Application of-chitosan-a-natural-aminopolysaccharide-for-dye-removal-from-aqueous-solutions-by-adsorption-processes-using-batch-studies-a-review-of-r

ARTICLE IN PRESS

0079-6700/$ - se

doi:10.1016/j.pr

Abbreviation:

113, acid blue 1

51, acid orange

acid red 87; AR

3, basic blue 3;

14; DB 71, dire

yellow 4; IC, in

methyl orange;

reactive blue RN

RO, reactive or

141, reactive re

yellow GR; RY�CorrespondE-mail addr

Prog. Polym. Sci. 33 (2008) 399–447

www.elsevier.com/locate/ppolysci

Application of chitosan, a natural aminopolysaccharide, for dyeremoval from aqueous solutions by adsorption processes using

batch studies: A review of recent literature

Gregorio Crini�, Pierre-Marie Badot

Department of Chrono-Environment, University of Franche-Comte, UMR UFC/CNRS 6565, Place Leclerc, 25000 Besanc-on, France

Received 21 December 2006; received in revised form 9 November 2007; accepted 9 November 2007

Available online 17 November 2007

Abstract

Application of chitinous products in wastewater treatment has received considerable attention in recent years in the

literature. In particular, the development of chitosan-based materials as useful adsorbent polymeric matrices is an

expanding field in the area of adsorption science. This review highlights some of the notable examples in the use of chitosan

and its grafted and crosslinked derivatives for dye removal from aqueous solutions. It summarizes the key advances and

results that have been obtained in their decolorizing application as biosorbents. The review provides a summary of recent

information obtained using batch studies and deals with the various adsorption mechanisms involved. The effects of

parameters such as the chitosan characteristics, the process variables, the chemistry of the dye and the solution conditions

used in batch studies on the biosorption capacity and kinetics are presented and discussed. The review also summarizes and

attempts to compare the equilibrium and kinetic models, and the thermodynamic studies reported for biosorption onto

chitosan.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Chitosan; Biosorption; Dyes; Batch process; Modeling and thermochemistry of biosorption

e front matter r 2007 Elsevier Ltd. All rights reserved.

ogpolymsci.2007.11.001

AB, acid blue; AB 1, acid black 1; AB 15, acid blue 15; AB 25, acid blue 25; AB 40, acid blue 40; AB 62, acid blue 62; AB

13; AG 25, acid green 25; AG 27, acid green 27; AO 7, acid orange 7; AO 10, acid orange 10; AO 12, acid orange 12; AO

51; AR, acid red; AR 1, acid red 1; AR 14, acid red 14; AR 18, acid red 18; AR 73, acid red 73; AR 27, acid red 27; AR 87,

88, acid red 88; AR 138, acid red 138; AV 5, acid violet 5; AY 25, acid yellow 25; BB, basic blue; BB 1, basic brown 1; BB

BB 9, basic blue 9; BR, brilliant red M5BR2; BY 45, basic yellow 45; CV, crystal violet; DB, direct blue; DB 14, direct blue

ct blue 71; DO, direct orange; DR, direct red; DR 2, direct red 2; DR 81, direct red 81; DS, direct scarlet B; DY 4, direct

digo carmine; IR, iragalon rubine RL; MB, maxilon blue 4GL; MB 29, mordant blue 29; MB 33, mordant brown 33; MO,

MO 10, mordant orange 10; MY, metanil yellow; MY 30, mordant yellow 30; O II, orange II; Rb 5, reactive blue 5; RB,

; RB 5, reactive black 5; RB 2, reactive blue 2; RB 15, reactive blue 15; RB 19, reactive blue 19; RB 222, reactive blue 222;

ange; RO 16, reactive orange 16; R 6G, rhodamine 6G; RR, reactive red; RR B, reactive red RB; RR 2, reactive red 2; RR

d 141; RR 189, reactive red 189; RR 195, reactive red 195; RR 222, reactive red 222; RTB, reactive T-blue; RY, reactive

2, reactive yellow 2; RY 86, reactive yellow 86; RY 145, reactive yellow 145.

ing author. Tel.: +333 81 66 57 01; fax: +33 3 81 66 57 97.

ess: [email protected] (G. Crini).

www.elsevier.com/locate/ppolysci

dx.doi.org/10.1016/j.progpolymsci.2007.11.001

mailto:[email protected]

ARTICLE IN PRESSG. Crini, P.-M. Badot / Prog. Polym. Sci. 33 (2008) 399–447400

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400

2. General considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402

2.1. Batch experiments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402

2.2. Why to use chitosan as raw material? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403

2.3. Considerations on dye adsorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405

2.4. Why to use chitosan as a biosorbent for dye removal? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407

2.5. Raw chitosan and chitosan-based materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409

3. A brief review of the recent literature on the adsorption of dyes by chitosan . . . . . . . . . . . . . . . . . . . . . . . 412

4. Control of adsorption performances of chitosan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

4.1. Influence of the chitosan characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

4.1.1. Chitosan origin. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

4.1.2. Physical nature of the chitosan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

4.1.3. Chemical structure of chitosan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 418

4.2. Activation conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419

4.2.1. Chitosan preprotonation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419

4.2.2. Grafting reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419

4.2.3. Influence of crosslinking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420

4.2.4. Chitosan-based composite beads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422

4.3. Influence of process variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422

4.3.1. Effect of chitosan dosage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422

4.3.2. Effect of initial dye concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 422

4.3.3. Effect of contact time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423

4.3.4. Effect of stirring rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424

4.3.5. Effect of dryness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424

4.4. Chemistry of the dye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424

4.5. Effect of the solution conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425

4.5.1. Effect of pH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425

4.5.2. Effect of pH variation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427

4.5.3. pH sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427

4.5.4. Effect of ionic strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427

4.5.5. Effect of competitive molecules and ions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428

4.6. Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428

4.7. Desorption of dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

5. Adsorption mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 429

6. Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

6.1. Equilibrium isotherm models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432

6.2. Kinetic modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

6.3. Thermochemistry of biosorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

6.3.1. Effect of temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

6.3.2. Thermodynamic parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440

7. Economic aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

8. Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444

1. Introduction

Many industries, such as textile, paper, plasticsand dyestuffs, consume substantial volume of water,and also use chemicals during manufacturing anddyes to color their products. As a result, theygenerate a considerable amount of polluted waste-

water [1–5]. For example, pulp and paper millsgenerate varieties of pollutants depending upon thetype of the pulping process. Their toxic effluentsare a major source of aquatic pollution and willcause considerable damage to the receiving waters ifdischarged untreated [1]. This specific type ofpollution is characterized by high biochemical

ARTICLE IN PRESS

Nomenclature

aL Langmuir isotherm constant (l/mg)C intercept of the intraparticle diffusion

equation (mg/g)Ce liquid-phase dye concentration at equili-

brium (mg/l)Co initial dye concentration in liquid phase

(mg/l)DG Gibbs free energy change (kJ/mol)DH enthalpy change (kJ/mol)DS entropy change (J/molK)Ea activation energy (kJ/mol)KF Freundlich isotherm constant (l/g)KL Langmuir isotherm constant (l/g)k0 frequency factor (min�1)k1 equilibrium rate constant of pseudo-first-

order adsorption (min�1)

k2 equilibrium rate constant of pseudo-second-order adsorption (g/mgmin)

ki intraparticle diffusion rate constant(mg/gmin�1/2)

qe amount of dye adsorbed at equilibrium(mg/g)

qt amount of dye adsorbed at time t (mg/g)qmax maximum adsorption capacity of the

adsorbent (mg/g)m mass of adsorbent used (g)nF Freundlich isotherm exponentR universal gas constant (8.314 J/molK)T absolute temperature (1K)t time (min)te equilibrium time (min)V volume of dye solution (l)x amount of dye adsorbed (mg)

Table 1

Principal existing and emerging processes for dyes removal

Conventional treatment

processes

� Coagulation/floculation

� Precipitation/floculation

� Electrocoagulation/

electroflotation

� Biodegradation

� Adsorption on activated carbon

Established removal

methods

� Oxidation

� Electrochemical treatment

� Membrane separation

� Ion-exchange

� Incineration

Emerging recovery

technologies

� Advanced oxidation

� Selective bioadsorption

� Biomass

G. Crini, P.-M. Badot / Prog. Polym. Sci. 33 (2008) 399–447 401

oxygen demand (BOD), chemical oxygen demand(COD), suspended solids (mainly fibers), bad smell,toxicity (high concentration of nutrients, presenceof chlorinated phenolic compounds, sulfur andlignin derivatives, etc.), and especially color [1,2].Color is the first contaminant to be recognizedin wastewater and the presence of very smallamounts of dyes in water is highly visible andundesirable [4,5].

During the past three decades, several wastewatertreatment methods have been reported and at-tempted for the removal of pollutants from textile,pulp and paper mill effluents. The technologies canbe divided into three main categories: (i) conven-tional methods, (ii) established recovery processesand (iii) emerging removal methods (see Table 1). Inthe literature, there are a great number of feasibilitystudies concerning the treatment of dyeing effluentsby these methods [2–8].

It is known that wastewaters containing dyes arevery difficult to treat, since the dyes are recalcitrantmolecules (particularly azo dyes), resistant toaerobic digestion, and are stable to oxidizing agents.Another difficulty is treatment of wastewaterscontaining low concentrations of dye molecules. Inthis case, common methods for removing dyes areeither economically unfavorable and/or technicallycomplicated. Because of the high costs associatedwith their practical applications to remove traceamounts of impurities, many of the methods fortreating dyes in wastewater (Table 1) have not been

widely applied on a large scale in the paper andtextile industries. In practice, no single processprovides adequate treatment and a combination ofdifferent processes is often used to achieve thedesired water quality in the most economical way.Thus, there is a need to develop new decolorizationmethods that are effective and acceptable inindustrial use.

It is now recognized that adsorption usinglow-cost adsorbents is an effective and economicmethod for water decontamination. A large varietyof non-conventional adsorbents materials have been


proposed and studied for their ability to removedyes [6]. However, low-cost adsorbents with highadsorption capacities are still under development toreduce the adsorbent dose and minimize disposalproblems. Much attention has recently been focusedon various biosorbent materials such as fungal orbacterial biomass and biopolymers that can beobtained in large quantities and that are harmless tonature. Special attention has been given to poly-saccharides such as chitosan, a natural aminopoly-mer. It is clear from the literature that thebiosorption of dyes using chitosan is one of themore frequently reported emerging methods forthe removal of pollutants.

Chitosan has been investigated by several re-searchers as a biosorbent for the capture ofdissolved dyes from aqueous solutions. This naturalpolymer possesses several intrinsic characteristicsthat make it an effective biosorbent for the removalof color. Its use as a biosorbent is justified by twoimportant advantages: firstly, its low cost comparedto commercial activated carbon (chitosan is derivedby deacetylation of the naturally occurring biopo-lymer chitin which is the second most abundantpolysaccharide in the world after cellulose); sec-ondly, its outstanding chelation behavior (one of themajor applications of this aminopolymer is basedon its ability to tightly bind pollutants, in particularheavy metal ions).

In this paper, we review the use of chitosan fordye removal from aqueous solutions. Since thereview only presents data obtained using raw,grafted and crosslinked chitosans, the discussionwill be limited to these chitosan-based materials andtheir adsorption properties. The main objectives areto summarize some of the developments related tothe decolorizing applications of these polymericmaterials and to provide useful information abouttheir most important features. We give an overviewof several recent batch studies reported in theliterature, with the various mechanisms involved.To do so, an extensive list of recent literature hasbeen compiled. The effects of various parameterssuch as chitosan’s characteristics, the activationconditions, the process variables, the chemistry ofthe dye and the experimental conditions used inbatch systems, on biosorption are presented anddiscussed. The review also summarizes the equili-brium and kinetic models, and the thermodynamicstudies reported for biosorption onto chitosan,which are important to determine the biosorptioncapacity and to design treatment processes.

2. General considerations

2.1. Batch experiments

The change in the concentration of a pollutant(adsorbate) in the surface layer of the material(adsorbent) in comparison with the bulk phase withrespect to unit surface area is termed adsorption.The term ‘‘biosorption’’ is given to adsorptionprocesses, which use biomaterials as adsorbents(or biosorbents). The assessment of a solid-liquidadsorption system is usually based on two types ofinvestigations: batch adsorption tests and dynamiccontinuous-flow adsorption studies. The presentreview only presents data obtained using batchstudies. When studying adsorption from solutionson materials it is convenient to differentiate between‘‘adsorption from dilute solution’’ and ‘‘adsorptionfrom binary and multicomponent mixtures coveringthe entire mole fraction scale’’. To judge by thenumber of papers published annually on adsorptionfrom dilute solution, this subject is more importantthan adsorption from binary mixtures. Therefore,reference will be made hereafter to adsorption fromdilute aqueous solutions.

Batch studies use the fact that the adsorptionphenomenon at the solid/liquid interface leads to achange in the concentration of the solution.Adsorption isotherms are constructed by measuringthe concentration of adsorbate in the mediumbefore and after adsorption, at a fixed temperature.For this, in general, adsorption data includingequilibrium and kinetic studies are performed usingstandard procedures consisting of mixing a fixedvolume of dye solution with an known amount ofchitosan in controlled conditions of contact time,agitation rate, temperature and pH. At predeter-mined times, the residual concentration of the dye isdetermined by spectrophotometry at the maximumabsorption wavelength. Dye concentrations in solu-tion can be estimated quantitatively using linearregression equations obtained by plotting a calibra-tion curve for each dye over a range of concentra-tions. The adsorption capacity (adsorption uptakerate) is then calculated and is usually expressed inmilligrams of dye adsorbed per gram of the (dry)adsorbent. For example, the amount of dyeadsorbed at equilibrium, qe, is calculated fromthe mass balance equation given by Eq. (1). Thesymbols used in the equation are defined in theNomenclature section. In general, the experi-ments are conducted in triplicate under identical

ARTICLE IN PRESSG. Crini, P.-M. Badot / Prog. Polym. Sci. 33 (2008) 399–447 403

conditions and found reproducible:

qe ¼V ðCo � CeÞ

m. (1)

The equilibrium relationship between adsorbentand adsorbate, i.e. the distribution of dye moleculesbetween the solid adsorbent phase and the liquidphase at equilibrium, which are the basic require-ments for the design of adsorption systems, aredescribed by adsorption isotherms using any of themathematical models available. The equilibriumadsorption isotherm, usually the ratio between thequantity adsorbed and that remaining in solution ata fixed temperature at equilibrium, is fundamentallyimportant since the equilibrium studies give thecapacity of the adsorbent and describe the adsorp-tion isotherm by constants whose values express thesurface properties and affinity of the adsorbent (i.e.to study the interaction between the adsorbate andthe surface and to know about the structure of theadsorbed layer).

In the literature, batch methods are widely usedto describe the adsorption capacity and the adsorp-tion kinetics. These processes are cheap and simpleto operate and, consequently, often favoured forsmall- and medium-size process applications usingsimple and readily available mixing tank equipment.

O

NHCOCH3

OH

CH2OH

O

n

Chitin

DA

O

NHCOCH3

OH

CH2OH

O

Commercial CN-acetyl glucosamine unit

Fig. 1. Chemical structure of chitin [poly(N-acetyl-b-D-glucosamine)], chi

characterized by its average degree of acetylation (DA)).

Simplicity, well-established experimental methods,and easily interpretable results are some of theimportant reasons frequently evoked for the ex-tensive usage of these methods. Another interestingadvantage is the fact that, in batch systems, theparameters of the solution such as adsorbentconcentration, pH, ionic strength, temperature,etc. can be controlled and/or adjusted.

2.2. Why to use chitosan as raw material?

The majority of commercial polymers and ion-exchange resins are derived from petroleum-basedraw materials using processing chemistry that is notalways safe or environmental friendly. Today, thereis growing interest in developing natural low-costalternatives to synthetic polymers [6].

Chitin, found in the exoskeleton of crustaceans,the cuticles of insects, and the cells walls of fungi, isthe most abundant aminopolysaccharide in nature[9–11]. This low-cost material is a linear homo-polymer composed of b(1-4)-linked N-acetyl gluco-samine (Fig. 1). It is structurally similar to cellulose,but it is an aminopolymer and has acetamide groupsat the C-2 positions in place of the hydroxyl groups.The presence of these groups is highly advantageous,

O

NH2

OH

CH2OH

O

n

Chitosan

1-DAO

CH2OH

NH2

OH

hitosanglucosamine unit

O

tosan [poly(D-glucosamine)] and commercial chitosan (a copolymer


providing distinctive adsorption functions andconducting modification reactions. The raw poly-mer is only commercially extracted from marinecrustaceans primarily because a large amount ofwaste is available as a by-product of food proces-sing [9]. Chitin is extracted from crustaceans(shrimps, crabs, squids) by acid treatment todissolve the calcium carbonate followed by alkalineextraction to dissolve the proteins and by adecolorization step to obtain a colorless product[10,11] (Fig. 2).

Since the biodegradation of chitin is very slow incrustacean shell waste, accumulation of largequantities of discards from processing of crusta-ceans has become a major concern in the seafoodprocessing industry. So, there is a need to recyclethese by-products. Their use for the treatment ofwastewater from another industries could be helpfulnot only to the environment in solving the solidwaste disposal problem, but also to the economy.However, chitin is an extremely insoluble material.Its insolubility is a major problem that confronts thedevelopment of processes and uses of chitin [11],and so far, very few large-scale industrial uses havebeen found. More important than chitin is itsderivative, chitosan (Fig. 1).

Shellfish was

dem

depr

deco

hydrolysis

glucosamines

oligosaccharides

Chitin

deac

Chitosan

salts

acetylation

oligosaccharides

glucosamines

N-acetyl-D-glucosamines

Fig. 2. Simplified representation of preparation

Partial deacetylation of chitin results in theproduction of chitosan (Fig. 2), which is apolysaccharide composed by polymers of glucosa-mine and N-acetyl glucosamine. The ‘‘chitosanlabel’’ generally corresponds to polymers with lessthan 25% acetyl content. The fully deacetylatedproduct is rarely obtained due to the risks of sidereactions and chain depolymerization. Copolymerswith various extents of deacetylation and grades arenow commercially available. Chitosan and chitinare of commercial interest due to their highpercentage of nitrogen compared to syntheticallysubstituted cellulose. Chitosan is soluble in acidsolutions and is chemically more versatile thanchitin or cellulose. The main reasons for this areundoubtedly its appealing intrinsic properties, asdocumented in a recent review [11], such asbiodegradability, biocompatibility, film-formingability, bioadhesivity, polyfunctionality, hydrophi-licity and adsorption properties (Table 2). Most ofthe properties of chitosan can be related to itscationic nature [9–12], which is unique amongabundant polysaccharides and natural polymers.These numerous properties lead to the recognitionof this polyamine as a promising raw material foradsorption purposes.

tes

ineralization

oteinization

loration

etylation carb oxymethylchitin

carb oxymethylation

chitosan derivatives

derivatization

of chitin, chitosan and their derivatives.

