosphate concentration: 200 mg l�1, flow rate: 5 ml min�1).

X. Xu et al. / Bioresource Technology 102 (2011) 5278–5282 5281

Based on the discussion above, it seems that the suitable influ-ent pH range could be selected in the range of 5.0–10.0.

3.7. Desorption tests in filter bed

HCl, NaCl and NaOH solutions with various concentrations wereused as the eluents and their desorption efficiencies are presentedin Table 1. When the concentration of HCl, NaCl and NaOH solu-tions is 0.001 mol l�1, their desorption efficiencies are about65.4%, 51.3% and 47.8%, respectively. The desorption efficienciesincrease with the further increase in the concentrations of desorp-tion eluents, and the desorption efficiencies almost reach the max-imum of 100% at concentrations of 0.1 mol l�1. Based on the factorsin costs and efficiencies, a suitable concentration of the eluentscould be selected in the range of 0.01–0.1 mol l�1 for the desorp-

Table 1Desorption efficiencies of desorption agents at different influent concentrations.

Concentrations (mol l�1) Desorption efficiency (%)

HCl NaCl NaOH Distilled water

0.1 100% 100% 100%0.01 92.2% 86.9% 83.0% 3.5%0.001 65.4% 51.3% 47.8%

tion test in filter bed. In addition, the desorption of phosphate bydistilled water is only 3.5%; this illustrates the insignificance ofphysical bonding in the sorption of phosphate onto GR basedadsorbent.

The desorption efficiencies of various eluents follow the orderas: HCl > NaCl > NaOH solutions, attributed to the smaller hydrat-edradius of HCl and more effective interaction between amine sitesand Cl�. In particular, neutral salt solution of NaCl has demon-strated a similar desorption capacity as compared to the strongacid solution of HCl; this indicates that the desorption of phos-phate ions by NaCl solution is most probably through a reactionof ion-exchange with high concentration of Cl- from the NaCl dis-placing phosphate ions from the surface of the adsorbent.

For further discussion on the desorption property of variousdesorption eluents, dynamic desorption tests were designed andtheir elution curves are shown in Fig. 4 (Lodeiro et al., 2006). It isapparent that one of the most important advantages of thesedesorption agents is their high desorption rate, particularly in thefirst hour; the peaks of phosphate concentrations in desorptionsolutions occur at 21, 15 and 17 min for desorption agents ofHCl, NaCl and NaOH solutions, respectively. When the elutionreaches to 60 min, the effluent phosphate concentration is lowerthan 25 mg l�1. As desorption time reaches to 90 min, the residuephosphate concentration approaches to zero and the elution pro-cess is finished. Lodeiro reported a weight loss (27%) of protonatedSargassum muticum after the eleventh sorption–desorption cycles

20 40 60 80 100 1200

50

100

150

200

250

300

Desorption time (min)

Eff

luen

t pho

spha

te c

once

ntra

tion

(m

g L

-1)

HCl solution NaCl solution NaOH solution

Fig. 4. Curves of HCl, NaCl and NaOH (Concentrations of the desorption agents:0.01 mol l�1; flow rate: 5 ml min�1).

5282 X. Xu et al. / Bioresource Technology 102 (2011) 5278–5282

by using 0.1 mol l�1 of HNO3 solution as the eluent (Lodeiro et al.,2006). In this work, a weight loss (1–3%) of the GR based adsorbentis observed after the elution of HCl, NaCl and NaOH solution (oneregeneration cycle). It seems that a long desorption time couldcause the destruction of cellulose/hemicellulose structure in GRbased adsorbent. On contrary, a short elution process could be inef-fective, so it is important to appropriately balance this procedure.

4. Conclusion

Adsorbent prepared from GR was used for the sorption of phos-phate in this work. Influent pH demonstrated an essential effect onthe performance of the filter bed as compared to other influentconditions (flow rates and influent concentrations). Lower flowrate or longer contact time would be required for phosphate sorp-tion. Dynamic elution tests validated the ion exchange mechanismfor the sorption of phosphate onto the adsorbent. In addition,desorption eluents of HCl, NaOH and NaCl solutions demonstratedtheir high desorption rates for the regeneration of the adsorbent,with slight losses in their initial column sorption capacities.

Acknowledgements

The research was supported by the National Natural ScienceFoundation of China (50878121), Key Projects in the National Sci-ence & Technology Pillar Program in the Eleventh Five-year PlanPeriod (2006BAJ08B05-2) and National Major Special TechnologicalProgrammes Concerning Water Pollution Control and Managementin the Eleventh Five-year Plan Period (2008ZX07010-008-002).

References

Chand, R., Watari, T., Inoue, K., Torikai, T., Yada, M., 2009. Evaluation of wheat strawand barley straw carbon for Cr (VI) adsorption. Sep. Pur. Technol. 65, 331–336.

