A

hseupntiupa©

K

1

icefsaiatma

0d

Analytica Chimica Acta 585 (2007) 211–218

Preparation of silica-supported porous sorbent for heavy metal ionsremoval in wastewater treatment by organic–inorganic hybridization

combined with sucrose and polyethylene glycol imprinting

Feng Li, Ping Du, Wei Chen, Shusheng Zhang ∗College of Chemistry and Molecular Engineering, Qingdao University of Science and Technology, Qingdao 266042, China

Received 20 October 2006; received in revised form 21 December 2006; accepted 22 December 2006Available online 10 January 2007

bstract

A new porous sorbent for wastewater treatment of metal ions was synthesized by covalent grafting of molecularly imprinted organic–inorganicybrid on silica gel. With sucrose and polyethylene glycol 4000 (PEG 4000) being synergic imprinting molecules, covalent surface coating onilica gel was achieved by using polysaccharide-incorporated sol–gel process starting from the functional biopolymer, chitosan and an inorganicpoxy-precursor, gamma-glycidoxypropyltrimethoxysiloxane (GPTMS) at room temperature. The prepared porous sorbent was characterized bysing simultaneous thermogravimetry and differential scanning calorimeter (TG/DSC), scanning electron microscopy (SEM), nitrogen adsorptionorosimetry measurement and X-ray diffraction (XRD). Copper ion, Cu2+, was chosen as the model metal ion to evaluate the effectiveness of theew biosorbent in wastewater treatment. The influence of epoxy-siloxane dose, buffer pH and co-existed ions on Cu2+ adsorption was assessedhrough batch experiments. The imprinted composite sorbent offered a fast kinetics for the adsorption of Cu2+. The uptake capacity of the sorbent

mprinted by two pore-building components was higher than those imprinted with only a single component. The dynamic adsorption in columnnderwent a good elimination of Cu2+ in treating electric plating wastewater. The prepared composite sorbent exhibited high reusability. Easyreparation of the described porous composite sorbent, absence of organic solvents, cost-effectiveness and high stability make this approachttractive in biosorption.

2007 Elsevier B.V. All rights reserved.

; Silic

tarcbiacht

eywords: Organic–inorganic; Hybrid; Chitosan; Imprinting; Sol–gel; Sorbent

. Introduction

Heavy metal contamination of water resources has been ands nowadays of great concern as it poses not only severe car-inogenic risks to humans, but also potentially unacceptablecological risks to plants, animals and microorganisms. There-ore, efficient removal of toxic heavy metals from aqueousolution by appropriate treatment technologies has long beencrucial issue [1–3]. Among the developed technologies (e.g.,

on exchange, filtration, coagulation and adsorption), however,dsorption has been shown to be the most effective one [4]. With

he increasing demand for economic large-scale water treat-

ent applications, the development of novel, low-cost, stablend efficient sorbent, therefore, is of great significance.

∗ Corresponding author. Tel.: +86 532 84022750; fax: +86 532 84022750.E-mail address: shushzhang@qust.edu.cn (S. Zhang).

mcrcmta

003-2670/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.aca.2006.12.047

a gel; Porous

Currently, biopolymers are industrially attractive becausehey are capable of binding transition metal ions, widely avail-ble and environmentally safe. Amongst them, chitosan (CS)emains special attentions as a natural polysaccharide. Chitosanould be derived from chitin, which is the second most abundantiopolymer in nature and waste product of seafood processingndustries. Additionally, repetitive functional groups, principalmino groups (–NH2) in the polyaminoglucosan matrix, makehitosan a promising biosorbent with excellent capacity to entrapeavy metal ions [5–7]. However, it is difficult to directly applyhis biomass in the form of flakes or powder in wastewater treat-

ent due to the disadvantages, e.g., swelling, solubility in acidiconditions, unsatisfying mechanical property and mass transferesistance. Progress has been made to produce cross-linked

hitosan particles/beads so that they can be regenerated afteretal adsorption and reused in recycling operations. However,

he procedures, conventional emulsion cross-linking methods,re commonly not suitable for large-scale water treatment

2 mica

atmpaFsticigftDtca[

itnufbaottcamfocaamb

potlf[sdlrmcwpdO

c[

ni4lsctma

2

2

wuCs4(hlnct

2

oefaUfXptcp

2c

fsaf

12 F. Li et al. / Analytica Chi

pplications [8]. As a more practical alternative, prepara-ion of composite materials by coating chitosan on preformed

icroparticle, silica gel, followed by chemical cross-linking hasroved successful [9,10]. This synthetic approach is convenientnd economically attractive based on four interesting properties.irstly, specific functionality of chitosan and advantages ofilica gel, e.g., large surface area and excellent mechanical resis-ance are combined. Secondly, much lower quantity of chitosans needed to build the sorbent, while the overall metal adsorptionapacity may not be affected. Thirdly, the composite sorbents expected to possess increased accessibility of complexingroups due to the presence of polysaccharide on the bead sur-ace. Fourthly, coating of cross-linked chitosan has been provedo enhance stability of silica gel in alkaline solutions [9,10].espite successes achieve with those composite sorbents,

oxicity problems, caused by the use of conventional chemicalross-linking reagents, e.g., glutaraldehyde, epichlorohydrinend ethylene glycol diglycidyl ether, should be overcome11].