ARTICLE IN PRESS

Table 2

Intrinsic properties of chitosan

Physical and

chemical properties

� Linear aminopolysaccharide with

high nitrogen content

� Rigid D-glucosamine structure; high

crystallinity; hydrophilicity

� Capacity to form hydrogen bonds

intermolecularly; high viscosity

� Weak base; the deprotonated amino

group acts a powerful nucleophile

(pKa 6.3)

� Insoluble in water and organic

solvents; soluble in dilute aqueous

acidic solutions

� Numerous reactive groups for

chemical activation and crosslinking

� Forms salts with organic and

inorganic acids

� Chelating and complexing properties

� Ionic conductivity

Polyelectrolytes (at

acidic pH)

� Cationic biopolymer with high

charge density (one positive charge

per glucosamine residue)

� Flocculating agent; interacts with

negatively charged molecules

� Entrapment and adsorption

properties; filtration and separation

� Film-forming ability; adhesivity

� Materials for isolation of

biomolecules

Biological

properties

� BiocompatibilityJ Non-toxicJ BiodegradableJ Adsorbable

� BioactivityJ Antimicrobial activity (fungi,

bacteria, viruses)J Antiacid, antiulcer, and

antitumoral propertiesJ Blood anticoagulantsJ Hypolipidemic activity

� Bioadhesivity


The interest in chitin and chitosan is reflected byan increasing number of articles published (Fig. 3),and of meetings in Europe, Asia and America onthis topic. Table 3 summarizes the main applica-tions of chitin and chitosan. Currently, thesepolymers and their numerous derivatives are widelyused in pharmacy [21,36,37], medicine [11,21,23–29],biotechnology [10,21,30], chemistry [21,31–34], cos-metics and toiletries [11,21], food technology [35],and the textile [21], agricultural [12,20,21], pulp and

paper industries [21] and other fields [21,38,39] suchas enology, dentistry and photography. The poten-tial industrial use of chitosan is widely recognized.These versatile materials are also widely applied inclarification and water purification, and water andwastewater treatment as coagulating [13–15], floc-culating [16,17] and chelating agents [19–22]. How-ever, despite a large number of studies on the use ofchitosan for pollutant recovery in the literature, thisresearch field has failed to find practical applica-tions on the industrial scale: this aspect will bediscussed later.

2.3. Considerations on dye adsorption

Synthetic dyes are an important class of recalci-trant organic compounds and are often found in theenvironment as a result of their wide industrial use.These industrial pollutants are common contami-nants in wastewater and are difficult to decolorizedue to their complex aromatic structure andsynthetic origin. They are produced on a largescale. Although the exact number (and also theamount) of the dyes produced in the world is notknown, there are estimated to be more than 100,000commercially available dyes. Many of them areknown to be toxic or carcinogenic.

Generally, dyes can be classified with regard totheir chemical structure (e.g. azo, anthraquinone,indigo, triphenylmethane), with regard to themethod and domain of usage (e.g. direct, reactive,chromic, metal-complexes, disperse, mordant, sul-fur, vat, pigments), and/or with regard to theirchromogen (e.g. n-p*, donor–acceptor, cyanine,polyenes). Mishra and Tripathy [40] proposed asimplified classification as follows: anionic (direct,acid and reactive dyes), cationic (basic) dyes andnon-ionic (disperse) dyes. As mentioned, there aremany structural varieties such as acidic, disperse,basic, azo, diazo, anthraquinone-based and metalcomplex dyes. Azo and anthraquinone colorants arethe two major classes of synthetic dyes andpigments. Together they represent about 90% ofall organic colorants.

Fig. 4 gives some examples of dyes currently usedin the textile industry. Reactive Black 5, a diazo dye,has two sulfonate groups and two sulfatoethylsul-fon groups in its molecular structure that havenegative charges in aqueous solution. Basic Blue 3, amonoxazine dye, possesses an overall positivecharge because it tends to ionize in solution. Theanthraquinonic dyes Reactive Blue 19 and Disperse

ARTICLE IN PRESS

4%7%

3%

28%

1%4%

53%

coagulation

precipitation

adsorption

membranes

flocculation

flotation

filtration

0

50

100

150

200

250

300

1998 1999 2000 2001 2002 2003 2004 2005

Num

ber

of art

icle

s

Fig. 3. A Scopus database literature survey of the wastewater applications of chitosan and chitin: (a) research articles published from 1998

to 2005 (the survey did not include patents) and (b) main domains of chitosan and chitin in the removal of pollutants from solutions.

G. Crini, P.-M. Badot / Prog. Polym. Sci. 33 (2008) 399–447406

Blue 14 have an anionic and non-ionic character,respectively. Basic Green 4 is an N-methylateddiaminotriphenyl methane dye, which has a cationiccharacter. It is important to note that dye moleculeshave many different and complicated structures,and their adsorption behavior is directly related tothe chemical structure, the dimensions of the dyeorganic chains, and the number and positioning ofthe functional groups of the dyes. This is one of themost important factors influencing adsorption.However, to the weay adsorption is affected bythe chemical structure of the dyes was not clearlyidentified: this aspect will be discussed in thefollowing sections.

Generally, a suitable adsorbent for adsorptionprocess of dye molecules should meet severalconditions:

�
low cost, � readily available,
�
large capacity and rate of adsorption, � high selectivity for different concentrations, � and efficient for removal of a wide variety of
target dyes.

Recently, numerous low-cost adsorbents havebeen proposed for dye removal. Among them,non-conventional activated carbons from solidwastes, industrial by-products, agricultural solidwastes, clays, zeolites, peat, polysaccharides andfungal or bacterial biomass deserve particularattention as recently summarized in a review byCrini [6]. Each has advantages and drawbacks.However, at the present time, there is no singleadsorbent capable of satisfying the above require-ments. Thus, there is a need for new systems to bedeveloped. In addition, the adsorption processprovides an attractive alternative treatment, espe-cially if the adsorbent is selective and effective forremoval of anionic, cationic and non-ionic dyes.

ARTICLE IN PRESS

Table 3

Applications of chitin and chitosan

Fields Applications

Agriculture Protection of plants

Increase of crop yields (reduces the growth

of phytopathogenic fungi)

Seed and fertilizer coating; soil treatment

Biomedical

engineering

Biological activities (antifungal,

antimicrobial, antiinfectious); antitumor

agent

Hemostatic effects; enhances blood

coagulation

Promotes tissue growth; stimulates cell

proliferation; artificial skin

Sutures/bandages

Ophthalmology, contact lenses

Biotechnology Enzyme and cell immobilization

Cell-stimulating materials

Matrix for affinity chromatography or

membranes

Chemical

industry

Water purification (metal chelation); water

engineering (flocculation, filtration,

adsorption); sludge treatment

Reverse osmosis, filtration membranes; gas

separation

Production of biodegradable packaging

films

Catalysis

Cosmetics and

toiletries

Hair spray, lotion; hand and body creams;

shampoo, moisturizer

Food industry Diet foods and dietary fiber;

hypocholesterolemic activity (binds

cholesterol, fatty acids and

monoglycerides)

Preservation of foods from microbial

deterioration

Bioconversion for the production of value-

added food products

Recovery of waste material from food-

processing discards

Clarification and deacidification of fruit

juices and beverages

Emulsifying agent; colour stabilization

Animal feed additive

Pharmaceutics Controlled drug delivery carriers

Microcapsules (forming gels and capsules

with anionic polymers)

Dermatological products (treats acne)

Others Textiles (anti-bacterial properties)

Pulp and paper (wet strength)

Enology (clarification, deacidification)

Dentistry (dental implants)

Photography (paper)


Now, the amounts of dyes adsorbed on the aboveadsorbents are not very high, some have capacitiesbetween 100 and 600mg/g and some even lowerthan 50mg/g [6]. To improve the efficiency andselectivity of the adsorption processes, it is essentialto develop more effective and cheaper adsorbentswith higher adsorption capacities.

2.4. Why to use chitosan as a biosorbent for dye

removal?

As already mentioned, a growing number ofpapers have been published since the 1980s con-cerning chitosan for wastewater treatment. Inparticular, chitosan has received considerable inter-est in heavy metal chelation due to its relatively lowcost compared with commercial activated carbon,its excellent metal-binding capacities and interestingselectivity, as well as its possible biodegradabilityafter use. It is frequent to reach adsorptioncapacities as high as 3mmol metal per gramchitosan for Cu (i.e. 200mg/g), 1–2mmol metalper gram for Pt and Pd, and up to 7–10mmol metalper gram for Mo and V [18,19]. In accordance withthe very abundant data in the literature, liquid-phase adsorption using chitosan is one of the mostpopular methods for the removal of heavy metalsfrom wastewater since proper design of the adsorp-tion process will produce a high-quality treatedsolution. Readers interested in a detailed discussionof the interaction of metal ions with chitosan shouldrefer to the excellent comprehensive review byGuibal [18].

Besides being natural and plentiful, chitosanpossesses interesting characteristics that also makeit an effective biosorbent for the removal of colorwith outstanding adsorption capacities. Comparedwith conventional commercial adsorbents such ascommercial activated carbons (CAC) for removingdyes from solution, adsorption using chitosan-basedmaterials as biosorbents offers several advantages(Table 4). In particular, three factors have specifi-cally contributed to the growing recognition ofchitosan as a suitable biomaterial for dye removal:

�
First is the fact that the chitosan-based polymersare low-cost materials obtained from naturalresources and their use as biosorbents is extre-mely cost-effective. In many countries, fisherywastes were used as excellent sources to producechitosan. Since such waste is abundantly avail-able, chitosan may be produced at relatively low

ARTICLE IN PRESS

HO

H2N

N

N SO3Na

SO3Na

N

N

Reactive Black 5

NaO3SOCH2CH2O2S

NaO3SOCH2CH2O2S

O

N

(C2H5)2N N(C2H5)2

Cl-+

Basic Blue 3

O

O

NH2

SO3Na

HN

SO2CH2CH2OSO3NaReactive Blue 19

N

N(CH3)2

+ O

O

HO

-O

Basic Green 4

O

O

NHCH3

NHCH3

Disperse Blue 14

Fig. 4. Examples of commonly used dyestuffs in the textile industry.


cost. The volume of biosorbent used is alsoreduced as compared to conventional adsorbentssince they are more efficient.
� Second is the high adsorption capacities re-
ported. The biosorbents posses an outstandingcapacity and high rate of adsorption, and alsohigh selectivity in detoxifying both very dilutedor concentrated solutions. They also have anextremely high affinity for many varieties of dyes.
� The third factor is the development of new
complexing materials because chitosan is versa-tile: it can be manufactured into films, mem-branes, fibers, sponges, gels, beads andnanoparticles, or supported on inert materials.The utilization of these materials presents manyadvantages in terms of applicability to a widevariety of process configurations.

Of course, there are, also disadvantages of usingchitosan in wastewater treatment (Table 4). Thisresearch field fails to find practical application at theindustrial scale. There are several reasons forexplaining this difficulty in transferring the process

to industrial applications [10,11,18,20]. The adsorp-tion properties depend on the different sources ofchitin (the quality of commercial chitin available isnot uniform) and performance is also dependent onthe type of material used. Another importantcriterion to be taken into account concerns thevariability and heterogeneity of the polymer (thedifficulty of controlling the distribution of the acetylgroups along the backbone makes it difficult to getreproducible initial polymers). There is a need for abetter standardization of the production process tobe able to prepare reproducible initial polymershaving the same characteristics. Changes in thespecifications of the polymer may significantlychange adsorption performance. Another problemwith chitosan derivatives is their poor physicochem-ical characteristics, in particular low surface areaand porosity. In addition, although chitosan ismuch easier to process than chitin or other low-costadsorbents, the stability of chitosan materials isgenerally lower, owing to their more hydrophiliccharacter and, especially, pH sensitivity. Being abiopolymer, chitosan is biodegradable and this may

ARTICLE IN PRESS

Table 4

Advantages and disadvantages of chitosan and chitosan-based

materials used as biosorbent for the removal of dyes from

aqueous solutions

Advantages Disadvantages

� Low-cost hydrophilic

biopolymer

� Very abundant material

and widely available in

many countries

� Renewable resource

� Cationic polysaccharide (in

acidic medium)

� Environmentally friendly,

publicly acceptable

material

� Extremely cost effective

� Outstanding dye-binding

capacities of a wide range

of dyes

� Fast kinetics

� High selectivity in

decolorizing both very

dilute or concentrated

solutions

� Versatile biosorbent

� Variability in the polymer

characteristics

� The performance depends

of the origin and treatment

of the polymer, and also its

degree of N-acetylation

� Nonporous sorbent

� Requires chemical

derivatization to improve

its performance

� Not effective for cationic

dyes (except after

modification)

� pH sensitivity

� Its use in sorption columns

is limited (hydrodynamic

limitations and column

fouling)

� Non-destructive process


be a serious drawback for long-term applications.These problems can rebut industrial users. Readersinterested in a detailed discussion of these problemsshould refer to the work of Guibal [18]. However,the opportunity now exists to consider chitosan foremerging applications where other technologieswould be unsuitable.

Different reviews of chitosan-based biomaterialshave been reported concerning adsorption andseparation, including metal complexation [18,19],complexing adsorbent matrices [21,22,41,42], andmembranes [33]. Obviously, chitosan has also beeninvestigated as a biosorbent for the capture ofdissolved dyes from aqueous solutions in numerousarticles. The effectiveness of chitin and chitosan toadsorb dye molecules has been reported by numer-ous workers [43–57]. For example, as long ago as1958, Giles et al. [43] investigated the bindingbehavior of dyes to chitin. In 1982–1985, extensivestudies on the adsorption of dyes on chitin byMcKay et al. [44–48] also revealed that chitin canadsorb substantial quantities of dyestuffs fromaqueous solutions. The interaction of chitosan withdyes was studied by several workers [49–57]. Theseearlier papers clearly demonstrated that raw materi-als have an intrinsically high affinity and selectivity

for a wide range of dyes, although several contra-dictory observations have been reported. However,a few review articles on the potential of chitosan fordye removal have been published. The applicationof the adsorption of pollutants including dyes ontochitosan has been reviewed by Ravi Kumar [21] andNo and Meyers [22]. Various chitosan-based com-posites and membranes have been also developedand proposed for adsorption and separation pur-poses [33,42]. To avoid repetition, in the followingchapters, only raw, grafted and crosslinked chit-osans will be discussed. This review focuses on therecent developments related to decolorizing applica-tions of the chitosan-based materials and reports themain advances published over the last 10 years. Thisis an ambitious project since the very large numberof groups working around the world forces us tomake a selection from the most significant results.Table 5 lists some of the researchers whose resultsare discussed in this review and the dyes theyinvestigated [58–116].

2.5. Raw chitosan and chitosan-based materials

Practical use of chitosan has been mainlyconfined to the unmodified forms. For a break-through in its utilization, chemical derivatizationonto polymer chains has been proposed to producenew materials. Derivatization is a key point whichwill introduce the desired properties to enlarge thefield of its potential applications. Chitosan has threetypes of reactive functional groups, an amino groupas well as both primary and secondary hydroxylgroups at the C-2, C-3 and C-6 positions, respec-tively (Fig. 1). Its advantage over other polysac-charides is that its chemical structure allows specificmodifications without too many difficulties, espe-cially, at the C-2 position [11]. These functionalgroups allow direct substitution reactions andchemical modifications, yielding numerous usefulmaterials for different domains of application.The most commonly used chemical activations arecarboxymethylation, acetylation and grafting. Thevariety of groups which can be attached tothe polymer is almost unlimited. To control boththe physical, mechanical and chemical properties,various techniques can be used, and often, themethods are adapted from the cellulose world [11].The chitosan derivatives can be classified into fourmain classes of materials: modified polymers, cross-linked chitosans, chitosan-based composites andmembranes (Table 6).

ARTICLE IN PRESS

Table 5

Authors of recent research on dye removal by chitosan (selected papers)

Corresponding author Country Dye(s) Reference(s)

Airoldi C. Brazil BB 9 [58]

Annadurai G. Iran BB 9, DS [59,60]

Cestari AR. Brazil IC, RB, RN, RR, RY [61–63]

Chen DH. Taiwan AG 25, AO 12 [64]

Chen L. China AB, BB [65]

Chiou MS. Taiwan AO 7, AO 12, AR 14, DR 81 MY, RB 2, RB 15, RR 2,

RR 189, RR 222, RY 2, RY 86

[66–70]

Cho SY. Korea RB 5 [71]

Crini G. France BB 3, BB 9 [72,73]

de Favere VT. Brazil RO 16 [74]

Dutta PK. India DB [75]

El-Tahlawy KF. Egypt BR, IR, MB [76,77]

Fahmy HM. Egypt DR [78]

Guibal E. France AB 1, AB 113, AG 25, AV 5, AY 25, DB 14, DB 71,

DY 4, MB 29, MB 33, MO 10, RB 5

[79–82]

Guha AK. India AR 87 [83]

Hebeish R Egypt AR, BY 45, DO, RO [84,85]

Juang RS. Taiwan AO 51, BB 9, RB 222, RR 222, RY 145, R 6G [86–93]

Li HY. Taiwan RR 189 [94]

Martel B. France AB 15, AB 25, AB 62, DR 81, MY 30, RB 5, RB 19 [95]

Manolova N. Bulgaria RR [96]

McKay G. Hong Kong AG 25, AO 10, AO 12, AR 18, AR 73 [97–99]

Miyata K. Japan AB 40, AR 18, AR 88, DR 2 [100]

Prado AGS. Brazil IC [101]

Saha TK. Bangladesh azo dye [102]

Shimizu Y. Japan AO 7, AR 1, AR 88, AR 138, BB 9, CV [103–105]

Shyu SS. Taiwan BB 1, BB 3 [106]

Stevens WF. Thailand BB 9, CV, MO, O II [107,108]

Thiravetyan P. Thailand RR 141 [109]

Szeto YS. Hong Kong AG 27 [110,111]

Uzun I. Turkey CV, O II, Rb 5, RB 5, RY 2 [112–115]

Wen YZ. China RR 195 [116]


An important class of chitosan derivatives are thecrosslinked materials, from gel types to bead typesor particles (including microparticles, microspheresand nanoparticles). Gels are often divided into threeclasses depending on the nature of their network,namely entangled networks, covalently crosslinkednetworks and networks formed by physical interac-tions. Berger et al. [26] suggested the followingmodified classification for chitosan gels; i.e. aseparation of chemical and physical gels. Physicalgels are formed by various reversible links andchemical gels are formed by irreversible covalentlinks, as in covalently crosslinked chitosan gels.