Gisbert, R., García, G., Koper, M.T.M., 2010. Adsorption of phosphate species on poly-oriented Pt and Pt (1 1 1) electrodes over a wide range of pH. Electrochim. Acta.55, 7961–7968.

Goel, J., Kadirvelu, K., Rajagopal, C., Garg, V.K., 2005. Removal of lead (II) byadsorption using treated granular activated carbon: batch and column studies.J. Hazard. Mater. 125, 211–220.

Hu, J., Chen, G.H., Lo, I.M.C., 2005. Removal and recovery of Cr(VI) from wastewaterby maghemite nanoparticles. Water Res. 39, 4528–4536.

Huang, L.Y., Ou, Z.Y., Boving, T.B., Tyson, J.L., Xing, B.S., 2009. Sorption of copper bychemically modified aspen wood fibers. Chemosphere 76 (8), 1056–1061.

Huang, X.P., Chang, X.J., He, Q., Cui, Y.M., Zhai, Y.H., 2008. Tris (2-aminoethyl) aminefunctionalized silica gel for solid-phase extraction and preconcentration of Cr(III), Cd (II) and Pb (II) from waters. J. Hazard. Mater. 157, 154–160.

Jeppesen, T., Shu, L., Keir, G., Jegatheesan, V., 2009. Metal recovery from reverseosmosis concentrate. J. Clean. Prod. 17, 703–707.

Liu, Y., Sun, X.M., Li, B.H., 2010. Adsorption of Hg2+and Cd2+ by ethylenediaminemodified peanut shells. Carbohyd. Polym. 81 (2), 335–339.

Lodeiro, P., Herrero, R., Sastre de Vicente, M.E., 2006. Batch desorption studies andmultiple sorption–regeneration cycles in a fixed-bed column for Cd(II)elimination by protonated Sargassum muticum. J. Hazard. Mater. B137, 1649–1655.

Mohamed, A.M.O., 2002. Development of a novel electro-dialysis based techniquefor lead removal from siltyclay polluted soil. J. Hazard. Mater. 90, 297–310.

Obaja, D., Macé, S., Costa, J., Sans, C., Mata-Alvarez, J., 2003. Nitrification,denitrification and biological phosphorus removal in piggery wastewaterusing a sequencing batch reactor. Bioresour. Technol. 87 (1), 103–111.

Orlando, U.S., Baes, A.U., Nishijima, W., Okada, M., 2002. Preparation of agriculturalresidue anion exchangers and its nitrate maximum adsorption capacity.Chemosphere 48 (10), 1041–1046.

Ong, S.T., Lee, C.K., Zainal, Z., 2007. Removal of basic and reactive dyes usingethylenediamine modified rice hull. Bioresour. Technol. 98, 2792–2799.

Özacar, M., 2003. Adsorption of phosphate from aqueous solution onto alunite.Chemosphere 51 (4), 321–327.

Panthapulakkal, S., Zereshkian, A., Sain, M., 2006. Preparation and characterizationof wheat straw fibers for reinforcing application in injection moldedthermoplastic composites. Bioresour. Technol. 97, 265–272.

Riahi, K., Thayer, B.B., Mammou, A.B., Ammar, A.B., Jaafoura, M.H., 2009. Biosorptioncharacteristics of phosphates from aqueous solution onto Phoenix dactylifera L.date palm fibers. J. Hazard. Mater. 170 (2–3), 511–519.

Sarin, V., Singh, T.S., Pant, K.K., 2006. Thermodynamic and breakthrough columnstudies for the selective sorption of chromium from industrial effluent onactivated eucalyptus bark. Bioresour. Technol. 97 (2006), 1986–1993.

Suksabye, P., Thiravetyan, P., Nakbanpote, W., 2008. Column study of chromium(VI)adsorption from electroplating industry by coconut coir pith. J. Hazard. Mater.160, 56–62.

Tsai, W.T., Yang, J.M., Lai, C.W., Cheng, Y.H., Lin, C.C., Yeh, C.W., 2006.Characterization and adsorption properties of eggshells and eggshellmembrane. Bioresour. Technol. 97 (3), 488–493.

Wahab, M.A., Jellali, S., Jedidi, N., 2010. Ammonium biosorption onto sawdust: FTIRanalysis, kinetics and adsorption isotherms modeling. Bioresour. Technol. 101,5070–5075.

Wartelle, L.H., Marshall, W.E., 2006. Quaternized agricultural by-products as anionexchange resins. J. Environ. Manage. 78 (2), 157–162.

Xu, X., Gao, B.Y., Wang, W.Y., Yue, Q.Y., Wang, Y., 2009. Adsorption of phosphatefrom aqueous solutions onto modified wheat residue: characteristics, kineticand column studies. Colloids Surf. B 70, 46–52.

Yoon, I.H., Meng, X.G., Wang, C., Kim, K.W., Bang, S., Choe, E., Lippincott, L., 2009.Perchlorate adsorption and desorption on activated carbon and anion exchangeresin. J. Hazard. Mater. 164, 87–94.