Along with the development of sol–gel chemistry, organic–norganic hybrid materials have recently attracted much atten-ion in both industries and researches. Taking inspiration fromatural biomineralization processes, that are believed to be reg-lated by proteins and polysaccharides, many recent researchesocus on incorporation of polysaccharide, into inorganic networky polysaccharide-manipulated sol–gel process starting fromlkoxysiloxane precursors [12,13]. Currently, chitosan-basedrganic–inorganic hybrid materials take on an increasing impor-ance in developing functional biomaterials [14]. Those bioma-erials compare favorably to materials obtained by conventionalhemical cross-linking reagents, as they offer advantages suchs low toxicity and high biocompatibility. Very recently, a novelethod proposed by Liu et al. and Shirosaki et al. shows potential

or developing chitosan-based hybrid materials through covalentrganic–inorganic hybridization [15–17]. In situ cross-linking ofhitosan and simultaneous formation of chitosan-silica hybridre revealed by a set of investigation, e.g., 29Si NMR, scanningnd transmission electronic microscopy. In addition, the hybridembranes show high hydrophilicity and good cytocompati-

ility.To obtain rapid uptake kinetics and high binding capacity,

orous materials have been synthesized by the combinationf one or more pore-building process. The sol–gel coatingechnology appears to be universal in nature and molecu-arly imprinted sol–gel coating has considerable potentialor creating materials with desired morphologic properties18,19]. Coating chitosan-based organic–inorganic hybrid onilica gel via imprinted sol–gel process is very promising forirect preparation of porous sorbent with available functionaligand, excellent mechanical resistance and low mass transferesistance. Very recently, we prepare new silica gel supportedatrices based on surface imprinting coating technique and

hitosan-incorporated sol–gel process. To provide macropores

ith adequate pore accessibility and high biocompatibility forrotein, PEG 20000 has been used as the imprinting moleculeue to its effects on silica gelling and phase separation [20].n the same time, sorbent suitable for selective cadmium

ss3w

Acta 585 (2007) 211–218

lean-up is obtained when we use ion-imprinting concept21].

This work attempts to disclose a new technology to prepare aew silica gel supported porous sorbent for removal of metal ionsn wastewater treatment. With sucrose and polyethylene glycol000 (PEG 4000) being synergic imprinting molecules, cova-ent surface coating was achieved through chitosan-incorporatedol–gel process at room temperature. Copper ion, Cu2+, washosen as the model metal ion to evaluate the effectiveness ofhe new biosorbent in wastewater treatment. The preparation

ethodology, its main characteristic features and applicationre described and discussed in detail.

. Experimental

.1. Reagents

Chitosan, with 98% deacetylation and an average moleculareight of 8 × 105 g mol−1 (Yuhua biomedical Corp., China) wassed to functionalize silica gel (100–200 mesh, Qingdao Oceanhemical Co., China). Gamma-glycidoxypropyltrimethoxy-

iloxane (GPTMS, Alfa aesar), polyethylene glycol 4000 (PEG000, Shanghai Bioengineering Corp., China) and sucroseQingdao Biomedical Co., China) were used in this study. Nin-ydrin reagent (ninhydrin, hydrindantin, dimethyl sulfoxide andithium acetate at pH 5.2, Sigma) was used to determine theon-hybrid chitosan. All other chemicals used were on analyti-al grade. Doubly deionized water (DDW) was used throughouthe work.

.2. Instrumentation

Scanning electron microscopy (SEM) images were obtainedn a Hitachi S-4100 (Hitachi, Japan) field emission scanninglectron microscope. Simultaneous thermogravimetry and dif-erential scanning calorimeter (TG/DSC) were conducted with

NETZSCH STA 409 thermogravimetric analyzer (Bruker,SA). Nitrogen adsorption porosimetry measurement was per-

ormed on an Omnisorp 100CX (Coulter, USA) apparatus.-Ray diffraction (XRD) was performed on Siemens D 5005owder X-ray diffractometer. A Hitachi 180-80 atomic absorp-ion spectrometer (Hitachi, Japan) was used to measure theoncentration of metal ions in aqueous solution. A peristalticump (Jingke Corp., China) was used in this study.