Hydrogels and beads can be formed covalentlycrosslinking polymer with itself. In this chemicaltype of crosslinking reaction, the crosslinking agentsare molecules with at least two reactive functionalgroups that allow the formation of bridges bet-ween polymer chains. To date, the most common

crosslinkers used with chitosan are dialdehydes suchas glyoxal, formaldehyde and in particular glutar-aldehyde (GLU) [26]. GLU reacts with chitosan andit crosslinks in inter and intramolecular fashionthrough the formation of covalent bonds mainlywith the amino groups of the polymer. Its reactionwith chitosan is very well documented. The maindrawback of GLU is that it is considered to betoxic, even if the presence of free unreacted GLU ingels is improbable since the materials are purifiedbefore use. Other crosslinkers of chitosan areepoxides such as epichlorohydrin (EPI) and ethy-lene glycol diglycidyl ether (EGDE), isocyanates(hexamethylenediisocyanate) and other agents (car-boxylic acids, genipin). Covalent crosslinking, andtherefore the crosslinking density, is influenced byvarious parameters, but mainly dominated by theconcentration of crosslinker. It is favoured whenchitosan molecular weight (MW) and temperature

ARTICLE IN PRESS

Table 6

The four main classes of chitosan derivatives

I. Modified polymers

� Carboxymethylchitosans

� Alkylated chitosans

� Chitosan sulfate derivatives

� Carbohydrate-branched chitosans

� Grafted chitosans

� Ligand-bound chitosan

II. Crosslinked chitosan

� Covalently crosslinked particles

� Ionically crosslinked particles

� Nanoparticles

� Physical gels

III. Chitosan-based composites

� Chitosan-dendrimer hybrids

� Chitosan-supported on inert materials (silica gel, glass beads,

alumina, etc.)

IV. Membranes

Table 7

Some methods for preparation of chitosan particles

Crosslinking with chemicals

� (Single) emulsion crosslinking

� Multiple emulsion

� Precipitation/crosslinking

Crosslinking and interactions with charged ions, molecules and

polymers

� Ionotropic gelation

� Wet-phase inversion

� Emulsification and ionotropic gelation

� Emulsification and solvent evaporation

� Simple or complex coacervation (precipitation, complexation)

Miscellaneous methods

� Thermal crosslinking

� Solvent evaporation method

� Neutralization method

� Spray drying

� Freeze drying

� Reverse micellar

� Coating

� Interfacial acylation


increased. Moreover, since crosslinking requiresmainly deacetylated reactive units, a high degreeof deacetylation (DD) of chitosan is favorable.

The crosslinked polymeric materials have a three-dimensional network structure and can swell con-siderably in aqueous medium without dissolution.Their synthesis and properties have been recentlydescribed in detail [41]. Various methods have beendeveloped for the chemical crosslinking of chitosan,which commonly result in gel formation. Thecrosslinking step is a well-documented reactionand an easy method to prepare chitosan-basedmaterials with relatively inexpensive chemicals.

Generally, a crosslinking step is required toimprove mechanical resistance and to reinforce thechemical stability of the chitosan in acidic solutions,modifying hydrophobicity and rendering it morestable at drastic pH, which are important features todefine a good adsorbent. However, it decreases thenumber of free and available amino groups on thechitosan backbone, and hence the possible liganddensity and the polymer reactivity. It also decreasesthe accessibility to internal sites of the material andleads to a loss in the flexibility of the polymerchains. So, the chemical step may cause a significantdecrease in dye uptake efficiency and adsorptioncapacities, especially in the case of chemical reac-tions involving amine groups, since the aminogroups of the polymers are much more activethan the hydroxyl groups and can be much moreeasily attacked by crosslinkers. Consequently, it is

important to know, control and characterize theconditions of the crosslinking reaction since theydetermine and allow the modulation of the cross-linking density, which is the main parameterinfluencing interesting properties of gels [26]. Theseconditions are useful for a better comprehension ofthe adsorption mechanisms. For example, the lossin flexibility of the polymer resulting from thecrosslinking may explain some diffusion restric-tions, and the decrease observed in the intraparticlediffusivity.

Table 7 outlines various methods and approacheswhich have been proposed for the preparation ofchitosan particles including microspheres/micropar-ticles, and nanoparticles. Selection of any of themethods depends upon factors such as particlesize requirement, thermal and chemical stability. Inpractice, the methods are often combined anddifferent follow-up treatments are carried out [33].The emulsion crosslinking method is widely used forthe synthesis of microspheres. This method isschematically represented in Fig. 5. With thismethod, the size of the particles can be controlledby modifying the size of the aqueous droplets.Another interesting method is spray drying. This isa complex operation with the movement of count-less droplets/particles in turbulent drying mediumflows under changing temperature and humidity

ARTICLE IN PRESS

hardening ofdroplets

chitosan aqueoussolution oil phase

emulsification

crosslinking agent

stirring

particles

separation

Fig. 5. Schematic representation of preparation of chitosan

particles by emulsion crosslinking.


conditions. Chitosan microspheres obtained by thistechnique are characterized by a high degree ofsphericity and specific surface area, parameters thatare important for application as adsorbents.

Ionic crosslinking reactions have also beenemployed by using ionotropic gelation to formhydrogels, beads and nanoparticles. Aside from itscomplexation with negatively charged ions ormolecules, an interesting property of chitosan is itsability to gel on contact with specific polyanions.This gelation process is due to formation of interand intramolecular crosslinks mediated by thesepolyanions. Tripolyphosphate (TPP) is commonlyused to provoke the ionotropic gelation of chitosan.The particles can be obtained by the addition of achitosan solution to a solution of TPP or vice versa,under strirring. In either case, the size of theparticles is strongly dependent on the concentrationof the solutions. Chiou and Li [68] and Szeto’sgroup [110,111] recently reported the ionotropicgelation of chitosan with TPP. They preparedchitosan particles by adding an alkaline phasecontaining TPP into an acidic phase containingchitosan. (Nano)particles are formed immediatelyupon mixing the two phases through molecularlinkages created between TPP phosphates andchitosan amino groups. The solution of TPP wasused to produce more rigid materials. They reportedthat TPP had no effect on dye adsorption. To

stabilize chitosan in acid solutions, Chiou and Li[68] also proposed an ionotropic gelation processfollowed by a chemical crosslinking step.

Chitosan is usually used in a flaked or powderedform that is both soluble in acidic media and non-porous. Moreover, the low internal surface area ofthe non-porous polymer limits access to interioradsorption sites and hence lowers dye adsorptioncapacities and adsorption rates. To overcome thisobstacle, porous beads were synthesized. Indeedan interesting characteristic of the chitosan is itsexcellent ability to be processed into porousstructures.

3. A brief review of the recent literature on the

adsorption of dyes by chitosan

There is abundant literature concerning theevaluation of adsorption performances of rawchitosan, especially in terms of adsorption capacity(amount of dye adsorbed) or uptake. In a batchsystem, the determination of the dye uptake rate bya chitosan-based material is often based on theequilibrium state of the adsorption system. At least100 dyes, mainly anionic dyes, have been so farstudied. Chitosan has an extremely high affinity formany classes of dyes (Table 8). In particular, it hasdemonstrated outstanding removal capacities foranionic dyes such as acid, reactive and direct dyes.This is due to its unique polycationic structure.

The effectiveness of chitosan for its ability tointeract with dyes has been studied by numerousworkers. Juang and co-workers [89–93] demon-strated the usefulness of chitosan for the removal ofreactive dyes. They found that the maximumadsorption capacities of chitosan for RR 222, RB222 and RY 145 were 1653, 1009 and 885mg/g,respectively [90]. Annadurai [59,60] and Crini et al.[72] also reported that chitosan may be a usefuladsorbent for the effluent of textile mills because ofits high adsorption capacity. Uzun and Guzel[112–115] noted that chitosan can be used in thestudies of dyestuff adsorption in comparison withmost other adsorbents. This polysaccharide showeda higher capacity for adsorption of dyes than CACand other low-cost adsorbents, as reviewed by Crini[6]. Kim and Cho [71] also indicated that theamount of RB 5 adsorbed on chitosan beads ismuch greater than on CAC. Similar conclusionswere reached by Lima et al. [58] for the BB 9adsorption. McKay’s group [97–99] recently pub-lished a series of papers on the ability of chitosan to

ARTIC

LEIN

PRES

STable 8

Results of batch studies for various dyes using chitosan

Dye Chitosan Effective pre-

treatment of

chitosan

Particle size Sspa pH T (1C) Equilibrium

time

Equilibrium

model

qmb Kinetic model Adsorption

mechanism

Reference

AB protonation 450–900mm 3.6 20 4 h 296 Diffusionc [65]

AG 25 Crab shell 355–500 mm 4 25 24 h Langmuir 645.1 Lagergren [97–99]

AG 25 Protonation 3 4 days Langmuir 525 [81]

AG 27 Nanoparticle 180 nm 25 24 h Langmuir 2103.6 [110]

AO 7 Bead (crab) Crosslinking 4 30 5 days Langmuir 1940 Ho and McKay [67]

AO 10 Crab shell 355–500 mm 4 25 24 h Langmuir 922.9 Lagergren [97–99]

AO 12 Bead (crab) Crosslinking 3 30 5 days Langmuir 1954 Ho and McKay [67]

AO 12 Crab shell 355–500 mm 4 25 24 h Langmuir 973.3 Lagergren [97–99]

AO 51 Wet bead 30 3 days Langmuir 656 Elovich Chemisorption [86]

AO 51 Dried bead 30 3 days Langmuir 494 Elovich Chemisorption [86]

AR 14 Bead (crab) Crosslinking 3 30 5 days Langmuir 1940 Ho and McKay [67]

AR 18 Crab shell 355–500 mm 4 25 24 h Langmuir 693.2 Lagergren [97–99]

AR 73 Crab shell 355–500 mm 4 25 24 h Langmuir 728.2 Lagergren [97–99]

AR 87 Bead (shrimp) 6 30 Langmuir 76 Ho and McKay Chemisorption [83]

BB Protonation 450–900 mm 9.6 20 4 h 50 Diffusionc [65]

BB 3 Powder (crab) Grafting 25 3 25 40min Langmuir 166.5 Ho and McKay Chemisorption [73]

BB 9 Wet bead 30 3 days Langmuir 222 Elovich Chemisorption [86]

BB 9 Dried bead 30 3 days Langmuir 202 Elovich Chemisorption [86]

BB 9 Powder (crab) Grafting 25 3 25 40min Langmuir 121.9 Chemisorption [72]

BB 9 0.177 10 9.5 60 24 h Lagergren Diffusionc [59]

BB 9 Grafting 0.99 5.5 25 3 h Langmuir [58]

DB 6 26 5 h Langmuir Lagergren [75]

DR 81 Bead (crab) Crosslinking 3 30 5 days Langmuir 2383 Ho and McKay [67]

DS 0.206 8.47 47.5 24 h 37.18 [60]

IC Shrimp shell 60–100 mesh 6 35 2 h Langmuir [63]

MY Bead (crab) Crosslinking 4 30 5 days 1334 Ho and McKay Diffusionc [66]

RB Bead Crosslinking 0.24 2 60–200min Avrami [62]

RB 2 Bead (crab) Crosslinking 3 30 5 days Langmuir 2498 Ho and McKay [67]

RB 5 3 2 days Langmuir 1100 [79]

RB 5 Flake Crosslinking 2mm 350 6 25 5 days Freundlich [71]

RB 15 Bead (crab) Crosslinking 4 30 5 days 722 Ho and McKay Diffusionc [66]

RB 222 Swollen bead 2.8mm 30 4 days Langmuir 1009 Ho and McKay Chemisorption [90]

RB 222 Flake 1–1.41mm 11.8 30 4 days Langmuir 199 Ho and McKay Chemisorption [90]

RB 222 Bead (lobster) 0.715mm 12.3 30 Diffusionc [89]

RO 16 Crosslinking 25 mm 2 25 24 h Langmuir 30.4 [74]

RR 2 Bead (crab) Crosslinking 3 30 5 days Langmuir 2422 Ho and McKay [67]

RR 141 Shrimp shell 850 mm–1mm 11 60 24 h Langmuir 156 [109]



RR 189 Bead Crosslinking 2.3–2.5mm 3 30 5 days Langmuir 1936 Ho and McKay Chemisorption [94]

RR 189 Bead Crosslinking 2.3–2.5mm 3 30 5 days Langmuir 1834 Ho and McKay Diffusionc [68]



RR 189 Bead 2.3–2.5mm 6 30 5 days Langmuir 1189 Ho and McKay Chemisorption [94]

G.

Crin

i,P

.-M.

Ba

do

t/

Pro

g.

Po

lym

.S

ci.3

3(

20

08

)3

99

–4

47

413

ARTIC

LEIN

PRES

STable 8 (continued )

Dye Chitosan Effective pre-

treatment of

chitosan

Particle size Sspa pH T (1C) Equilibrium

time

Equilibrium

model

qmb Kinetic model Adsorption

mechanism

Reference

RR 189 2.3–2.5mm 6 30 5 days Langmuir 950 Ho and McKay Diffusionc [68]

RR 222 Bead Crosslinking 3 30 2 days Langmuir 2252 Ho and McKay Chemisorption [69]

RR 222 Bead 30 5 days Freundlich 1965 Lagergren Diffusionc [87]

RR 222 Swollen bead 2.8mm 30 4 days Langmuir 1653 Ho and McKay Chemisorption [90]

RR 222 Wet bead 30 3 days Freundlich 1498 Elovich Chemisorption [86]

RR 222 Dried bead 30 3 days Freundlich 1215 Elovich Chemisorption [86]

RR 222 Bead (crab) 3.11mm 30–40 30 5 days Langmuir 1106 Diffusionc [91]

RR 222 Bead (shrimp) 2.39mm 30–40 30 5 days Langmuir 1026 Diffusionc [91]

RR 222 Bead (lobster) 2.93mm 30–40 30 5 days Langmuir 1037 Diffusionc [91]

RR 222 Flake (shrimp) 16–30 mesh 4–6 30 5 days Langmuir 494 Diffusionc [91]

RR 222 Flake (lobster) 16–30 mesh 4–6 30 5 days Langmuir 398 Diffusionc [91]

RR 222 Flake 1–1.41mm 11.8 30 4 days Langmuir 339 Ho and McKay Chemisorption [90]

RR 222 Flake (crab) 16–30 mesh 4–6 30 5 days Langmuir 293 Diffusionc [91]

RR 222 Bead Crosslinking 4.01 30 3 days Freundlich [88]

RR 222 Bead (lobster) 0.715mm 12.3 30 Diffusionc [89]

RY Bead Crosslinking 0.24 2 60–200min Avrami [62]

RY 2 Bead (crab) Crosslinking 4 30 5 days Langmuir 2436 Ho and McKay [67]

RY 86 Bead (crab) Crosslinking 3 30 5 days Langmuir 1911 Ho and McKay [67]

RY 145 Swollen bead 2.8mm 30 4 days Langmuir 885 Ho and McKay Chemisorption [90]

RY 145 Flake 1–1.41mm 11.8 30 4 days Langmuir 188 Ho and McKay Chemisorption [90]

RY 145 Bead (lobster) 0.715mm 12.3 30 Diffusionc [89]

aSpecific surface area in m2/g.bAdsorption capacities in mg/g.cIntraparticle diffusion model.

G.

Crin

i,P

.-M.

Ba

do

t/

Pro

g.

Po

lym

.S

ci.3

3(

20

08

)3

99

–4

47

414


act as an effective adsorbent for the removal of aciddyestuffs from aqueous solution. The monolayeradsorption (saturation) capacities were determinedto be 973.3, 922.9, 728.2 and 693.2mg of dye pergram of chitosan for AO 12, AO 10, AR 73 and AR18, respectively [99]. The interaction betweenchitosan and anionic dyes has also been intensivelyinvestigated by Guibal and co-workers [79–82].Their investigations clearly indicated that chitosanhad a natural selectivity for dye molecules and wasvery useful for the treatment of wastewater. Theyreported that adsorption capacities ranged between200 and 2000 mmol/g for chitosan and between 50and 900 mmol/g for CAC [82]. They concluded thatchitosan exhibited a twofold or more increase in theadsorption capacity compared to CAC in the case ofacid, direct, reactive and mordant dyes. The bestchoice for the adsorbent between CAC and chitosandepends on the dye, however, it was impossible todetermine a correlation between the chemicalstructure of the dye and its affinity for either carbonor chitosan.

It is evident from this brief literature survey thatchitosan can be utilized as an interesting tool for thepurification of dye-containing wastewater becauseof its outstanding adsorption capacity.

4. Control of adsorption performances of chitosan

The data from the literature show that the controlof adsorption performances of a chitosan-basedmaterial in liquid-phase adsorption depends on thefollowing factors:

(i)
the origin and nature of the chitosan such asits physical structure, chemical nature andfunctional groups;
(ii)
the activation conditions of the raw polymer(physical treatment, chemical modifications);
(iii)
the influence of process variables such ascontact time, initial dye concentration, polymerdosage and stirring rate;
(iv)
the chemistry of the dye (e.g. its pKa, polarity,MW size and functional groups);
(v)
and finally, the solution conditions, referring toits pH, ionic strength, temperature and presenceof impurities.
These aspects will be described in the following.However, the reader is encouraged to refer to theoriginal papers for complete information on experi-mental conditions in the batch studies used.

4.1. Influence of the chitosan characteristics

It is very important to note that tuning the chitosanmanufacturing process can ernable the production ofpolymers with varying chemical characteristics andMW distributions. As stated in the introduction,chitosan is a ‘‘collective term’’ applied to deacetylatedchitins in various stages of deacetylation anddepolymerization [37]. Commercial chitosan is usual-ly offered as flakes or powders. Products of variouscompanies differ in purity, salt-form, color, granula-tion, water content, DD or degree of acetylation(DA), amino group content, MW, crystallinity andsolubility [10–12,18]. These parameters determined bythe conditions selected during the preparation arevery important because they control the swelling anddiffusion properties of chitosan and also influence itscharacteristics [117]. In particular, numerous studieshave demonstrated that the MW and DD influencethe adsorption properties of this polymer. Therefore,these factors must be considered carefully during theadsorption optimization process.