.3. Preparation of the silica gel supported porousomposite sorbent

Silica gel was heated up at 110 ◦C for 1 h to activate the sur-ace. Chitosan was dissolved in 1 mol L−1 acetic acid aqueousolution with a 3 wt.% CS concentration. Then, 5 wt.% sucrosend 10 wt.% polyethylene glycol 4000 were added. After stirringor 1 h, certain amount of gamma-glycidoxypropyltrimethoxy-

iloxane was added to the transparent solution. The mixture wastirred for 2 h and subsequently bathed in an ultrasonic bath for0 min. Then, the activated silica gel was added. The wet beadas allowed to evaporate at room temperature to complete the

mica A

cw6i(pb

2

ttetass

w

Tiww

adrflbmtbpoapabpsb

2

esdfaoCbwa

metcCeoc

tsst

Q

CCso

2e

etFp0ma1it

2

bw0C

3

3c

ssp

F. Li et al. / Analytica Chi

ross-linking reaction and gelation. The dry product obtainedas treated with a 0.5 mol L−1 ammonia solution at 80 ◦C forh to extract the imprinting sucrose and PEG [22,23]. The result-

ng material was filtered, washed with doubly deionized waterDDW), and dried under vacuum. The controlling blank wasrepared in parallel without adding of sucrose, PEG 4000 andoth two imprinting molecules, respectively.

.4. Characterization of the sorbent

The primary amino groups on the sorbent were assayed byitrimetrically [24]. The stability of the sorbents in acid solu-ion, which were prepared using different siloxane dose, wasvaluated by measuring the amount of CS released in acid extrac-ion solution. Every sorbent was shaken separately in 1 mol L−1

cetic acid aqueous solution overnight. Extracted CS was mea-ured using ninhydrin reagent [22]. Water content of the preparedorbent in DDW was calculated as follows:

ater content = W − Wd

W(1)

he weight of the wet sample (W) was determined after remov-ng the surface water by blotting with moistured filter paper. Theeight of the dry sample (Wd) was determined after the sorbentas dried under vacuum.The thermal and decomposition characteristics of the materi-

ls were determined using simultaneous thermogravimetry andifferential scanning calorimeter (TG/DSC) in the temperatureange of 50–750 ◦C at a heating rate of 20 ◦C min−1, with airushed at 200 mL min−1. The surface morphology of the sor-ent was examined with the field emission scanning electronicroscope at 5.0 kV. Samples were vacuum-dried in a desicca-

or and gold coated by a vacuum electric sputter coater beforeeing gluemounted onto the sample stud. X-ray diffraction waserformed to identify the change in the crystal structure. X-raysf 1.5408 A wavelength were generated by a Cu K source. Thengle of diffraction varied from 5◦ to 40◦. Nitrogen adsorptionorosimetry measurement was used to provide data on surfacerea and pore diameter. Nitrogen isotherms were obtained inoth adsorption and desorption modes. The surface areas of sup-orts and the sorbent were determined by the BET method. Poreize distribution curves were established from the desorptionranches of the isotherms using the BJH model.

.5. Batch adsorption experiments

The effect of pH on the adsorption of Cu2+ was tested byquilibrating 1 g of the prepared sorbent with 10 mL of bufferolutions containing 1 mg mL−1 of Cu2+ under different pH con-itions for 1 h. The pH of the solutions was adjusted using theollowing buffers: sodium acetate/hydrochloric acid for pH 2–3,nd sodium acetate/acetic acid for pH 4–6. The concentrationf every buffer solution was 100 mmol L−1. Uptake kinetics of

u2+ to the prepared sorbent was performed by adding 1 g sor-ent to 10 mL of 1 mg mL−1 Cu2+ solution at pH 6. Samplesere regularly collected at appropriate time intervals, separated

nd analyzed for Cu2+ content. Adsorption isotherm was deter-

tp

s

cta 585 (2007) 211–218 213

ined to measure the adsorption capacity. One gram sorbent wasquilibrated with 10 mL of various concentrations of Cu2+ solu-ions buffered at pH 6 at room temperature (25 ◦C). The effect ofo-existed ions, including Na+, Mg2+, Ni2+, Cl− and SO4

2−, onu2+ adsorption capacity was studied. One gram sorbent wasquilibrated with 10 mL binary mixture containing Cu2+ andne additive ion, both at 1 mg mL−1, at pH 6 for 1 h. Then, theapacity of Cu2+ adsorption was measured.