4.1.1. Chitosan origin

From a practical viewpoint, crustaceans shells arethe potential sources for chitin production. Chit-osan is commonly prepared by deacetylating chitinusing 40–50% aqueous alkali at 110–115 1C for afew hours [12]. Chitin occurs in a wide variety ofspecies, from fungi to animals. Depending on thechitin source, chitosan varies greatly in its adsorp-tion properties and solution behavior, as reportedby Juang and co-workers [89–93]. For example, theadsorption capacities of RR 222 on different typesof chitosan prepared from three fishery wastes(shrimp, crab and lobster shells) were compared.The monolayer adsorption capacities were deter-mined to be 293, 398 and 494mg of dye per gram offlake-type of chitosan for crab, lobster and shrimp,respectively [91]. This demonstrates that the adsorp-tion capacity of chitosan depends on its origin.Rinaudo [11] also reported in a recent review thatthe origin of chitin influences not only its crystal-linity and purity but also its polymer chainsarrangement, and hance its properties. In particular,the chitin resulting from crustaceans needs to begraded in terms of purity and color since residualprotein and pigment can cause problems [10,11].

4.1.2. Physical nature of the chitosan

The adsorption capacity of chitosan also dependson its physical structural parameters such as


crytallinity, surface area, porosity, particle type,particle size and water content. These parametersare determined by the conditions selected during thepreparation and polymer conditioning.

Three crystalline forms are known for chitin:a-, b- and g-chitins. The most abundant and easilyaccessible form is a-chitin [11,91]. Chitosan is alsocrystalline and shows polymorphism depending onits physical state. Depending on the origin of thepolymer and its treatment during extraction fromraw resources, the residual crystallinity may varyconsiderably. Crystallinity is maximum for bothchitin (i.e. 0% deacetylated) and fully deacetylatedchitosan (i.e. 100%). Generally, commercial chit-osans are semi-crystalline polymers and the degreeof crystallinity is a function of the DD. Crystallinityplays an important role in adsorption efficiency asreported by Trung et al. [108]. They demonstratedthat decrystallized chitosan is much more effectivein the adsorption of anionic dyes. Crystallinitycontrols polymer hydratation, which in turn deter-mines the accessibility to internal sites. This para-meter strongly influences the kinetics of hydratationand adsorption. Dissolving the polymer breaks thehydrogen bonds between polymer chains. Thereduced polymer crystallinity can be maintainedthrough freeze-drying of the chitosan solution,while air-drying or oven-drying partially reestab-lishes polymer crystallinity. The conditioning of thepolymer and physical modification can stronglyreduce the influence of this important parameterand improve diffusion properties [18]. The gelformation procedure also allows an expansion ofthe polymeric network, a decrease in steric hin-drance phenomena and a decrease in the crystal-linity of raw materials which enhance masstransport. The case of dye adsorption with cross-linked chitosan is a typical example of the influenceof particle size. When crosslinked with GLU, thenetwork formed makes the sorption performancesbecome dependent on the size of particles. Thisdependence disappears when chitosan particles aremodified by gel formation. Hebeish et al. [84,85]indicated that the crosslinking step changes thecrystalline nature of chitosan and decrease theparticle size of the crystallites, enhancing itsadsorption capacity. The crosslinking reactiondestroys the crystalline structure at low levels ofcrosslinking. The authors assumed that moreaccessible domains are created as a result of changesin the physical and chemical structures of chitosanduring the modification by GLU, and consequently

these effects increased dye adsorption [85]. How-ever, Cestari et al. [62] recently noted that after thecrosslinking reaction, there is a small increase in thecrytallinity of chitosan beads with increased accessto the small pores of the material.

Among the other parameters that have a greatimpact on dye adsorption is particle type. Chitosancan be presented as gels, flakes, powders andparticles. Chitosan beads are preferred since flakeand powder forms of polymer are not suitable foruse as adsorbents due to their low surface area andlack of porosity, as indicated by Varma et al. [19].Beads are usually prepared by dropping high-viscosity chitosan salt solutions into a basic solutionwith slow stirring. The diameters of the drops aswell as the solution flow rate control the diameter ofthe beads. Wu et al. [91] reported that bead-typechitosan gives a higher capacity for dye adsorptionthan the flake type by a factor of 2–4 depending onthe source of fishery waste. For example, acomparison of the maximum adsorption capacity(qmax) for RR 222 by chitosan flakes and beadsprepared from a crab source showed 293mg/g forflakes and 1103mg/g for beads. The authorsexplained this result by the fact that the beadspossessed a greater surface area (i.e., more loosepore structure) than the flakes. They also reportedthat the adsorption capacity of chitosan depends onits source. The qmax were determined to be 1106,1037 and 1026mg of dye per gram of bead-type ofchitosan for crab, lobster and shrimp, respectively[91]. Again, it can be noted that the order of qmax forthe different sources is exactly identical to that ofthe surface area of the whole animal, i.e., crab4lobster4shrimp. Chang and Juang [86] also notedthat chitosan in the bead form significantly im-proves the adsorption performance of RR 222, AO51 and BB 9 compared to that in the flake form.Guibal et al. [82] indicated that it would beinteresting to use chitosan gel beads instead offlakes since the production of gel beads decreasesthe residual crystallinity of polymer which enhancesboth the porosity and the diffusion properties of thematerial, due to the expansion of the chitosannetwork and the increase in the specific surface area.Crini et al. [72] observed that compared to chitosanflakes, chitosan beads exhibited a twofold or moreincrease in the adsorption capacity for BB 9. One ofchitosan’s most promising features is its excellentability to be processed into nanostructures. Thesenanochitosans can also be used in batch studies, asreported by Hu et al. [110]. They noted that an


adsorption capacity of 2103.6mg of AG 27 pergram chitosan was achieved, which was significantlyhigher than that of the chitosan microparticles.

Previously, it has been demonstrated that theparticle size of chitosan also influences its adsorp-tion profile. For example, Park et al. [56] showedthat of the smaller particle size, the more dye wasabsorbed. As adsorption is a surface phenomenon,this can be attributed to the relationship betweenthe effective specific surface area of the adsorbentparticles and their sizes. The surface area valuesusually increased as the particle size decreased and,as a consequence, the saturation capacity per unitmass of adsorbent increased. Decreasing the size ofparticles improves the adsorption properties of thechitosan, especially when chitosan is crosslinked.However, small particle sizes are not compatiblewith large-scale applications. For example, in fixed-bed columns, small particles are inappropriate sincethey induce head loss and column blocking andcause serious hydrodynamic limitations [32]. Thereare a large number of studies that highlight thecorrelation between adsorption performance andsize of particle. Annadurai [59,60] used chitosan forthe removal of basic and direct dye from solutions.The results indicated that the adsorption efficiencydepends upon the particle size, dosage and tem-perature. In particular, the adsorption capacityincreased with a decrease in the particle size andthe dye molecules were preferably adsorbed on theouter chitosan surface. The author suggested thatthis observation can be attributed to the larger totalsurface associated with smaller particles [60]. Incontrast to the findings of Annadurai, Guibal andco-workers [80–82] observed that the adsorptionoccurred not only at the surface of the materialdue to rapid surface adsorption but also in theintraparticle network of the polymer. In particular,the large external surface area for small particlesremoves more dye in the initial stages of theadsorption process than the large particles, con-firming the previous results reported by McKayet al. [44,45]. They studied the adsorption of AG 25on chitosan and reported that the size of adsorbentparticles influenced both the adsorption kinetics andequilibrium [81] because of the resistance tointraparticle diffusion. The greater the particle size,the greater the contribution of intraparticle diffu-sion resistance to the control of the adsorptionkinetics for materials of low porosity. In otherworks [80,82], they indicated that the time requiredto reach equilibrium increased on increasing the size

of the adsorbent particles. This means that intra-particle diffusion greatly influences the accessibilityof dye molecules to internal sites. With rawchitosan, the differences were more marked thanwith protonated material [80]. Due to resistance tointraparticle mass transfer in raw chitosan, it isusually necessary to use very small particles toimprove adsorption kinetics. When the dyes havestrong interactions with chitosan, this allows largeradsorbent particle sizes to be used to get the sameadsorption rate. They concluded that this wasespecially interesting for large-scale applicationssince it was easier to manage large adsorbentparticles rather than fine powders [82]. Juang et al.[93] also observed that the adsorption capacitystrongly depended on the particle size of chitosan.At a chitosan particle size of 250–420 mm, the valueswere 380, 179 and 87mg/g for RR 222, RY 145 andRB 222, respectively. These results were signifi-cantly greater than those obtained using adsorbentssuch as CAC, natural clay, bagasse pith and maizecob, in which the capacity for reactive dyes wasoften less than 30mg/g. They concluded than thesmaller the chitosan particles, the greater thecapacity for dye. Li and co-worker [94] reportedsimilar conclusions for the adsorption of basic dyeson the adsorption of RR 189 on crosslinked beads.For example, the adsorption capacity of particleswith diameters 2.3–2.5, 2.5–2.7 and 3.5–3.8 were1936, 1686 and 1642mg/g, respectively, at pH 3 and30 1C. They also concluded that the dye uptakeincreased with a decrease in the particle sizesince the effective surface area was higher for thesame mass of smaller particles. Chiou and Chuang[66], using crosslinked chitosan for the removal ofdye from solutions, indicated that the increase inadsorption capacity with decreasing particle sizesuggests that the dye preferentially adsorbed on theouter surface and did not fully penetrate the particledue to steric hindrance of large dye molecules.Recently, Trung et al. [108] reported that no effectof the difference in particle size of decrystallizedchitosan on the decolorization capacity was ob-served. The size of particles has been shown to be akey parameter in the control of adsorption perfor-mances of several dyes on chitosan, in particularthis may be the main parameter to control dyeadsorption equilibrium. However, the relationshipof adsorption capacity to particle size also princi-pally depends on two criteria: (i) the chemicalstructure of the dye molecule (its ionic charge) andits chemistry (its ability to form hydrolyzed species)


and (ii) the intrinsic characteristic of the adsorbent(its crystallinity and porosity, the rigidity of thepolymeric chains, the degree of crosslinking), asshown by Guibal and co-workers [80–82].

Adsorption performance (in particular intrapar-ticle diffusion) is also controlled by polymerporosity (i.e. porous volume, porous distributionand pore size). CAC are well-known conventionalporous adsorbents and are characterized by a largespecific surface area and a great porosity that limitsthe resistance to intraparticle diffusion. The aggre-gation of dye molecules may involve a strongincrease in the size of the diffusing molecule, andthis effect may be reinforced by the influence of poresize in controlling intraparticle diffusion propertiesand accessibility to internal sites. Thus, the effi-ciency in adsorbing dyes onto a material such asCAC can be correlated to its surface characteristics.However, chitosan is known as a non-porouspolymer. It is characterized by a low surface areaand a low porosity that control the diffusion to thecenter of the particles, especially with large mole-cules. These features generally limit access tointerior adsorption sites. So, polymer porosity mayaffect the dye adsorption capacity of chitosan. Incrosslinked chitosan beads, usually prepared by achemical treatment with GLU, the materials aresubmicron to micron-sized, and need large internalpores to ensure adequate surface area for adsorp-tion. Indeed these chemical treatments involvesupplementary linkages that limit the transfer ofsolute molecules. In general, diffusion limitationwithin particles leads to the decreases in adsorption.These limiting effects can be compensated for by thephysical modification of the polymer. As alreadymentioned, an interesting characteristic of chitosanis its excellent ability to be processed into porousand nanoporous structures. Gel bead conditioningin addition to the decrease of polymer crystallinity,improves both swelling and diffusing properties, butalso allows expansion of the porous structure of thenetwork, which in turn enhances the transportof dyes. This physical modification allows boththe polymer network to be expanded (enhancingthe diffusion of large sized molecules) and thecrystallinity of the polymer to be reduced. Porousstructures can be formed by freeze-drying chitosan-acetic acid solutions in suitable molds. Exclusion ofchitosan acetate salt from the ice crystal phase andsubsequent ice removal by lyophilization generatesa porous material with a mean pore size that can becontrolled by varying the freezing rate and hence the

ice crystal size. Pore orientation can be directed bycontrolling the geometry of thermal gradientsduring freezing. The mechanical properties of theresulting material are mainly dependent on the poresizes and pore orientations. Another process con-sists in dissolving the polymer in acid solutionfollowed by a coagulation. Recently, Kim and Cho[71] proposed a sol–gel method to prepare porouschitosan beads with interesting high internal specificsurface areas, allowing better accessibility of dyes tointerior adsorption sites. Nanotechnology has beenalso proposed to prepare porous materials[110,118,119]. Compared to the traditional micron-sized materials, nano-sized adsorbents possess quitegood performance due to high specific area andporous structure, and the absence of internaldiffusion resistance.

4.1.3. Chemical structure of chitosan

The properties of chitosan also depend on itschemical nature (MW, DD), functional groups(ionic charge, variety, density, accessibility) andsolution behavior (purity, water content, salt-form,affinity for water). These parameters are alsodetermined by the conditions selected during thepreparation.

It is known that chitin samples have different DDdepending on their origin and mode of isolation[12]. Deacetylation takes place during isolation byalkaline treatment to remove proteins. To preparechitin with a fully N-acetylated polymer or auniform structure, selective N-acetylation of thefree amino groups is necessary. Chitosan is preparedby deacetylating chitin. Depending on the chitinsource and the methods of hydrolysis, commercialchitosan also varies greatly in its MW and distribu-tion, and therefore its solution behavior. The MWof chitosan is a key variable in adsorption propertiesbecause it influences the polymer’s solubility andviscosity in solution. It is an important factor forcharacterization, but poor solubility and structuralambiguities in connection with the distribution ofacetyl groups are major obstacles to quantitativelydetermining MWs [11]. It is also difficult todetermine the MW of native chitin.

Another important characteristic of chitosan isthe degree of N-acetylation (DA) or DD. The DDparameter is essential since, though the hydroxylgroups on the polymer may be involved in attractingdye molecules, the amine functions remain the mainactive groups and so can influence the polymer’sperformance. Guibal et al. [82] observed that


increasing the DD involved an increase in therelative proportion of amine groups, which wereable to be protonated, favoring dye adsorption.However, they indicate that the variation inadsorption properties was not proportional toDD, but changed with the type of dye, especiallywith chitin. Saha et al. [102], studying the adsorp-tion of an azo dye onto chitosan flakes, alsoreported that the results were found to be stronglydependent on the DD of the polymer. The higherDD chitosan provided a better adsorption. Re-cently, it has been reported that the solutionproperties of a chitosan depend not only on itsaverage DA but also on the distribution of theacetyl groups along the main chain [11]. However,Chiou and Li [68], studying the adsorptionof RR 189 on crosslinked chitosans reportedthat both the MW and the DD of the polymerwere almost without effect on the adsorptioncapacities.

An additional advantage of chitosan is the highhydrophilic character of the polymer due to thelarge number of hydroxyl groups present on itsbackbone. Depending on its MW and DD, chitosanin aqueous solution is expected to have the proper-ties of an amphiphilic polymer. With an increase inDD, the number of amino groups in the polymerincreases, and with an increase of MW, the polymerconfiguration in solution becomes a chain or a ball.In addition, adsorption is known to change theconformation of the chitosan polymer. The viscosityof chitosan also greatly influences the chitosanconditioning processes.

4.2. Activation conditions

4.2.1. Chitosan preprotonation

Because of its stable, crystalline structure, thepolyamine chitosan is insoluble in either water ororganic solvents. However, in dilute aqueous acids,the free amino groups are protonated and thepolymer becomes fully soluble below �pH 5. Sincethe pKa of the amino group of glucosamine residuesis about 6.3, chitosan is extremely positively chargedin acidic medium. So, treatment of chitosan withacid produces protonated amine groups along thechain and this facilitates electrostatic interactionbetween polymer chains and the negatively chargedanionic dyes, as previously observed by Maghamiand Roberts [50]. The pH-dependent solubility ofchitosan provides a convenient tool to improve itsperformance although solubility is a very difficult

parameter to control [11]. In fact, the solubility andits extent depends on the concentration and on thetype of acid. The polymer dissolves in hydrochloricacid and organic acids such as formic, acetic, lacticand oxalic acids. However, solubility decreases withincreasing concentrations of acid. Solubility is alsorelated to the DA, the ionic concentration, as well asthe conditions of isolation and drying of thepolymer [11]. In particular, the distribution of acetylgroups along the chain (random or blockwise) canstrongly influence the solubility of the polysacchar-ide and also the interchain interactions due toH-bonds and the hydrophobic character of theacetyl groups.

Trung et al. [108] proposed a pretreatment usingcitric acid to produce decrystallized chitosan with alow degree of crystallinity and a high anionic dye-binding capacity. The percentage crystallinity ofdecrystallized chitosan was 10%, significantly lowerthan that of raw chitosan (32%). This reduction isattributed to a probable rearrangement of polymerchains during precipitation in the presence of citrateions. They also indicated that the decrystallizedchitosan had the same degree of DD and MW as theoriginal chitosan. Decrystallized chitosan adsorbedanionic dyes almost twofold more efficiently thanraw chitosan, due to its more amorphous character,but showed decreased adsorption for cationic dyes.However, the presence of ash in decrystallizedchitosan could also play a role in increased dye-binding capacity. Gibbs et al. [81] showed that thepreliminary protonation of amine groups, obtainedby contact with a sulfuric acid solution, reducedthe variation of solution pH following adsorbentaddition. Crini et al. [72] found that a homogeneouschemical treatment such as a solubilization–reprecipitation process could give a chitosan pro-duct with a higher adsorption level for dyes thanone prepared by a heterogeneous process with thesame DD. They attributed this to an increase of thesurface area due to the conversion of the chitosanflakes into a powder.

4.2.2. Grafting reactions

Several workers have suggested that althoughchitosan as such is very useful for treating con-taminated solutions, it may be advantageous tochemically modify chitosans, e.g. by grafting reac-tions [72,73,77,95,103,106,113]. The modificationscan improve chitosan’s removal performance andselectivity for dyes, alter the physical and mechan-ical properties of the polymer, control its diffusion


properties and decrease the sensitivity of adsorptionto environmental conditions. Chemical graftingof chitosan with specific ligands has been reviewedby Jayakumar et al. [23] and by Prabaharan andMano [41].