In all above batch experiments, the mixtures were separatedhrough bead sedimentation. The copper ions adsorbed on theorbent was determined by mass balance between liquid andolid phases. The amount of Cu2+ adsorbed was calculated byhe following equation:

= (C0 − Ce)V

m(2)

0 and Ce represented initial and equilibration concentration ofu2+, respectively, which was measured using atomic absorption

pectrometer. V was the volume of solution and m was the weightf the sorbent.

.6. Cu2+ adsorption from wastewater using columnxperiment

The adsorption behavior of the imprinted sorbent in treatinglectric plating wastewater was studied using column infiltra-ion. The wastewater came from the Hongda Electro-Platingactory of Shandong Province. After pretreated by adjustingH value to 5–6, the wastewater was filtered with a Millipore.45 mm filter. Afterwards, the initial Cu2+ content was deter-ined. The prepared sorbent was packed into the column with

n inner diameter of 2.6 cm, yielding an approximate length of0 cm (bed volume 53 mL). The metal solutions were pumpednto the column using the peristaltic pump and Cu2+ concentra-ion in the effluent was determined.

.7. Reuse of the sorbent

The Cu2+ ion adsorbed on the sorbent was stripped byeing washed with 0.1 mol L−1 HCl. After rinsed several timesith DDW, the acid-treated sorbent was regenerated using.1 mol L−1 NaOH. Then, it was used for batch adsorption ofu2+ through five adsorption/regeneration cycles.

. Results and discussion

.1. Preparation of the silica gel supported porousomposite sorbent

An efficient sorbent for wastewater treatment of metal ionshould consist of a stable and insoluble porous matrix havinguitable active groups that interact with heavy metal ions. Thereparation strategy selected in present investigation relied on

he use of a surface imprinting coating technique combined witholysaccharide-incorporated sol–gel process.

Silica gel constituted of the supporting material for theurface coating. Silica gel is an amorphous inorganic polymer

2 mica

hgghmop

cobe[octmwomadB[istisagip

swtfpogtaia[bpcscpAsatm

3

itcahpvTtiolTccmlccRitpIawwTf

3

For the preparation of the sorbent, interaction between chi-tosan and inorganic siloxane lied in covalent organic–inorganichybridization by in situ cross-linking of CS and simultaneous

14 F. Li et al. / Analytica Chi

aving siloxane groups (Si–O–Si) in the bulk and silanolroups (Si–OH) on its surface [19]. It has been known as aood supporting material because of its large surface area,igh mass exchange characteristics, non-swelling and excellentechanical resistance. In addition, the reactive silanol groups

n the surface favor the covalent modification of this inorganicolymer, by the introduction of new functional groups.

In our surface imprinting coating technique, pore-buildingomponents are crucial for the preparation of porous layern the surface of silica gel. As a water-soluble and hydrogenonding forming polymer, polyethylene glycol (PEG), has beenxtensively used to render materials with suitable morphology10,22]. For example, Zeng et al. have reported the preparationf microporous membranes by selective dissolution of PEG inhitosan/PEG blend membrane. The pore structure and hencehe properties of the membrane can be controlled by altering the

olecular weight of PEG [23]. Moreover, PEG also concernsith recent progress in sol–gel chemistry considering effectsn silica gelling and phase separation for the preparation ofonoliths. Smatt et al. have already established that there islinear relationship between the logarithmic plot of the pore

iameter in the monoliths versus the PEG/siloxane ratio [25].esides PEG, Pang et al. pioneer a new pore-building process

26]. They synthesize uniform sucrose/silicate nanocompos-tes through direct organic/inorganic co-assembly of low-costucrose and tetraethyl orthosilicate (TEOS). By carbonization,he obtained composites convert into carbon/silica nanocompos-te thin films. After removal of carbon or silica, respectively, bothilica and carbon films show interconnective mesopore networknd high specific surface area. In our procedure, polyethylenelycol 4000 and sucrose, therefore, were chosen as synergicmprinting molecules to endow the composite sorbent withorous characteristic.

The synthesis of the new composite sorbent involved threeteps. In the first step, hydrophilic CS, sucrose and PEG 4000ere mixed. CS was molecularly dispersed at the atmosphere of

he imprinting molecules. The second step involved covalent sur-ace coating of silica gel through chitosan-incorporated sol–gelrocess at room temperature. The step started with the addingf GPTMS, the epoxy-siloxane having trimethoxy anchorroups. Silanol groups were subsequently in situ generatedhrough acid-catalyzed self-hydrolysis. The self-condensationnd co-condensation between silanols from siloxane and sil-ca gel surface, and chitosan-incorporation by addition ofmino groups of CS to epoxy rings took place simultaneously15,16,20]. Multiple hydrogen bonds thus form in this processetween each component in the system, including silica gel,olysiloxane network, CS, sucrose and PEG molecules. Theovalent interactions, formation of hydrogen bonds, and phase-eparation-inducing effect of PEG corporately created orderlyovalent coating on silica gel. Additionally, the domains fororogens were frozen in the gel along with solvent evaporation.fter treated with hot ammonium hydroxide solution, porous

tructure was obtained through effective extraction of sucrosend PEG by breakage of the H-bonds [20,23]. Moreover, a struc-ural reorganization in such treatment gave a chemically and

echanically stabilized polysiloxane network [27].FT

Acta 585 (2007) 211–218

.2. Effect of pH on Cu2+ adsorption

The major sources of heavy metal pollutants are usually fromndustries. Copper is extensively used in the electrical industry,he electroplating industry, agricultural insecticides. Althoughopper is an essential micronutrient, vital for the body in smallmounts, however, it exhibits high acute and chronic toxicity touman body when present at high concentrations. The effect ofH on Cu2+ adsorption onto prepared sorbent was studied byarying the pH between 2 and 6, and results are shown in Fig. 1.he used sorbent was prepared using a molar ratio of siloxane

o glucosamine of 1:2. The adsorption of Cu2+ increased as pHncreased, from a low value of 5.2% at pH 2 to its maximumf 99.6% at pH 6. The minimized extent of Cu2+ adsorption atow pH might be ascribed to the protonation of amine groups.erreux et al. have studied the interactions between Cu2+ andonstitutive elements of chitosan structure, one or several glu-osamine residues, by means of density functional theory (DFT)ethods. Results showed that Cu2+ mainly bind to glucosamine

igands through coordination effect using the amino-nitrogen asoordination site [28]. Yan and Bai measured zeta potentials ofhitosan hydrogel beads in solutions at different pH values [29].esults showed that the beads possessed positive zeta potentials

n acidic solutions. As the pH decreased from 6 to 2, the posi-ive zeta potentials significantly increased. As a result, a strongrotonation existed at much acidic medium, pH 4 and below.t remarkably decreased the coordination ability of functionalmino groups in chitosan. Thus, minimized Cu2+ adsorptionas obtained. When pH increased from 4.5 to 6, protonationeakened relatively and Cu2+ adsorption increased as a result.o avoid Cu2+ removal by precipitation, pH 6 was chosen forurther experiment.

.3. Influence of epoxy-siloxane agent dosage

ig. 1. Effect of pH on the uptake of Cu2+ onto the prepared composite sorbent.en milliliters of 1 mg mL−1 Cu2+ was equilibrated with 1 g sorbent.

F. Li et al. / Analytica Chimica A

Fs

fdbAedptgBdcasfCdhtcria

3

imA(lo(fli(w(owata

agfatwlsf8

tFhpap

ig. 2. Effect of siloxane dosage on Cu2+ adsorption capacity of the preparedorbent.

ormation of CS-silica hybrid through sol–gel process. Theosage of epoxy-siloxane agent was a key parameter to controloth the availability and the accessibility of adsorption sites.

series of sorbents were prepared with different dosage ofpoxy-siloxane. Fig. 2 shows the relationship between theosage of siloxane agent and Cu2+ adsorption capacity of therepared sorbents. It was noted that the adsorption capacities ofhe sorbent increased slowly as the used molar ratio of epoxyroup to NH2 (siloxane/glucosamine) increased from 1:8 to 1:2.eyond the point of 1:2, increasing molar ratio led to significantecrease of adsorption capacity. To evaluate the hybrid status,hitosan extracted from sorbents prepared using different silox-ne dosage was measured by ninhydrin test. Acetic acid aqueousolution at 1 mol L−1 was used as the extraction solution. Exceptor that prepared using a molar ratio 1:8, almost no releasedS was detected for all other sorbents. Thus, the favorableosage of epoxy-siloxane to obtain effective hybridization andigh adsorption capacity was selected at a molar ratio of epoxyo NH2 of 1:2. The sorbent prepared with a coating solution

ontaining 5 wt.% sucrose, 10 wt.% PEG 4000 and a molaratio of siloxane to glucosamine of 1:2, was selected for furthernvestigation. An analysis of the sorbent gave 0.21 mmol ofmine content per gram beads (water content, 23.1%).