It is known that the only class for which chitosan[106] and crosslinked chitosan [85] have low affinityis basic (cationic) dyes. To overcome this problem,Crini et al. [72,73] suggested the use of N-benzylmono- and disulfonate derivatives of chitosan inorder to enhance its cationic dye hydrophobicadsorbent properties and to improve its selectivity.The maximum adsorption capacities of theseadsorbents for BB 9 and BB 3 were 121.9 and166.5mg/g, respectively. These derivatives could beused as hydrophobic adsorbents in acidic mediawithout any crosslinking reaction. To fully developthe high potentials of chitosan, it is necessary tointroduce chemical substituents at a specific positionin a controlled manner as suggested by Lima et al.[58] and Chao et al. [106]. Lima et al. [58] proposedthe use of chitosan chemically modified withsuccinic anhydride in the BB 9 adsorption. Thischemical derivatization provides a powerful meansto promote new adsorption properties in particulartowards basic dyes in acidic medium. Chao et al.[106] suggested enzymatic grafting of carboxylgroups onto chitosan as a means to confer theability to adsorb basic dyes on beads. The presenceof new functional groups on the surface of beadsresults in increases in surface polarity and thedensity of adsorption sites, and this improves theadsorption selectivity for the target dye. Otherstudies showed that the ability of chitosan toselectively adsorb dyes could be further improvedby chemical derivatization. Shimizu et al. [103]proposed novel chitosan-based materials by react-ing chitosan with a higher fatty acid functionalizedwith a glycidyl moiety in order to introduce longaliphatic chains. They observed that these productscould be used as effective adsorption materials forboth anionic and cationic dyes. Martel et al. [95],and El-Tahlawy and co-workers [76,77] proposedthe use of cyclodextrin-grafted chitosan as newchitosan derivatives for the removal of dyes. Martelet al. [95] demonstrated that these materials arecharacterized by a rate of adsorption and a globalefficiency greater than that of the parent chitosanpolymer. Uzun and Guzel [113,114] reported thatcarboxymethylated chitosan is a rather betteradsorbent than raw chitosan for acidic dyestuffs,and its production is not costly.

4.2.3. Influence of crosslinking

Raw chitosan powders also tend to present somedisadvantages such as unsatisfactory mechanicalproperties and poor heat resistance. Anotherimportant limitation of the raw material is that itis soluble in acidic media and therefore cannot beused as an insoluble adsorbent under these condi-tions, except after physical and chemical modifica-tion. One method to overcome these problems is totransform the raw polymer into a form whosephysical characteristics are more attractive. So,crosslinked beads have been developed and pro-posed. After crosslinking, these materials maintaintheir properties and original characteristics [62],particularly their high adsorption capacity,although this chemical modification results in adecrease in the density of free amine groups at thesurface of the adsorbent in turn lowering polymerreactivity towards metal ions [80].

An important work on crosslinked chitosan wasdone by Chiou and co-workers [66–70]. Chitosanbeads were crosslinked with GLU, EPI or EGDE.The results showed that the chitosan-EPI beadspresented a higher adsorption capacity than GLUand EGDE resins [68,69]. They reported that thesematerials can be used for the removal of reactive,direct and acid dyes. It was found that 1 g chitosanadsorbed 2498, 2422, 2383 and 1954mg of RB 2,RR 2, DR 81 and AO 12, respectively [67]. It isimportant to specify that the adsorption capacitiesof CAC for reactive dyes generally vary from 278 to714mg/g [6]. Another advantage of EPI is that itdoes not eliminate the cationic amine function ofthe polymer, which is the major adsorption site toattract the anionic dyes during adsorption [69]. Thecrosslinking of chitosan with GLU (formation ofimine functions) or with EDGE decreases theavailability of amine functions for the complexationof dyes and with a high crosslinking ratio the uptakecapacity drastically decreases. They also indicatedthat the crosslinking ratio slightly affected theequilibrium adsorption capacity for the three crosslinkers under the range they studied [68]. Theamount of dye adsorbed was found to be higher inacidic than in basic solution. This was explained byconsidering the rate of diffusion from the swollenbeads in acidic and basic media. In basic medium, alimited swelling of the beads inhibited the diffusionof dyes at a faster rate as it occurred in acidicmedium. Among the conditions of the crosslinkingreaction that have a great impact on dye adsorptionare the chemical nature of the crosslinker, as


mentioned above, but also the extent of thereaction. In general, the adsorption capacity de-pends on the extent of crosslinking and decreaseswith an increase in crosslinking density. Whenchitosan beads were crosslinked with GLU underheterogeneous conditions, it was found that thesaturation adsorption capacity of RR 2 on cross-linked chitosan decreased exponentially from 200 to50mg/g as the extent of crosslinking increased from0 to 1.6mol GLU/mol of amine. This is because ofthe restricted diffusion of molecules through thepolymer network and reduced polymer chainflexibility. Also the loss of amino-binding sites byreaction with aldehyde is another major factor inthis decrease. However, Chiou and co-workersindicated that the crosslinking step was necessaryto improve mechanical resistance, to enhance theresistance of material against acid, alkali andchemicals, and also to increase the adsorptionabilities of chitosan. The removal performance ofcrosslinked chitosan and CAC for anionic dyes werecompared: the adsorption values were 3–15 timeshigher at the same pH. Chiou and co-workers[66–70] concluded that chitosan chelation was theprocedure of choice for dye removal from aqueoussolution. However, Kim and Cho [71], studying theadsorption of RB 5 on crosslinked chitosan beads,arrived at contrasting conclusions. They demon-strated that the adsorption capacity of non-cross-linked beads was greater than that of crosslinkedbeads in the same experimental conditions.

The materials, mainly crosslinked using GLU,have been also proposed as effective dye removersby several other workers [62,77,84,85,88,94,105]. Allthese studies showed that the reaction of chitosanwith GLU leads to the formation of imine groups,in turn leading to a decrease in the number of aminegroups, resulting in a lowered adsorption capacity,especially for dyes sorbed through ion-exchangemechanisms. However, this limiting effect of achemical reaction with GLU significantly dependson both the procedure used and the extent ofcrosslinking, as reported by Hebeish et al. [84,85]. Inheterogeneous conditions, chitosan (solid state) wassimply mixed with GLU solution, while in homo-geneous conditions chitosan was mixed with GLUsolution after being dissolved in acetic acid solution.An optimum aldehyde/amine ratio was found fordye adsorption, which depended on the crosslinkingoperation mode (water-soluble or solid-state solu-tion). The initial increase in dye adsorption wasattributed to the low levels of crosslinking in the

precipitates preventing the formation of closelypacked chain arrangements without any greatreduction in the swelling capacity. This increase inadsorption was interpreted in terms of the increasesin hydrophilicity and accessibility of complexinggroups as a result of partial destruction of thecrystalline structure of the polymer by crosslinkingunder homogeneous conditions. At higher levels ofcrosslinking, the precipitates had lower swellingcapacities, and hence lower accessibility because ofthe more extensive three-dimensional network andalso because of its more hydrophobic character withincreased GLU content. Juang et al. [88], studyingthe adsorption of RR 222 on crosslinked chitosanbeads, also observed that the adsorption capacitydepends on the extent of crosslinking and decreasesas crosslinking density increases. This result wasmainly interpreted by the fact that the crosslinkingreaction with GLU decreases the availability ofamine functions for the complexation of dyes. Theresults showed that the chitosan-GLU beads pre-sented a higher adsorption capacity than glyoxalbeads. Gaffar et al. [77] and Shimizu et al. [105]reported that the extent of crosslinking showed asignificant influence on adsorption properties. Theseauthors noted that the increase in the extent ofcrosslinking is accompanied by a decrease in dyeuptake, confirming the results of Hebeish et al.[84,85]. The adsorption capacity increased greatly atlow degrees of substitution but decreased withincreasing substitution. This phenomenon is inter-preted in terms of increased hydrophilicity causedby the destruction of the crystalline structure at lowcrosslinking densities, while this can be associatedwith an accompanying decrease in active sites,accessibility, and swellability of the adsorbent byincreasing the level of crosslinking. On the contrary,Chiou and Li [94], studying the adsorption of RR189 on EPI-crosslinked chitosan beads, reportedthat the crosslinking ratio did not affect theadsorption capacity.

Another study showed that the physical andmechanical properties of chitosan could be furtherimproved by crosslinking. Chitosan forms gelsbelow pH 5.5 and acid effluents could severely limitits use as an adsorbent in removing dyes from acideffluent. To solve this problem, Cestari et al. [62]proposed the use of homogeneously crosslinkedbeads. They reported that the beads were not onlyinsoluble in acid solution but also presented higherspecific surface areas (0.1 and 0.24m2/g before andafter the crosslinking reaction, respectively) and


stronger mechanical resistance than the raw chit-osan powder. The chemical, physical and mechan-ical behavior of the beads and also adsorptionproperties were enhanced by crosslinking withfunctional groups. The materials had a strongadsorption capacity for RY, RB and RR belowpH 5.5. The authors also noted that crosslinking canchange the crystalline nature of chitosan, assuggested by the XRD diffractograms. After thecrosslinking reaction, there was a small increase inthe crytallinity of the chitosan beads and alsoincreased accessability to the small pores of thematerial.

4.2.4. Chitosan-based composite beads

Practical industrial applications of raw chitosanin fixed-bed systems or packed in adsorptioncolumns are also limited. The characteristics of thepolymers can introduce hydrodynamic limitationsand column fouling, which limits their use for large-scale columns. For example, the flaked or powderedform swells (the crosslinked beads have lowerswelling percentage [120]) and crumbles easily, anddoes not function ideally in packed-column config-urations common to pump-and-treat adsorptionprocesses. Various chitosan-based composites havebeen designed to overcome these problems. Changand Juang [87] proposed the addition of activatedclay to chitosan to prepare composite beads in orderto improve its mechanical properties. Cestari et al.[61] also proposed the use of silica/chitosan hybridfor the removal of anionic dyes from aqueoussolutions: these materials are of interest becausethey combine the structure, strength and chemicalproperties of the silica with the specific character-istics of chitosan. Chang and Chen [64] proposedthe use of chitosan-conjugated Fe3O4 nanoparticlesfor the removal acid dyes from aqueous solutions.The adsorption capacities were 1883 and 1471mg ofdye/g of chitosan for AO 12 and AG 25. Panevaet al. [96] also proposed a novel effective route forincorporating magnetic material into chitosan beadsby capillary extrusion. They concluded that thematerial might be used for wastewater treatment inthe textile industry.

4.3. Influence of process variables

The amount of dye that can be removed from asolution by chitosan also depends on processvariables used in batch systems such as chitosan

dosage, initial dye concentration, contact time,agitation rate and dryness.

4.3.1. Effect of chitosan dosage

Of all the above factors, chitosan dosage isparticularly important because it determines theextent of decolorization and may also be used topredict the cost of chitosan per unit of solution to betreated. As expected, the adsorption density in-creases significantly as adsorbent dosage decreases.This is due to the higher amount of the dye per unitweight of adsorbent. Wen et al. [116] showed thatthe increasing chitosan dose had a dramatic positiveimpact on color removal and there was anapproximately linear relationship between chitosandose and color removal of the dye. Crini et al.[72,73] also observed that the increase in adsorptionwith adsorbent dosage can be attributed to in-creased adsorbent surface and availability of moreadsorption sites. However, if the adsorption capa-city was expressed in mg adsorbed per gram ofmaterial, the capacity decreased with the increasingamount of sorbent. This may be attributed tooverlapping or aggregation of adsorption sitesresulting in a decrease in total adsorbent surfacearea available to the dye and an increase in diffusionpath length. It was also indicated that the timerequired to reach equilibrium decreased at higherdoses of adsorbent.

4.3.2. Effect of initial dye concentration

Park et al. [56] and Knorr [121] previously foundsignificant correlations between dye concentrationand the dye-binding capacity of chitin or chitosan.The amount of the dye adsorbed onto chitosanincreased with an increase in the initial concentra-tion of dye solution if the amount of adsorbent waskept unchanged. This is due to the increase in thedriving force of the concentration gradient with thehigher initial dye concentration. In most cases, atlow initial concentration the adsorption of dyes bychitosan is very intense and reaches equilibriumvery quickly. This indicates the possibility of theformation of monolayer coverage of the moleculesat the outer interface of the chitosan. At a fixedadsorbent dose, the amount adsorbed increasedwith increasing concentration of solution, but thepercentage of adsorption decreased. In other words,the residual concentration of dye molecules will behigher for higher initial dye concentrations. In thecase of lower concentrations, the ratio of initialnumber of dye moles to the available adsorption


sites is low and subsequently the fractional adsorp-tion becomes independent of initial concentration[67,68,75,83]. At higher concentrations, however,the number of available adsorption sites becomeslower and subsequently the removal of dyes dependson the initial concentration. At the high concentra-tions, it is not likely that dyes are only adsorbed in amonolayer at the outer interface of chitosan. As amatter of fact, the diffusion of exchanging mole-cules within chitosan particles may govern theadsorption rate at higher initial concentrations.Recently, Gaffar et al. [77] reported that theadsorption percentage decreases on increasingthe dye concentration. This could be ascribed tothe accompanying increase in dye aggregationand/or depletion of accessible active sites on thematerial.

4.3.3. Effect of contact time

Contact time is another important variable inadsorption processes. Generally speaking, the ad-sorption capacity and the removal efficiency of dyesby chitosan become higher on prolonging thecontact time. However, in practice, it is necessaryto optimize the contact time, considering theefficiency of desorption and regeneration of theadsorbent.

During the process, the adsorbent surface isprogressively blocked by the adsorbate molecules,becoming covered after some time. When thishappens, the adsorbent cannot adsorb any moredye molecules. As each particle purifies a certainvolume of liquid, increasing the dosages rapidlypromotes an equilibrium between adsorbate andadsorbent because the number of particles to treatthe same volume of liquid is increased. In general,the adsorption capacity increases with time and, atsome point in time, reaches a constant value whereno more dye is removed from the solution. At thispoint, the amount of dye being adsorbed onto thematerial is in a state of dynamic equilibrium withthe amount of dye desorbed from the adsorbent.The time required to attain this state of equilibriumwas termed the equilibrium time (te) and the amountof dye adsorbed at te reflected the maximum dyeadsorption capacity of the adsorbent under theseconditions.

Chatterjee et al. [83] reported an equilibrium timeof 20 h for eosin adsorption onto chitosan. Theyobserved that the process was initially very fast andthen slowly reached equilibrium. Dutta et al. [75]noted that the maximum accumulation occurred

within 4–5 h for reactive and direct dyes onchitosan. Guibal et al. [82], studying the adsorptionof 12 anionic dyes on chitosan also observed thatequilibrium was reached within the first 12 h ofcontact and adsorption kinetics were relatively fast.Gibbs et al. [81] noted that with increasing AG 25concentration relative to a fixed adsorbent dosage,the time required to reach equilibrium stronglyincreased. Although 1–2 h was sufficient to achievecomplete recovery of the dye at initial concentra-tions of below 100mg/l. For the highest concentra-tion (200mg/l) with raw chitosan 8 h was necessaryto reach equilibrium and the complete eliminationof the dye. Fahmy et al. [79] reported an equilibriumtime of 45min for anionic dye adsorption oncrosslinked chitosan. Cestari et al. [61,62] foundreaction times of 60–200min. At the other extreme,Wu et al. [91] observed that it took several days forequilibrium to be attained. Crini et al. [72,73],studying adsorption of basic dyes (BB 9 and BB 3)with different chitosan concentrations and contacttimes observed that 40min of contact time wasenough to reach adsorption equilibrium in all theexperiments, while Chang and Juang [86] report that3 days were required for BB 9 adsorption ontochitosan. The difference might have been due todifferences in the properties of the material used orto differences in the aqueous solution treated. Criniet al. [73] used grafted chitosan with a solutioncontaining sodium chloride at pH 3 while Changand Juang [86] used chitosan that had not beenpretreated with a neutral solution. However, mostauthors seem to agree on a figure in the range3–5 days for most dye molecules (Table 8) and withthe fact that the adsorption of dyes is fast at theinitial stages of the treatment time, and thereafter,becomes slower near the equilibrium. It is obviousthat a large number of vacant surface sites areavailable for adsorption during the initial stage, andafter a lapse of time, it is difficult to occupy theremaining vacant surface sites due to repulsiveforces between dye molecules adsorbed on the solidand and those in the solution phase.

The contact time and adsorption rate aredependent on the initial dye concentration. Gibbset al. [81] observed that increasing the initial dyeconcentration increased the time required to achievecomplete recovery of the dye. As also expected,decreasing the adsorbent particle size leads to adecrease in the time required to reach the equili-brium and a strong increase in the initial adsorptionvelocity. The modification of chitosan by grafting


reactions [72,73] allows the effect of particle size tobe reduced. It is suspected that these variationsare caused by the specific surface area of the sizefractions, the wide size dispersion around themedian value and the effect of diffusion mechan-isms. The equilibrium time increases with thecrosslinking ratio [62,69,71,84,85,95]: GLU cross-linking involves the formation of new interchainlinkages and a loss of chain flexibility, and thussome restriction at the entrance to the polymernetwork. Therefore, the extent of crosslinking isexpected to play a great part in adsorption/diffusioncontrol, especially intraparticle diffusion. However,the differences are not as marked [62,79]: it isdifficult to find a homogeneous trend in theintraparticle diffusion coefficients with increasingcrosslinking ratio. Cestari et al. [61] also indicatedthat the adsorption behavior of anionic dyes wasdirectly related to the dimensions of the dye organicchains, the amount and position of the sulfonategroups, and the adsorption temperature. Maghamiand Roberts [50] previously reported that equili-brium was reached more rapidly with the smallestdye, being attained in less than 2 h with AO 7 whileapproximately 9 h were required with AR 27.However, they indicated that the ionic charge onthe dye appeared to have a negligible effect on thetime to equilibrium.

4.3.4. Effect of stirring rate

Stirring is an important parameter in adsorptionphenomena, influencing the distribution of thesolute in the bulk solution and the formation ofthe external boundary film. Generally, the rate ofdye removal is influenced by the degree of agitationand the uptake increased with stirring rate. Thedegree of agitation reduced the boundary-layerresistance and increased the mobility of the system.Increase in agitation by increasing stirrer speedlowers the external mass transfer effect. Uzun andGuzel [114] reported that the adsorption of the dyesO II and CV by chitosan must be studied at highshaking rate. In another recent work [112], theyindicated that there is a small effect of shaking rateon the adsorption of RY 2 and Rb 5 by chitosan.Wu et al. [91] also noted that agitation had littleeffect on adsorption.