HCch

Fig. 3. The TG/DSC analysis of chitosan (a), silica g

cta 585 (2007) 211–218 215

.4. Characterization of the sorbent

The thermal stability of the prepared sorbent was investigatedn contrast with pure CS and bare silica by simultaneous ther-

ogravimetry and differential scanning calorimetry (TG/DSC).s shown in Fig. 3, the thermogravimetric curve of chitosan

Fig. 3a) showed two main degradation stages. The first weightoss occurred around 100 ◦C, which was attributed to the lossf adsorbed water molecules. The second weight loss region200–400 ◦C) might be caused by its decomposition with airush, which correlated an exothermic peak in the DSC curve,

ndicating energy release with loss of weight. For silica gelFig. 3b), only one weight loss attributed to the loss of adsorbedater molecules was observed. For the new composite sorbent

Fig. 3c), trends of weight loss and DSC were similar with thatf pure CS, indicating the efficient surface coating. Comparedith pure CS, however, the starting temperature of weight loss

nd the exothermic peak showed a shift towards high tempera-ure region. It confirmed the hybridization between the organicnd inorganic parts.

Fig. 4 represents surface morphology of supporting silica gelnd imprinted sorbent. As can be seen, the supporting silicael was in an amorphous form with a smooth and rigid sur-ace (Fig. 4a). The BET surface area was 152.55 m2 g−1 andverage pore diameter was 6.7 nm, given by nitrogen adsorp-ion porosimetry. However, significantly different morphologyas observed after coating with the imprinted organic–inorganic

ayer (Fig. 4b). As seen, the surface coating endowed the hybridorbent with a dense layer with high porosity. The BET sur-ace area was 191.95 m2 g−1 and average pore diameter was4.9 nm.

Cross-linking process usually reduced crystalline domains inhe polysaccharide and then increased adsorption capacity [30].ig. 5 shows the XRD pattern of pure CS and non-supportedybrid material. The latter one was prepared using the samerocedure but casting the coating solution on a clean glass platend analyzing the scraped power. As shown in Fig. 5a, two strongeaks appeared in diffractogram of CS at 2θ = 11.4◦ and 20.2◦.

owever, no peak for crystallization regions compared with pureS were found for non-supported hybrid material (Fig. 5b), indi-ating significant decrease of crystallization. As compared toeterogeneous cross-linking after phase-inversion, CS treatment

el (b) and the prepared composite sorbent (c).

216 F. Li et al. / Analytica Chimica Acta 585 (2007) 211–218

F (a) an×

ihiisco1ea

Fm

3

huTe

ig. 4. Scanning electron microscopy image (5.0 kV) of supporting silica gel10000 (a, right and b, left) and ×20000 (b, right).

n the homogenous condition was expected to increase material’sydrophilicity, diffusion properties and metal binding capac-ty by more destruction of crystallization [21]. Moreover, thencrease of adsorption capacity at low dosage of hybrid agent,hown in Fig. 2, might be attributed to the destruction of CSrystallinity. However, excess cross-linking at much high levelsf siloxane (molar ratio of siloxane to glucosamine larger than

:2), decreased adsorption sites and their accessibility. Thus, theffect of sufficient crystallinity destruction was counteracted anddsorption capacity decreased.

ig. 5. X-ray diffraction patterns of chitosan (a) and non-supported hybridaterial (b).

apdp

FO

d the prepared composite sorbent (b). The magnification was ×100 (a, left),

.5. Cu2+ adsorption to the sorbent

The uptake kinetics of Cu2+ adsorption to the silica-supportedybrid sorbent was investigated. As shown in Fig. 6, the Cu2+

ptake by imprinted hybrid sorbent was successful and rapid.he adsorption occurred primarily within 25 min and then thequilibrium was achieved. As known, morphology of the sorbent

nd CS crystallinity played important roles for its adsorptionerformance, e.g., kinetics and adsorption capacity, due to theecision in the legend density and mass diffusion resistance. Theorous characterization of the imprinted surface might increase

ig. 6. Uptake kinetics for Cu2+ adsorption onto the prepared composite sorbent.ne gram sorbent was added to 10 mL of 1 mg mL−1 Cu2+ solution at pH 6.

mica Acta 585 (2007) 211–218 217

tct

bde

C(sacQ(wascwc

3

fciitiNctocwtbcacicmHoobppmtoa

Ff

aaca

3

iCtMit

3

rtbaatttoocCls

4

F. Li et al. / Analytica Chi

he adsorption kinetics. Moreover, the decrease of crystallizationaused by organic–inorganic hybridization favored the adsorp-ion.

Adsorption isotherm of Cu2+ to the prepared composite sor-ent was similar to the Langmuir sorption isotherm. Adsorptionata were well fitted to the linear form of the Langmuir equationxpressed as following:

CeQ

= 1KQmax

+ CeQmax

e and Q represented equilibrium concentration of Cu2+

mg mL−1) and the amount of adsorbed Cu2+ on per gram oforbent (mg g−1), respectively. Qmax was the maximum Cu2+

dsorption and K was the Langmuir adsorption equilibriumonstant (L g−1). From the slope and intercept, the values ofmax and K were estimated as 10.5 mg g−1 and 3.17 L g−1

γ2 = 0.9955). On the other hand, the sorbents imprinted onlyith a single component, PEG 4000 and sucrose gave Cu2+

dsorption at 9.1 and 3.2 mg g−1, respectively, whereas uncoatedilica exhibited almost no adsorption for Cu2+. Obviously, theapacity of sorbent imprinted with both sucrose and PEG 4000as slightly larger than those two imprinted with only a single

omponent.