4.3.5. Effect of dryness

To find the effect of dryness of beads on theadsorption rate, Chiou and Li [68] used dried beadsto evaluate the adsorption behavior. They reported

that the adsorption rate for wet beads is much fasterthan that of dry beads and the time lag to reachsimilar adsorption capacity is lower because it takestime for the dry beads to swell before adsorptioncan take place. Chang and Juang [86] also reportedthat the adsorption capacity of activated chitosanwith the wet composite beads was generally higherthan that with the dried beads, possibly because oftheir stronger affinity of water for the bead matrix.

4.4. Chemistry of the dye

In liquid-phase adsorption, the adsorption capa-cities of an adsorbent are commonly attributed tomany factors such as its origin, physical, chemicaland mechanical properties, solution conditions, etc.Dye–dye interactions also play an important role aswell as those between dye and aqueous solution. Asdescribed throughout this review, chitosan is beingstudied extensively as an adsorbent for dye removal.However, there is still much to be accomplished inunderstanding its mechanisms. In particular, dyemolecules have many different and complicatedstructures. This is one of the most important factorsinfluencing adsorption and mechanisms. In addi-tion, in most studies, experimental procedures donot take into account the change in pH due tochitosan addition and its effect on dye chemistry,leading to inaccurate interpretation of adsorptionproperties. Moreover, changing the experimentalconditions (i.e. the pH, the dye concentration andthe matrix of the solution) can considerably affectthe distribution of dye molecules and consequentlytheir ability to interact with chitosan. However, tothe best of our knowledge, the comparative effectsof important variables such as the kind of dye, itspKa, the differences in chemical dye structures onadsorption behavior on chitosan beads have beenlittle studied, excepted in recent years.

Guibal et al. [82], studying the adsorption of acid,direct, mordant and reactive dyes on chitosan,reported that both adsorption capacities andkinetics depended on the type of dye involved.However, any attempt to correlate adsorptionperformance to the structure of the dye failed. Theyfound the following order for anionic dyes: MO104RB 54AV 54AG 254AB 14AY 254DB714MB 29 for raw chitosan and AV 54AG254AY 254MO 104MB 294AB 14DB 71 forcrosslinked chitosan. Following to Maghami andRoberts [50], who postulated a 1:1 stoichiometry forthe interaction of sulfonic acid groups on the dyes


with the protonated amine groups of the chitosanfor mono-, di- and tri-sulfonated dyes, Guibal et al.[82] attempted to correlate the adsorption capacitiesof the dyes to their sulfonate content. They observedthat no meaningful correlation was apparent. Somedyes containing numerous sulfonic acid groupsexhibited adsorption capacities lower than otherdyes containing only one group per molecule. Theprediction of adsorption performance from onlysulfonic acid content was not possible and otherparameters (pKas, presence of other functionalgroups, hydrophobicity) may control adsorptioncapacities. They concluded that the differencesbetween the dyes may be due to different pKasand/or to the contribution of other interactions inthe adsorption mechanism such as hydrophilic andhydrophobic interactions due to the differentchemical dye structures. There was no correlationbetween the size of the dye and the better efficiencyof chitosan for its adsorption compared to CAC. Itis also important to point out that some dyes suchas RB 5 might be subject to hydrolysis, and thiswould explain the fact some sulfonic acid groups arenot able to react with chitosan [80].

Different conclusions have been reported byWong et al. [99] who attributed the differences inthe degree of adsorption mainly to the chemicalstructure of each dye. A detailed study of the orderof affinity of five acid dyestuffs for ion-exchangeonto chitosan was reported by the authors. Theynoted that there was a great variation in the affinityof chitosan for these dyes in a single solute system.The order of extent of decolorization was AO124AO 104AR 734AR 184AG 25. The differ-ences in adsorption capacities may be due to theeffect of molecular size and the number of sulfonategroups of each dye. They concluded that mono-valent and/or smaller dye molecules have greateradsorption capacities due to an increase in the dye/chitosan ratio in the batch system. The smaller dyemolecules are able to undertake a deeper penetra-tion of dye into the internal pore structure of thechitosan particles. A similar interpretation waspreviously given by McKay et al. [44] for theadsorption of basic dyes on chitin. Smith et al. [52]reported that the molecular size of the dye was amajor factor in adsorption characteristics. Theynoted that small, low MW dyes adsorbed best onchitosan. Crini et al. [72,73] also found significantvariations in affinity of the different cationic dyes(BB3 and BB 9) to grafted chitosan, with minimumadsorption for BB 3 (166.5mg/g) and maximum for

BB 9 (121.9mg/g). It was found that the molecularsize of the dye was a major factor in adsorptioncharacteristics. Juang et al. [93] observed that theadsorption capacity, defined as the amount atthe isotherm plateau, depended on the nature ofthe dye molecules. Under comparable experimentalconditions, the capacity decreased in the order RR2224RY 1454RB 222, confirming the role of dyestructure. They also added that the adsorption ratealso depended on the dye chemistry and followedthe same order. Cestari et al. [61] also indicated thatthe adsorption behavior of dyes was influenced bythe chemical structure of the dye molecules, but theyreported that the link with the dye structure was notclearly identified.

4.5. Effect of the solution conditions

4.5.1. Effect of pH

The pH of the dye solution plays an importantrole in the whole adsorption process and particu-larly on the adsorption capacity, influencing notonly the surface charge of the adsorbent, the degreeof ionization of the material present in the solutionand the dissociation of functional groups onthe active sites of the adsorbent, but also thesolution dye chemistry. It is important to indicatethat while the adsorption on CAC was largelyindependent of the pH, the adsorption of (anionic)dyes on chitosan was controlled by the acidity of thesolution.

pH affects the surface charge of the adsorbent.Chitosan is a weak base and is insoluble in waterand organic solvents, however, it is soluble in diluteaqueous acidic solution (pHo6.5) which can con-vert the glucosamine units into a soluble formR-NH3

+. It gets precipitated in alkaline solution orwith polyanions and forms gels at lower pH. Its pKa

depends on the DD, the ionic strength and thecharge neutralization of amine groups. In practice itlies within 6.5–6.7 for fully neutralized aminefunctions [32]. So, chitosan is polycationic in acidicmedium: the free amino groups are protonated andthe polymer becomes fully soluble and this facil-itates electrostatic interaction between chitosan andthe negatively charged anionic dyes. This cationicproperty will influence the adsorption procedure,especially in the case of anionic dyes, depending onthe charge and functions of the dye under thecorresponding experimental conditions [82]. In1994, Muzzarelli and co-workers [53,54], studyingthe interactions of dyes with chitosan in solution by


circular dichroism analysis, observed that the pH ofthe solution changed both the extent and the modeof the binding. They explained this change in themechanism by changes in the conformation ofthe polymer chains.

In the literature, the ability of the anionic dyes toadsorb onto chitosan beads is often attributed to thesurface charge which depends on the pH of theoperating batch system, as mentioned. Dye adsorp-tion occurred through electrostactic attraction onprotonated amine groups and numerous workersconcluded that the influence of the pH confirmedthe essential role of electrostatic interactions be-tween the chitosan and the target dye. For example,Chatterjee et al. [83] indicated that chitosan had apositively charged surface below pH 6.5 (point ofzero potential), and reducing the pH increased thepositivity of the surface, thus making the adsorptionprocess pH sensitive. Decreasing the pH makesmore protons available to protonate the aminegroup of chitosan with the formation of a largenumber of cationic amines. This results in increasingdye adsorption by chitosan due to increasedelectrostatic interactions. Differences in pH of thesolution have also been reported by Gibbs et al.[80,81] to influence the dye adsorption capacity ofchitosan and its mechanism. They noted that, at lowpH, chitosan’s free amino groups are protonated,causing them to attract anionic dyes, demonstratingthat pH is one of the most important parameterscontrolling the adsorption process. Crini et al.[72,73] also found that the adsorption capacityof cationic dyes on chitosan-grafted materialswas strongly affected by the pH of solution andwas generally significantly decreased by increasingthe pH.

pH is also known to affect the structural stabilityof dye molecules (in particular the dissociation oftheir ionizable sites), and therefore their colorintensity. For example, BG 4, a cationic dye(pKa ¼ 10.3) gets protonated in acidic mediumand deprotonated in basic medium. Consequently,the dye molecule has high positive charge density ata low pH. This indicates that the deprotonation(or protonation) of a dye must be take into account.If the dyes to be removed are either weakly acidicor weakly basic, then the pH of the medium affectstheir structure and adsorption. Initial pH alsoinfluences the solution chemistry of the dyes:hydrolysis, complexation by organic and/or inor-ganic ligands, redox reactions, and precipitation arestrongly influenced by pH, and on the other side

strongly influence speciation and the adsorptionavailability of the dyes.

Two interesting experimental facts must bepointed out. Firstly, the free amine groups inchitosan are much more reactive and effective forchelating pollutants than the acetyl groups in chitin,and there is no doubt that amine sites are the mainreactive groups for (anionic) dye adsorption, thoughhydroxyl groups (especially in the C-3 position) maycontribute to adsorption. Almost all functionalproperties of chitosan depend on the chain length,charge density and charge distribution and much ofits potential as biosorbent from its cationic natureand solution behavior. However, at neutral pH,about 50% of total amine groups remain proto-nated and theoretically available for the adsorptionof dyes. The existence of free amine groups maycause direct complexation of dyes co-existing withanionic species, depending on the charge of the dye.As the pH decreases, the protonation of aminegroups increases together with the efficiency.The optimum pH is frequently reported in theliterature to be around pH 3–6 (Table 10). Belowthis range, usually a large excess of competitoranions limits adsorption efficiency. This competitoreffect is the subject of many studies aiming todevelop materials that are less sensitive to thepresence of competitor anions and to the pH of thesolution, as described in the next two paragraphs.Secondly, it is not really the total number of freeamine groups that must be taken into account butthe number of accessible free amine groups. Thereare several explanations for this. The availability ofamine groups is controlled by two importantparameters: the crystallinity of polymer and thediffusion properties. It is known that some ofthe amine sites on chitosan are involved in boththe crystalline area and in inter or intramolecularhydrogen bonds. Moreover the residual crystallinityof the polymer may control the accessibility toadsorption sites. The DD also controls the fractionof free amine groups available for interactions withdyes. Indeed, the total number of free amine groupsis not necessarily accessible for dye uptake. Actu-ally, rather than the fraction or number of freeamine groups available for dye uptake, it would bebetter to consider the number of accessible freeamine groups. Guibal et al. [79] recently showedthat not all the amino groups are really available orat least accessible. They also concluded that thehydrogen bonds linked between monomer units ofthe same chain (intramolecular bonds) and/or


between monomer units of different chains (inter-molecular bonds) decrease their reactivity. Theweakly porous structure of the polymer and itsresidual crystallinity are critical parameters for thehydratation and the accessibility to adsorption sites.

4.5.2. Effect of pH variation

To date it remains difficult to establish cleartrends in the adsorption properties of chitosanmaterials for the recovery of dyes. The variability inpublished results may deter potential users. Thereare many reasons for this difficulty in comparingadsorption performances such as differences inchitosan and above all experimental conditions(equilibrium time, adsorbent dose), and also under-estimation of the influence of pH, especially pHvariation effects during the process [81]. Indeed,frequently, batch systems used in the literature donot take into account the change in pH.

Few studies have been published on the inter-pretation of the pH variation during adsorptionprocess. Sakkayawong et al. [109] reported that thesystem pH changed during RR 141 adsorption bychitosan. The explanation for this was that underacidic conditions hydrogen atoms in the solutioncould protonate the amine groups of the polymerand thus causes the increased pH. In addition, theadsorption efficiency was systematically greater forsolutions whose pH was controlled during theadsorption than for solutions for which pH variedalong the uptake process. At increasing adsorbentdosage, Gibbs et al. [81] observed that the additionof chitosan to the solution strongly increased its pHand noted that the pH variation increased exponen-tially. The authors interpreted this phenomenon asrelated to the increase of the number of aminegroups available for protonation, consequentlycausing a marked increase in the pH. Increasingadsorbent dose also reduced polymer saturation.

To diminish the influence of pH change on theinterpretation and modeling of adsorption perfor-mance, chitosan can be conditioned by contact witha solution of sulfuric acid [81] or other chemicals[83], as shown below.

4.5.3. pH sensitivity

It is well known that chitosan hydrobeads losetheir integrity as a result of partial dissolution inacidic solvent, since protonation of amine groupscauses the polymer to dissolve, making themunsuitable for reuse. Crosslinking of the chitosanbeads has been proposed but can result in poor

adsorption. Thus, it is of interest to increase theintegrity of the beads as well as adsorption proper-ties at acidic pH, and also to diminish the influenceof the pH change during adsorption process.

Recently, several authors [72,81,83] chose tomodify raw chitosan without crosslinking in orderto diminish the pH variation. Crini et al. [72]proposed the use of new chitosan derivatives asadsorbents in acidic medium. The materials areprepared by grafting reactions followed by asolubilization/reprecipitation step, converting thechitosan flakes into a powder having an advanta-geous specific surface area. The originality of thematerials is their inversed pH domain of solubilitycompared to the parent polymer. This characteristicallowed adsorption capacity to be studied in strongaqueous acidic media (pH 2/3), without furthercrosslinking or pH sensitivity. Gibbs et al. [81]showed that the preliminary protonation of aminegroups, obtained by contact with a sulfuric acidsolution, reduces the variation of solution pHfollowing adsorbent addition. In this case, pHvariation during the adsorption process was sig-nificantly lower. A direct correlation can beobserved between the theoretical neutralizationcurve of chitosan, calculated from the pKa, andthe adsorption performance at different pHs. Theyalso observed that when the initial dye concentra-tion increased, pH variation increased, indicatingthat the protonation of the dye plays an importantrole in this variation. However, as previouslymentioned, this pretreatment (preprotonation)strongly reduced adsorption performance at bothequilibrium and kinetic levels. In particular, thetime necessary to reach equilibrium increased up tothreefold depending on the experimental conditions.Chatterjee et al. [83] proposed the conditioning ofthe beads with ammonium sulfate to reduce the pHsensitivity of the process with interesting adsorptionproperties in the case of eosin Y. This modificationwould probably form complexes between cationicamino groups and ammonium sulfate, which under-goes interactions with dye anions.

4.5.4. Effect of ionic strength

For almost all the treatment strategies, animportant factor which has not yet been adequatelycharacterized is the effect of typical wastewatercontaminants on decolorization efficiencies [21]. Intypical dyeing systems it is well known that certainadditives such as salts and surfactants can eitheraccelerate or retard dye adsorption processes. For


example, sodium chloride is often used as astimulator in dyeing processes. Salts may have twofunctions: (i) they may screen the electrostaticinteraction of opposite charges in adsorbents andthe dye molecules, and an increase in salt concen-tration could decrease the amount of dye adsorbed;(ii) they may enhance the degree of dissociationof the dye molecules and facilitate the amount ofpollutant adsorbed. The former function seems tobe dominant in the literature, confirming thatelectrostatic interaction is responsible for adsorp-tion of acid dyes by chitosan. The situation becomesmuch more significant when both the adsorbate andadsorbent are charged bodies. So, the ionic strengthmay be another important factor in the adsorptionof certain dyes onto chitosan [68,72,95,100,105].

Crini et al. [72] discussed the strong effect ofsodium chloride on the adsorption process. Theyconcluded that the capacities depended on the ionicstrength of the solution. Added salts affectedadsorption via two mechanisms, either by screeningthe coulombic potential between the adsorbing mole-cule and charged adsorbents, or by adsorbingpreferentially on the active sites of the adsorbent.They also noted that the increase in NaCl concen-tration reduced the dose of chitosan necessary toadsorb the pollutant. In another work, Martel et al.[95] reported that the addition of salts in batchsystems diminished the solubility of the dyestuff inthe solution and thus favored its precipitation ontopolymer, suggesting the presence of an aggregationmechanism increasing the adsorption capacity in thepresence of salts. Miyata et al. [100] also confirmedthe fact that the addition of sodium chloride greatlyaffected the adsorption of acid and direct dyes bychitosan in agreement with the prediction of thechemisorption mechanism. Shimizu et al. [105]noted that the addition of sodium chloride greatlyaffected the adsorption of AR 1 by crosslinkedchitosan while no effect was observed on theadsorption of AR 138, indicating the role of thedye structure. They concluded that the adsorptionof AR 138 also involved hydrophobic interactionsbetween the alkyl groups in the dye and thehydrophobic functions in the adsorbent. However,these interactions are strongly influenced by the pHof the solution. Chiou and Li [68] also reported thataddition of inorganic salts to the adsorption systemis an effective way of influencing adsorption.However, they observed that the presence of NaClin the batch solution lowers the adsorption capacity,slows down the initial adsorption rate and increases

the equilibrium time. They suggested that theaddition of NaCl reduced the electrostatic interac-tion between dye and crosslinked chitosan.

4.5.5. Effect of competitive molecules and ions

Competitive adsorption occurs where the adsorp-tion of a mixture of adsorbates is carried out on onesurface. Some of the components in the effluent mayinduce the adsorption of others or may coadsorbalong with another components. As mentionedabove, the variability of wastewater must be takeninto account in the design of any decolorizationsystem. However, only very limited information isavailable on the competitive adsorption of dyemolecules with chitosan-based materials.

In an effort to further understand the adsorptionprocess of dye molecules onto chitosan, Chiou et al.[68,69] evaluated the effect of mixed-dye solutions.They reported that the presence of other moleculesmight affect the adsorption of a particular molecule.They concluded that several dyes in a solutionwould compete against each other for availablesites. Those having the greatest ionic potentialwould be removed first, and if the sites were stillundersaturated, then those having lower ionicpotential would be removed in sequence. The moreelectronegative molecules are attracted to the sur-face more strongly. Although the presence of morethan one dye in a solution creates competition foradsorption sites, the total adsorption capacity hasbeen found to increase. The presence of counter-ions, interferents, and/or other substances in solu-tion affects the adsorption of dyes by chitosan, asrecently reported by Wen et al. [116]. They notedthat the presence of appreciable quantities of Na+

did not have any effect on RR 195 removal bychitosan. Addition of other cation species (Ca+,Mg2+, Fe2+) decreased color removal. This wasattributed to chelation between cations and chitosanchains, which decreased the electrostatic interactionbetween RR 195 and chitosan. Compared withFe2+ alone, the combination of Fe2+/HCO3

�

increased color removal. However, the mechanismsof combination need to be explored.

4.6. Stability

In view of industrial developments of the variouskinds of chitosan derivatives, the stability of thematerials is of utmost importance. Unfortunately,in this area, the literature produces very littleinformation except in recent publications [18,62,72].


Being a biopolymer, chitosan is biodegradable.This property may be a serious drawback for long-term applications in adsorption processes, in parti-cular in dynamic systems using fixed-bed columns.The low thermal stability of chitosan and itsdegradation resulting from acidic hydrolysis areanother important criterion to take into account.Generally, it is possible to reinforce the chemicalstability of the chitosan by crosslinking treatmentsor grafting reactions [62,68,73,84,88]. However,despite the number of studies dealing with chitosangel beads, their mechanical stability is significantlyless documented [18]. Guibal [18] pointed out therole of crosslinking reactions on chitosan beadstability. He reported that, in the case of GLUtreatment, the beads obtained lose their elasticityand deformability, and under strong strain thematerials break and small granules are formed. Heconcluded that this was a serious drawback forlarge-scale applications.

4.7. Desorption of dyes

Polymeric adsorbents present considerable ad-vantages such as their high adsorption capacity,selectivity and also the facility of regeneration. Theregeneration of the adsorbent may be cruciallyimportant for keeping the process costs down and toopen the possibility of recovering the pollutantextracted from the solution. For this purpose, it isdesirable to desorb the adsorbed dyes and toregenerate the chitosan derivative for another cycleof application. Generally, the regeneration ofsaturated chitosan for non-covalent adsorptioncan be easily achieved by using an acid solution asthe desorbing agent.

Chatterjee et al. [83] proposed to desorb the dyefrom the beads by changing the pH of the solutionand they showed that the beads could be reused fivetimes without any loss of mechanical or chemicalefficacy. The rate of desorption was found toincrease with an increase in pH of the eluent. Theyconcluded that the chitosan was recyclable. This isinteresting because adsorption processes could beconsidered as potential methods for the decontami-nation of the effluents of textile industries since dyescan be selectively adsorbed, concentrated, and thenrecycled. Though CAC has received a great deal ofattention, this material is frequently non-selectiveand difficult to reuse [82]. Hu et al. [110] reportedthat the dyes adsorbed on chitosan nanoparticlescould be desorbed in an alkaline medium (causing

the de-aggregation of the particles) but not in aneutral medium. Trung et al. [108] showed thatdecrystallized chitosan can be regenerated bysufuric acid and was reusable more than 10 times.They hypothesized that the protons of the sulfuricacid are effective in reducing the dissociation of theanionic groups of the dyes. Lima et al. [58] reportedthat grafted chitosan can be regenerated withaqueous solutions containing sodium chloride with-out the use of organic solvents or pH modification.Since the interaction between BB 9 and graftedchitosan are driven mainly by chemisorption,organic solvents could be good candidates for theregeneration of the materials, as suggested by Criniet al. [72]. After saturation, the materials are easilyregenerated using ethanol as extraction solvent. Theauthors indicated that the adsorption capacitiesremained unchanged after this treatment: thisshowed both the chemical stability of the materialsand reproducibility of the values. Guibal [18]pointed out that adsorbent recycling was necessaryfor making the use of chitosan cost-efficient forenvironmental applications. However, as yet there islittle literature on this topic.

5. Adsorption mechanisms

The first major challenge for the adsorption fieldis to select the most promising types of adsorbent,mainly in terms of efficiency and low cost. The nextreal challenge is to clearly identify the adsorptionmechanism(s), in particular the interactions occur-ring at the adsorbent/adsorbate interface.

Two mechanisms are clearly established for theinterpretation of metal adsorption on chitosanmaterials, i.e. electrostatic interactions in acid media(ion-exchange) and metal chelation (coordination),although the formation of ion pairs has also beenreported [18,19,21,22]. Metal ion adsorption isassumed to occur through single or mixed mechan-isms including coordination on amino groups in apendant fashion or in combination with vicinalhydroxyl groups, and ion-exchange with protonatedamino groups through proton exchange or anionexchange, the counter ion being exchanged with themetal anion. The nature of the reaction dependsupon several parameters related to the adsorbent(ionic charge), to the solution (pH) and thechemistry of the metal ion (ionic charge, ability tobe hydrolyzed and to form polynuclear species). Formore details on these mechanisms, two recentreviews can be consulted [18,19].


On the contrary, for dye molecules, the mechanismsby which adsorption onto chitosan occurs has been amatter of considerable debate with surface adsorption,chemisorption, diffusion and adsorption-complexationbeing the prevalent theories. Among the large numberof papers dedicated to the removal of dyes by chitosan-based materials, most focus on the evaluation ofadsorption performances and only a few aim atgaining a better understanding of adsorption mechan-isms. Different studies have reached different conclu-sions. This can perhaps be explained by the fact thatdifferent kinds of interactions such as chemicalbonding, ion-exchange, hydrogen bonds, hydrophobicattractions, van der Waals force, physical adsorption,aggregation mechanisms, dye–dye interactions, etc.,can act simultaneously. It is important to note thatvariation in chitosan preparation and actual metho-dology often adds to the complication. Wide ranges ofchemical structures, pH, salt concentrations and thepresence of ligands also makes the comparison ofresults difficult.

In general, the mechanism for dye removal byadsorption on an adsorbent material may beassumed to involve the following four steps:

(i)
bulk diffusion: migration of dye from the bulkof the solution to the surface of the adsorbent;
(ii)
film diffusion: diffusion of dye through theboundary layer to the surface of the adsorbent;
(iii)
pore diffusion or intraparticle diffusion: trans-port of the dye from the surface to within thepores of the particle;
(iv)
chemical reaction: adsorption of dye at an activesite on the surface of material via ion-exchange,complexation and/or chelation.
Heat transfer may, in theory, be added. However,due to the heat transfer properties of water, kinetic

chitosan

NH2

H+

SO3Na

H2Odye dye

protonation

dissociation

electrostatic int

+ SO3-dyeNH3

+

chitosan

Scheme 1. Mechanism of anionic dye adsorpt

limitations due to heat transfer can be neglected.Providing sufficient stirring to avoid particle andsolute gradients in the batch system also makes itpossible to ignore bulk diffusion, which can beconsidered instantaneous. So, the most importantsteps are film diffusion, pore diffusion and chemicalreaction.

Previously, several studies [49–56] showed thatamine sites were the main reactive groups for dyes,though hydroxyl groups (especially in the C-3position) might contribute to adsorption, and theintermolecular interactions of the dye molecules aremost probable in the chitosan-dye systems. Theoccurrence of an interaction in aqueous solutionsbetween the hydrophilic chitosan and the anionichydrophobic dyes was proved by Muzzarelli and co-workers [53,54] using optical and thermodynamictechniques. It is now recognized that chemisorption(ion-exchange, electrostatic attractions) is the mostprevalent mechanism with the pH as the main factoraffecting adsorption. Chemisorption, a strong typeof adsorption in which molecules are not exchangedbut electrons may be exchanged, is commonly citedas the main mechanism for the adsorption ofanionic dyes in acidic conditions. Scheme 1 brieflydescribes the mechanism: in the presence of H+, theamino groups of chitosan become protonated; also,in aqueous solution, the anionic dye is first dissolvedand the sulfonate groups in the case of acid orreactive dyes dissociate and are converted to anionicdye ions; the adsorption process then proceeds dueto the electrostatic attraction between these twocounterions. In general, as the initial dye concentra-tion increases, the equilibrium pH decreases. This isconsistent with the principles of ion-exchange sinceas more dye molecules are adsorbed onto material,more hydrogen ions are released, thereby decreasingthe pH.

NH3+

SO3- + Na+

eraction

NH3+ dye-O3S

ion by chitosan under acidic conditions.


The Scheme 1 is accepted by numerous authors asthe main adsorption mechanism. Juang and co-workers [86–93] carried out different studies tochallenge the theory that amino groups wereprimarily responsible for dye binding in crosslinkedchitosan. Their investigations centered on theadsorption of anionic dyes by chitosan. Theyprepared several types of chitosan including flakes,beads and composite materials, and found a strongcorrelation between the functional groups presenton the adsorbent surface and their adsorptioncapacity for dye molecules. Since chitosan wassoluble in water at acid pH, the electrostaticproperties of chitosan were pH responsive and thepH played an important role in chitosan-basedadsorbent processes. They concluded that thematerials had particularly high selectivity withrespect to the extraction of dyes by ion-exchange.Sakkayawong et al. [109] reported that the mechan-ism of adsorption of RR 141, a reactive dye, bychitosan under acidic conditions was by chemisorp-tion, while under caustic conditions it was both byphysical interaction (rapid surface adsorption) andchemical adsorption. Surface adsorption is anothermechanism by which dye molecules may be boundto chitosan. This mechanism is a surface reactionwhere a molecule is attracted to a charged surfacewithout the exchange of ions or electrons. Otherstudies have found evidence that chitosan takes updyes by surface adsorption. Uzun and Guzel [115]reported that both physical adsorption and chemi-sorption occurred simultaneously between O II andchitosan. They concluded that physical surfaceinteraction occurring with multilayer adsorption,played an important role in the adsorption mechan-ism. In another work [112], they reported that,because the BET specific surface area of chitosanwas very low (0.65m2/g), mechanisms were mainlycontrolled by surface diffusion, and at lowertemperatures, surface diffusion was even moredominant. Using kinetic studies, they showed thatthe mechanism of action was surface adsorptionrather than ion-exchange. On the contrary, Cestariet al. [62] arrived at different conclusions. Theypointed out the importance of site accessibility andthat the low specific surface areas implied thatpurely physical adsorption onto the surface of thebeads was not significant.

Other authors [72,73,80,81,83,91,95,102,103] con-cluded that the uptake of dyes on flake- and bead-types of chitosan, and grafted chitosan may proceedmainly through ion-exchange mechanisms. Gibbs

et al. [81] reported that the adsorption of AG 25 onraw and preprotonated chitosan consisted of twomain steps: a rapid surface adsorption followed bydiffusion and chemisorption of dye molecules in thepolymer network. In another work, they alsoindicated that RB 5 was bound to chitosan notonly through electrostatic attractions but also a dyeaggregation mechanism which can play a role undercertain experimental conditions [80]. Crini et al.[72,73], studying the adsorption of cationic dyes bygrafted chitosan, confirmed that the mechanism waschemisorption of the dye via the formation ofelectrostatic interactions. The adsorption phenom-enon mainly depends on the interaction between thesurface of the grafted chitosan and the adsorbedspecies. However, they added that the mechanismwas also due to physical surface adsorption andhydrogen bonding because of the polymer network.Adsorption increases as the surface area of theadsorbent increases. They concluded that themechanism was a multistep complex process sinceother interactions such as diffusion and hydropho-bic and steric interactions could play an importantrole. Martel et al. [95] studied the adsorption ofacid, direct, mordant and reactive dyes on rawchitosan, chitosan beads and cyclodextrin-graftedchitosan. They observed that the materials did notinterfere with the dye molecule in the same mannerand suggested the presence of different interactionsin the adsorption mechanism simultaneously. Sahaet al. [102], studying the adsorption of an azo dyeonto chitosan flakes by kinetic studies concludedthat the adsorption mechanisms were both trans-port- and attachment-limited. In other words,the mechanism is separated in two processes: thetransport of dye from the bulk solution to thesurface of the flakes followed by the attachment ofthe dye to chitosan by chemisorption. Wu et al. [91]also showed that intraparticle diffusion plays animportant role in the adsorption mechanism,together with ion-exchange mechanisms. Similarconclusions have been reached in the case ofcrosslinked chitosan beads. Chiou et al. [67]demonstrated that the strong electrostatic interac-tions in acidic medium between the –NH3

+ groupspresent in crosslinked materials and dye anions canexplain the adsorption mechanism. Hebeish et al.[84,85] also indicated that crosslinked materialspossesses greater affinity for the adsorption ofreactive and direct dyes than for basic dyes. Theyindicated that the anionic character of the formerdyes was responsible for the greater dye adsorption,


unlike the basic dye with its cationic character,confirming the chemisorption mechanism.

Evidence has also been found that aggregation, astrong type of interaction depending on the pH, canbe involved in chitosan-dye binding. The aggrega-tion mechanism affects the size of the dye particlesand their ability to diffuse into the internal porousnetwork. Hu et al. [110] reported that nanoparticlesquickly aggregated after interacting with the dyemolecules, suggesting the replacement of the hydro-gen bonds between polymer chains by electrostaticinteractions between dyes and chains. So, anaggregation mechanism could be included. Increas-ing the dye concentration influences the aggregationmechanism as suggested by Gibbs et al. [80]. Thesize of the dye aggregate can influence its diffusivity,especially for low-porosity chitosans. The aggrega-tion phenomenon is enhanced by the presence of anelectrolyte in the solution and by ionic strength.Hydrophilic [53,54] and hydrophobic [51,72,73,105]interactions, depending on the chitosan structure,have been also proposed by several authors. Forexample, Seo et al. [51] previously pointed out thepredominant contribution of hydrophobic interac-tions. Shimizu et al. [105] also showed that theadsorption of AR 138 by crosslinked chitosaninvolved hydrophobic interactions between the(hydrophobic) dye and the hydrophobic functionsin the polymer network. Crini et al. [72,73] reportedthat dye adsorption on grafted chitosan in solutionwas essentially an exchange process and moleculesadsorbed not only because the were attracted bymaterials but also because the solution might rejectthem due to the presence of strong dye–dyehydrophobic interactions.

The difference in the degree of adsorption mayalso be attributed to the chemical structure of eachdye, as proposed by Wong et al. [97–99]. Theydemonstrated that the dye molecules, when ad-sorbed on chitosan, are more or less attached tochitosan chains in a flat or layered manner, that is,covering long chitosan macromolecules with ben-zene rings oriented parallelly (as far as possible) tothe polyamine chain of chitosan. If the attachmentof the dye were at one point only (electrostaticreaction between amino and sulfonate groups), thedye molecule would be expected to be more spatiallyoriented. This result confirms that, in addition toelectrostatic binding, there is a strong possibility ofhydrogen bonding between chitosan and dyemolecules [99]. Gibbs et al. [80] also reported thatdye molecules with phenyl groups can adopt a

planar structure and readily tend to form inter-molecular interactions that facilitate permanentaggregation under certain experimental conditions(presence of surfactant, pH).

Numerous authors concluded that the bindingwas a chemisorption reaction and the adsorptionphenomenon mainly depended on the interactionsbetween the surface of the adsorbent and theadsorbed species. However, all the studies arrive atcontrasting conclusions showing the difficulty ofusing simple models for the interpretation of theinteractions of these polymeric materials with dyes.Much work is necessary to clearly demonstrate theadsorption mechanism for the different types ofchitosan-based materials.

6. Modeling

6.1. Equilibrium isotherm models

Adsorption properties and equilibrium data,commonly known as adsorption isotherms, describehow pollutants interact with adsorbent materialsand so, are critical in optimizing the use ofadsorbents. In order to optimize the design of anadsorption system to remove dye from solutions, itis important to establish the most appropriatecorrelation for the equilibrium curve. An accuratemathematical description of equilibrium adsorptioncapacity is indispensable for reliable prediction ofadsorption parameters and quantitative comparisonof adsorption behavior for different adsorbentsystems (or for varied experimental conditions)within any given system.

Adsorption equilibrium is established when theamount of dye being adsorbed onto the adsorbent isequal to the amount being desorbed. It is possible todepict the equilibrium adsorption isotherms byplotting the concentration of the dye in the solidphase versus that in the liquid phase. The distribu-tion of dye molecule between the liquid phase andthe biosorbent is a measure of the position ofequilibrium in the adsorption process and cangenerally be expressed by one or more of a seriesof isotherm models [122–126]. The shape of anisotherm may be considered with a view topredicting if a sorption system is ‘‘favorable’’ or‘‘unfavorable’’. The isotherm shape can also pro-vide qualitative information on the nature of thesolute–surface interaction. In addition, adsorptionisotherms have been developed to evaluate thecapacity of chitosan materials for the adsorption


of a particular dye molecule. They constitute thefirst experimental information, which is generallyused as a convenient tool to discriminate amongdifferent materials and thereby choose the mostappropriate one for a particular application in givenconditions.

The most popular classification of adsorptionisotherms of solutes from aqueous solutions hasbeen proposed by Giles et al. [125,126]. Fourcharacteristic classes are identified, based on theconfiguration of the initial part of the isotherm(i.e., class S, L, H, C). The subgroups relate to thebehavior at higher concentrations. The Langmuirclass (L) is the most widespread in the case ofadsorption of dye compounds from water, and it ischaracterized by an initial region, which is concaveto the concentration axis. Type L also suggests thatno strong competition exists between the adsorbateand the solvent to occupy the adsorption sites.However, the H class (high affinity) results fromextremely strong adsorption at very low concentra-tions giving rise to an apparent intercept on theordinate. The H-type isotherms suggest the uptakeof pollutants by materials thgrough chemical forcesrather than physical attraction.

There are several isotherm models available foranalyzing experimental data and for describing theequilibrium of adsorption, including Langmuir,Freundlich, BET, Toth, Temkin, Redlich-Peterson,Sips, Frumkin, Harkins-Jura, Halsey, Henderson

Table 9

The two most popular equilibrium isotherm equations and their linear

Isotherm Equation Assumptions

Langmuir qe ¼x

m¼

KLCe

1þ aLCe

� Monolayer adsorption

� The sorption takes place

adsorbent

� Once a dye molecule occ

� The adsorbent has a fini

equilibrium, a saturation

adsorption can occur)

� All sites are identical an

� The adsorbent is structu

Freundlich qe ¼ KF C1=nFe � Multilayer adsorption

� The model applies to ads

with interaction between

� The adsorption energy e

completion of the sorpti

� This is an empirical equ

heterogeneous systems.

and Dubinin-Radushkevich isotherms. These equi-librium isotherm equations are used to describeexperimental adsorption data. The different equa-tion parameters and the underlying thermodynamicassumptions of these models often provide insightinto both the adsorption mechanism, and thesurface properties and affinity of the adsorbent.Therefore, it is important to establish the mostappropriate correlation of equilibrium curves tooptimize the condition for designing adsorptionsystems.

Various researchers have used these isotherms toexamine the importance of different factors on dyemolecule sorption by chitosan. However, the twomost frequently used equations applied in solid/liquid systems for describing sorption isotherms arethe Langmuir [123] and the Freundlich [124] modelsand the most popular isotherm theory is theLangmuir one which is commonly used for thesorption of dyes onto chitosan (see Table 8). Table 9reports the corresponding equations that can beused for fitting experimental data. The symbols usedin the equations are defined in the Nomenclaturesection. Linear regression was frequently used todetermine the most frequently used model through-out the years.