.6. The effect of co-existed ions on Cu2+ adsorption

In wastewater streams, the offending metal ion is usuallyound in matrices containing various co-existed ions. Theseomponents might interact with heavy metal ion and thus modifyts adsorption behavior towards the sorbent used. The effect ofons, including Na+, Mg2+, Ni2+, Cl− and SO4

2− on the adsorp-ion capacity of the prepared sorbent towards Cu2+ was studiedn batch conditions. Due to the binding ability to chitosan,i2+ ion was chosen as a species competitive with Cu2+. The

hoice of other ions was due to their general presence in indus-rial wastewaters [31]. To facilitate the comparison of resultsbtained from positive ions, all data have been obtained withhloride. The Cu2+ adsorption capacity in the presence of Na+

as 10.4 mg g−1. Obviously, Na+ did not significantly disturbhe adsorption of Cu2+ by hybrid chitosan. A similar trend hadeen reported by Martins et al. [32]. In other words, disturbedircumstance for the adsorption dominant by electrostatic inter-ction did not occur. Therefore, Cu2+ adsorption to the preparedomposite sorbent was mainly coordination effect. Concern-ng Mg2+, its influence was slightly stronger (Cu2+ adsorptionapacity, 9.8 mg g−1) than that observed in the case of Na+. Itight be ascribe to the competition between Cu2+ and Mg2+.owever, alkaline-earth metals were not significantly adsorbedn chitosan without chemical derivation. Thus, the influencef Mg2+ was not significant. On the contrary, the competitionecame remarkable in the presence of relatively strong com-lexing ion, Ni2+ (Cu2+ adsorption capacity, 5.3 mg g−1). Thehenomenon indicated the versatility for removal of different

etal ions using the prepared sorbent. To investigate effects of

he two classic negative ions, studies were performed in the formf sodium salts since sodium did not disturb Cu2+ adsorption,s confirmed above. We could see that no significant changes on

ei

ig. 7. In-column dynamic adsorption of Cu2+ using prepared porous sorbentor treating electric plating wastewater.

dsorption capacity were observed in the presence of Cl− (Cu2+

dsorption capacity, 10.4 mg g−1) and SO42− (Cu2+ adsorption

apacity, 10.3 mg g−1), which indicated the coordination mech-nism again.

.7. Cu2+ adsorption in wastewater media

The dynamic adsorption behavior of the prepared sorbentn treating electric plating wastewater was studied. The initialu2+ concentration was 28.6 mg L−1. As could be seen in Fig. 7,

he prepared sorbent underwent a good elimination of Cu2+.oreover, wastewater could be treated by dynamic adsorption

n column using this material, reducing the concentration of Cu2+

o less than 1 mg L−1.

.8. Desorption and reusability

To make the sorbent economically competitive, the mate-ial should be reused. Adsorbed Cu2+ could be stripped byhe introduction of protons that competed with metal ions forinding sites and made amine groups final protonation. Thedsorption ability was resumed after the regeneration of suchcid-treated sorbent using alkali solution. In test of recycling,he sorbent was used to adsorb Cu2+ through five adsorp-ion/regeneration cycles. The capacity of the prepared sorbenthrough five cycles was found to be (94 ± 3)% of the freshne. Obviously, only very slight decrease of adsorption capacityccurred in recycling studies of the prepared silica-supportedomposite sorbent. The covalent surface coating and uniqueS treatment in homogenous condition, including both cross-

inking and organic–inorganic hybridization, made the sorbenttable in recycling.

. Conclusions

With sucrose and low weight polyethylene glycol being syn-rgic pore-building components, the successful preparation ofmprinted sol–gel coating materials demonstrated the feasibility

2 mica

oTiwow

A

FttT

R

[[

[

[[

[[

[

[[[[

[[[

[[

[

[

[

18 F. Li et al. / Analytica Chi

f direct formation of porous matrix with functional ligands.he method is simple and mild. All procedures are performed

n aqueous medium without addition of any organic solvents,hich is essential from environmental and economical pointf view. This methodological study and application in low-costastewater treatment will be an important field.

cknowledgements

This research was supported by the National Natural Scienceoundation of China (No. 20475030), the Program for New Cen-

ury Excellent Talents in Universities (No. NCET-04-0649) andhe Special Project of Qingdao for Leadership of Science andechnology (No. 05–2-JC-80).

eferences

[1] W. Yantasee, Y.H. Lin, G.E. Fryxell, K.L. Alford, B.J. Busche, C.D. John-son, Ind. Eng. Chem. Res. 43 (2004) 2759.

[2] S. Wang, R.F. Zhang, Anal. Chim. Acta 575 (2006) 166.[3] J. Nan, X.P. Yan, Anal. Chim. Acta 536 (2005) 207.[4] Y. Zhang, C. Banks, Water Res. 40 (2006) 788.[5] G. Crini, Prog. Polym. Sci. 30 (2005) 38.[6] E. Guibal, Sep. Sci. Technol. 38 (2004) 43.[7] K. Oshita, M. Oshima, Y.H. Gao, K.H. Lee, S. Motomizu, Anal. Chim.

Acta 480 (2003) 239.[8] V.R. Sinha, A.K. Singla, S. Wadhawan, R. Kaushik, R. Kumria, K. Bansal,

S. Dhawan, Int. J. Pharm. 274 (2004) 1.

[9] Q.H. Shi, Y. Tian, X.Y. Dong, S. Bai, Y. Sun, Biochem. Eng. J. 16 (2003)

317.10] F.N. Xi, J.M. Wu, J. Chromatogr. A 1057 (2004) 41.11] C.C. Tsai, R.N. Huang, H.W. Sung, H.C. Liang, J. Biomed. Mater. Res. 52

(2000) 58.

[

[[

Acta 585 (2007) 211–218

12] J.E. Meegan, A. Aggeli, N. Boden, R. Brydson, A.P. Brown, L. Car-rick, A.R. Brough, A. Hussain, R.J. Ansell, Adv. Funct. Mater. 14 (2004)31.

13] Y.A. Shchipunov, T.Y. Karpenko, Langmuir 20 (2004) 3882.14] K. Molvinger, F. Quignard, D. Brunel, M. Boissiere, J.M. Devoisselle,

Chem. Mater. 16 (2004) 3367.15] Y.L. Liu, Y.H. Su, J.Y. Lai, Polymer 45 (2004) 6831.16] Y.L. Liu, Y.H. Su, K.R. Lee, J.Y. Lai, J. Membr. Sci. 251 (2005)

233.17] Y. Shirosaki, K. Tsuru, S. Hayakawa, A. Osakaa, M.A. Lopes, J.D. Santos,

M.H. Fwenandes, Biomaterials 26 (2005) 485.18] S. Dai, Y.S. Shin, C.E. Barnes, L.M. Toth, Chem. Mater. 9 (1997) 2521.19] G.Z. Fang, J. Tan, X.P. Yan, Anal. Chem. 77 (2005) 1734.20] F. Li, X.M. Li, S.S. Zhang, J. Chromatogr. A 1129 (2006) 223.21] F. Li et al., An ion-imprinted silica-supported organic-inorganic hybrid

sorbent prepared by a surface imprinting technique combined with apolysaccharide incorporated sol-gel process for selective separation ofcadmium(II) from aqueous solution, Talanta (2007), doi:10.1016/j.talanta.2006.07.023.

22] L. Yang, W.W. Hsiao, P. Chen, J. Membr. Sci. 197 (2002) 185.23] M.F. Zeng, Z.P. Fang, C.W. Xu, J. Appl. Polym. Sci. 91 (2004) 2840.24] T.Y. Guo, Y.Q. Xia, G.J. Hao, B.H. Zhang, G.Q. Fu, Z. Yuan, B.L. He, J.F.

Kennedy, Carbohydr. Polym. 62 (2005) 214.25] J.H. Smatt, S. Schunk, M. Linden, Chem. Mater. 15 (2003) 2354.26] J.B. Pang, X. Li, D.H. Wang, Z.W. Wu, V.T. John, Z.Z. Yang, Y.F. Lu, Adv.

Mater. 16 (2004) 884.27] N. Ishizuka, H. Minakuchi, K. Nakanishi, N. Soga, N. Tanaka, J. Chro-

matogr. A 797 (1998) 133.28] R. Terreux, M. Domard, C. Viton, A. Domard, Biomacromolecules 7 (2006)

31.29] W.L. Yan, R.B. Bai, Water Res. 39 (2005) 688.

30] P.K. Dutta, J. Dutta, M.C. Chattopadhyaya, V.S. Tripathi, J. Polym. Mater.

21 (2004) 321.31] H. Benaissa, B. Benguella, Environ. Pollut. 130 (2004) 157.32] A.O. Martins, E.L. da Silva, E. Carasek, N.S. Goncalves, M.C.M. Laran-

jeira, V.T. de Favere, Anal. Chim. Acta 521 (2004) 157.