The Langmuir model was found to be the mostappropriate to describe the adsorption process inthe case of (i) RB 5 [79,81], BB 9 [73], DB [75], eosin[83], AG 27 [110], O II [115] and CV [115] on

forms and equation parameters

Linear form

at specific sites within the

upies a site

te capacity for the adsorbate (at

point is reached where No further

d energetically equivalent

rally homogeneous

Ce

qe

¼1

KL

þaL

KL

Ce

orption on heterogeneous surfaces

adsorbed molecules

xponentially decreases on

onal centres of an adsorbent

ation employed to describe

ln qe ¼ ln KF þ1

nF

ln Ce


chitosan, (ii) RR 189 [68], RR 222 [69], RO 16 [74]and AR 138 [105] on crosslinked chitosan, (iii) BB 9[58], BV 3 [106] and O II [113] on grafted chitosan.Gibbs et al. [81], studying the adsorption of RB 5 onchitosan reported that the adsorption isothermswere characterized by a steep increase in theadsorption capacity (indicating a great affinity ofthe adsorbent for the dye), followed by a plateaurepresenting the maximum capacity at saturation ofthe monolayer. This shape corresponds to thetypical Langmuir-type equation. They concludedthat the adsorption was very favorable and almostirreversible. Similar conclusions have been reportedby Saha et al. [102] who studied the adsorption of anazo dye onto chitosan flakes. They concluded thatthe isotherm was dominated by a monolayeradsorption process, in which the flakes wereprotonated and rapidly interacted with the dye toform complexes in accordance with Scheme 1. Theyassumed that the adsorbed layer was one moleculethick and the sites were homogeneous, confirmingthe applicability of the Langmuir model [123].Sakkayawong et al. [109] demonstrated that RR141 adsorption on chitosan obeyed the Langmuirmodel which suggested that biosorption occurredon the homogeneous surface. Wong et al. [98,99]reported that the Langmuir model was found toprovide the best theoretical correlation of theexperimental data for the adsorption of five aciddyes. They noted the high degree of correlation forthe linearized Langmuir relationship suggested asingle surface reaction with constant activationenergy and was the predominant adsorption step.However, they also reported that, if the wholeconcentration range was divided into three differentregions, excellent fits to the experimental data couldbe observed with the Freundlich model, especially atthe lower concentrations [99]. Hu et al. [110],studying the adsorption of AG 27 on nanochitosanfound that the sorption fitted the Langmuir modelwell, especially when the concentration was high.Crini et al. [72,73], studying the adsorption of BB 9and BB 3 by grafted chitosan, used both theLangmuir and the Freundlich models to describethe adsorption equilibrium. It was found thatthe data fitted better to Langmuir model and erroranalysis investigations highlighted the non-linearmethod as a better way to obtain the isothermparameters. Dos Anjos et al. [63] observed that bothFreundlich and Langmuir models fitted IC adsorp-tion on chitosan well, which indicated adsorptionby combined mechanisms onto a heterogeneous

surface. In other cases, the Freundlich equation waspreferred. Kim and Cho [71], studying the adsorp-tion of RB 5 on crosslinked chitosan, reported thatthe Freundlich isotherm best fitted the data over theentire pH and temperature range of the solution.Chang and Juang [86], studying the adsorption ofRR 222, AO 51 and BB 9 onto chitosan, reportedthat in a simple way the two-parameter Langmuirand Freundlich equations can treat the isotherms.The Freundlich equation was preferred for thedescription of the isotherms of reactive dyes and theisotherms of acidic and basic dyes were better fittedby the Langmuir model, depending on the dyeconcentration. However, they confirmed that theFreundlich was an empirical approach applicable tothe adsorption of single solutes within a fixed rangeof concentration [124]. This model is generallysuitable for high- and middle-concentration envir-onments and is not suitable for low concentrationsbecause it does not meet the requirements ofHenry’s law [86].

Two important points must be pointed out. Thefirst is that, despite its highly idealistic simplicity,the two-parameter isotherm model remains a usefuland convenient tool for the comparing results fromdifferent sources on a quantitative basis. Manyliquid adsorption studies on chitosan have beencarried out by fitting the Langmuir and Freundlichisotherm parameters to the experimental data asshown in Table 8. However, while this approach isconvenient for the characterization of data, it haslimited potential only for predicting behavior underconditions within the ranges of experimental mea-surements, because the assumptions of these modelsare not closely based on the actual adsorptionprocesses. Both models suffer from the drawbackthat equilibrium data over a wide concentrationrange cannot be fitted with a single set of constants.In addition, the Langmuir and Freundlich models,initially developed for modeling gas solutes onmetallic surfaces, are based on the hypothesis ofphysical adsorption. In the case of dye adsorption,which is more chemical than physical, it would bemore appropriate to consider dye adsorption withmodels based on chemical reactions, in order to takeinto account the real phenomena between thechitosan material and the dye, and also the dye–dye interactions. Therefore, the three-parameterisotherm equations such as the Redlich-Petersonand the Sips models which combine the features ofboth Langmuir and Freundlich models, are pre-ferred. These models are still being used and much


more work is required. It was demonstrated thatthe three-parameter models fitted the experimentaldata better than the two-parameter models becausethey take into account additional parameters(pH, temperature) and other interactions in theadsorption mechanism (dye–dye interactions). De-spite the number of papers published, there is as yetlittle literature containing a full study comparingvarious models and this topic clearly needs furtherdetailed research. Some conclusions can be found inRefs. [93,97,98].

The second point is related to the mathematicalmodel used. There is no doubt that mathematicalmodeling is an invaluable tool for the analysis anddesign of adsorption systems and also for thetheoretical evaluation and interpretation of thermo-dynamic parameters. However, an isotherm mayfit experimental data accurately under one set ofconditions but fail entirely under another. Inaddition, no single model has been found to begenerally applicable. This is readily understandablein the light of the assumptions associated with theirrespective deviations. In the single-component iso-therm studies, the optimization procedure requiresan error function to be defined in order toquantitatively compare the applicability of differentmodels in fitting data. To determine isothermconstants for two-parameter isotherms such as theLangmuir and the Freundlich models, two methodsare available: fitting the isotherm equation tothe data in its non-linear form or converting theequation into a linear form by transforming theisotherm variables. In the literature, linear regres-sion is the most commonly used method to estimateadsorption, and linear coefficients of determinationare preferred. However, the use of this method islimited to solving linear forms of equation whichmeasure the difference between experimental dataand theoretical data in linear plots only, but not the

Table 10

The three most popular kinetic model and their linear forms

Model Equation

Lagergrenlog

qe

qe � qt

� �¼

k1

2:303

Ho and McKay 1

ðqe � qtÞ¼

1

qe

þ k2t

Webber and Morris

errors in isotherm curves. Much work is alsorequired to this area. Recently, several studies haveshown that the linearization of a non-linearisotherm expression produces different outcomes[73,97]. Crini et al. [73] reported that linearregression and the non-linear Chi-square analysisgave different models as the best-fitting isotherm forthe given data set, thus indicating a significantdifference between the analytical methods. Theyshowed that the non-linear Chi-square test provideda better determination for the experimental data.Wong et al. [97] also reported that the values of theindividual isotherm constants changed with theerror methodology selected. They obtained contra-dicting results from linearization using differenterror functions.

6.2. Kinetic modeling

An ideal adsorbent for wastewater pollutioncontrol must not only have a large adsorbatecapacity but also a fast adsorption rate. Therefore,the adsorption rate is another important factor forthe selection of the material and adsorption kineticsmust be taken into account since they explain howfast the chemical reaction occurs and also providesinformation on the factors affecting the reactionrate. The kinetics of adsorption is also another areaof debate, and once again, differences in chitosantype, preparation, dyes and methodology examinedmakes any comparison of results difficult.

Three kinetic models (Table 10) have been widelyused in the literature for adsorption processes: (i)pseudo-first-order kinetic model (Lagergren model)[127]; (ii) pseudo-second-order kinetic model (Hoand McKay model) [128]; (iii) and intraparticlediffusion model (Webber and Morris model) [129].These kinetic models are used to examine thecontrolling mechanism of adsorption process such

Linear form

t logðqe � qtÞ ¼ log qe �k1

2:303t

t

qt

¼1

k2q2eþ

1

qe

t

qt ¼ kit1=2 þ C


as adsorption surface, chemical reaction and/ordiffusion mechanisms (Table 8). The parameters ofthe kinetic models can be obtained by suitablelinearization procedures followed by both linearand/or non-linear regression analysis. Recent workby Crini et al. [73] has shown that non-linearregression gives a more accurate determination ofparameters than linear methods.

When adsorption is preceded by diffusionthrough a boundary, the kinetics in most casesfollows the pseudo-first-order rate equation ofLagergren. Dutta et al. [75] reported that theadsorption of reactive and direct dyes on chitosanfollowed first-order kinetics and the Lagergren plotswere linear for a wide range of concentrations andcontact periods. Chang and Juang [86] and Wonget al. [97] also indicated that the pseudo-first-orderequation could well describe the adsorption pro-cesses. However, the Lagergren model was notproved to be effective in representing the experi-mental kinetic data for the entire adsorption period.In some cases though the Lagergren model providedan excellent fit with the experimental kinetic data, itfailed to predict the amount of dye adsorbedtheoretically thereby deviating from theory. So,the pseudo-second order was preferred.

The second order (and also the first order) isbased on the adsorption capacity: it only predictsthe behavior over the ‘‘whole’’ range of studiessupporting the validity, and is in agreement withchemisorption being the rate-limiting step. Thekinetics of adsorption of many dye species ontovarious chitosan materials was also found to be ofsecond-order in the literature: adsorption of AO 7[67], DR 81 [67], RR 222 [69] and RR 189 [68,94] oncrosslinked chitosan, AR 87 [83], RR 2 [67] and RB222 [90] on raw chitosan (Table 8). The applicabilityof the pseudo-second-order model suggested thatchemisorption might be the rate-limiting step thatcontrols these adsorption processes. In general, thismodel is interesting and useful since the Ho andMcKay [128] equation was found to explain thekinetics of most adsorption systems very well for theentire range of adsorption periods using differentconcentrations and chitosan dosages. In addition, ithas the following advantage: the adsorption capa-city, the pseudo-second-order rate constant, and theinitial adsorption rate can be determined from theequation without knowing any parameters before-hand [133]. Crini et al. [73] used different kineticmodels for the characterization of the adsorption ofBB 9 and BB 3 by grafted chitosan. The kinetic

measurements and their modeling showed that bothprocesses were rapid because of rapid surfacephysical adsorption and the Ho and McKayequation was more accurate at fitting the experi-mental data. They reported that, at all initial dyeconcentrations, the adsorption data were wellrepresented by the Lagergren model for only thefirst 60min and thereafter they deviated fromtheory. The adsorption data were well representedonly in the region where rapid adsorption tookplace. This confirmed that it was not appropriate touse the Lagergren kinetic model to predict theadsorption kinetics of BB 3 onto chitosan for theentire adsorption period. The adsorption systemobeys the pseudo-second-order kinetic model forthe entire adsorption period and thus supports theassumption behind the model that the adsorption isdue to chemisorption. They also showed that thekinetic parameters decreased markedly with increas-ing initial dye adsorption. The adsorption of dyeprobably takes place via surface exchange reactionsuntil the surface functional sites are fully occupied;thereafter dye molecules diffuse into the polymernetwork for further interactions and/or reactions.

Both the Lagergren, and Ho and McKay modelsbasically include all steps of adsorption (i.e. externalfilm diffusion, adsorption and intraparticle diffu-sion), they are thus pseudo-models [86]. However,using the so-called pseudo-first and pseudo-second-order equations for data interpretation isquestionable since the equations have no physicalsignificance. It is more reasonable to interpret thekinetic data in terms of mass transfer [130–132].During the past several decades, a large number ofstudies of batch adsorption have been reported inthe literature and a summary of these studies can befound in the excellent compilation reported by Tien[122]. Because the above two-lumped kinetic pseu-do-models cannot identify adsorption mechanisms,several investigators proposed to use the diffusionmechanisms such as intraparticle diffusion using theWeber and Morris equation, the Avrami model andthe Elovich equation [86]. The former modeloriginates from Fick’s second law. The validity ofthe Elovich equation suggests that the chemisorp-tion (chemical reaction) mechanism is probably ratecontrolling in the adsorption mechanism. Changand Juang [86] indicated that although the pseudo-first-order equation could describe the adsorptionprocesses well, from a lumped point of view, thebetter-fit of kinetic data by the Elovich equationinstead of by intraparticle diffusion suggested the


significance of chemisorption mechanism during theprocesses. They supposed that the coordination andreaction between the dyes and the amino andhydroxy groups on chitosan chains would besignificant and chemisorption controlled the pro-cess. Cestari et al. [62,63] indicated that the pseudo-models did not take into account the influenceof parameters such as temperature and kind ofdye. So, they suggested using the Avrami model,which is the best kinetic model to evaluate multistepadsorption phenomena at the solid/solution inter-face. However, this model cannot give interactionmechanisms.

Mass transfer involves several steps including (i)bulk diffusion, (ii) film diffusion, (iii) intraparticlediffusion and (iv) (physical and/or chemical) ad-sorption reactions. Numerous authors consider thatbulk and film diffusion can be ignored if a sufficientstirring speed is used. This is correct for bulkdiffusion but is more controversial regarding filmdiffusion. Moreover, it is usually accepted that, inthe case of physical adsorption, the adsorption itselfcan be considered as an instantaneous processes,and the adsorption kinetics are controlled either byexternal or intraparticle diffusion or by bothdiffusion mechanisms at the same time [122]. Inthe case of chemical reactions, their own kineticrates may interfere in the control of the adsorptionrate. For complete modeling of adadsorptionkinetics it would be necessary to take into accountnot only the diffusion equations but also boundaryconditions including the adsorption isotherm equa-tion [18,122]. This means that the system ofequations is very complex but, generally, it ispossible to simplify the system by separatingdiffusion steps or taking into account only diffusionsteps in the control of kinetic rates. In differentadsorption studies, the diffusion mechanisms wereconsidered independently in accordance with theassumptions that the kinetics was controlled byexternal diffusion at the beginning of the experimentand then controlled by intraparticle diffusion.

McKay [47] observed that diffusion within theparticle is much slower than the movement of thedye from solution to the external solid surfacebecause of (i) the greater mechanical obstruction tomovement presented by the surface molecules orsurface layers and (ii) the restraining chemicalattractions between dye and adsorbent. Duringadsorption of the dye from a batch system, dyemolecules arrive at the adsorbent surface morerapidly than they can diffuse away into the solid.

The dye accumulates at the surface and a (pseudo)-equilibrium is established, and further adsorption ofdye can take place only at the same rate as thesurface concentration is depleted by inward adsorp-tion. The dye uptake can be correlated to the squareof time over a large adsorption zone to getdiffusivity of the dye in adsorbent particles. Indiffusion studies, it is possible to define a rateparameter by plotting the adsorption capacity as afunction of the square root of time [47,93]. The roottime dependence may be expressed by the equationproposed by Weber and Morris [129], assuming thatthe mathematical dependence is obtained if theprocess is considered to be influenced by simplediffusion in the particles and convective diffusion inthe solution. If intraparticle diffusion is involved inthe adsorption process, then the plot of the squareroot of time versus the uptake would result in alinear relationship, and intraparticle diffusionwould be the rate-limiting step if this line passedthrough the origin. When the plots do not passthrough the origin, this is indicative of some degreeof boundary layer control and further shows thatthe intraparticle diffusion is not the only rate-controlling step, but that other processes maycontrol the rate of adsorption. The Webber–Morrisplot is also an empirically relationship but widelyused in the literature.

Several different steps in the process have beencharacterized by this simple mathematical modeland different linear sections have been identified inthe adsorption of dyes on chitosan as reported bynumerous authors [59,66,68,73,87,89,91]. All thestudies showed that the kinetics results can be usedto determine if particle diffusion is the rate-limitingstep for dye adsorption onto a material. In general,the Webber–Morris plots present a multilinearity,which indicates that two or more steps occur in theprocess. In the plots, there are three differentportions, representing the different stages in ad-sorption: an initial curved portion followed bylinear portion and then a plateau. The initial curveportion is due to surface adsorption and rapidexternal diffusion (boundary layer diffusion). Thesecond linear portion is the gradual adsorptionstage where the intraparticle diffusion is rate-controlled. The plateau (third portion) is the finalequilibrium stage, where the intraparticle diffusionstarts to slow down due to the low soluteconcentration in solution. Juang and co-workers[90,91,93] reported that adsorption kinetics werecontrolled by different mechanisms, the most


limiting of which were the diffusion mechanismsincluding the external and the intraparticle masstransfer resistances and the reaction rate. TheWebber–Morris plots gave three-stage sections,which mean an instantaneous adsorption stage, agradual adsorption stage and final equilibrium stagein sequence. As a first approximation, externaldiffusion controls the initial stage of adsorptionprocess while the second stage of the process iscontrolled by the intraparticle diffusion. Similarconclusions have been reported by other authors[72,73,81,83]. Crini et al. [72,73] reported that themultilinearity obtained using the Webber–Morrismodel showed a contribution of film diffusion onthe control of adsorption kinetics and the intrapar-ticle diffusion played an important role but was notthe rate-determining step. Two diffusion mechan-isms are involved in the adsorption rate: porediffusion (diffusion within the pore volume) andsurface diffusion (diffusion along the surface of thepores). Pore diffusion and surface diffusion occur inparallel within the adsorbent particle. They con-cluded that the mechanism was complex, involvingadsorption on the external surface, diffusion intothe bulk, chemisorption and other interactions(mainly hydrophilic and hydrophobic interactions).Gibbs et al. [81] also observed that the adsorption ofAG 25 on chitosan appeared to occur not only atthe surface of the material but in its intraparticlenetwork with chemisorption the rate-limiting step.They concluded that the resistance to intraparticlediffusion also plays an important role in the controlof mass transfer.

Several factors can affect the reaction kinetics ofdye adsorption onto chitosan. These factors includethe chemical structure of the target dye, thecharacteristics of the adsorbent (in particular, itsparticle size) and/or the experimental solutionconditions. Guibal and co-workers [80–82] reportedthat adsorption kinetics were strongly influencednot only by intraparticle diffusion resistance butalso by the affinity of the dye for the material. Theaffinity of the dye molecule for the adsorbentchanged the relative importance of the intraparticlediffusion on the control of the overall kinetics. Theconcentration of the dye could also strongly affectthe kinetics [80]. The strong effect of particle sizealso confirmed that the contribution of intraparticlediffusion resistance to the control of kinetics cannotbe neglected [82]. The greater the particle size, thegreater the contribution of intraparticle diffusionresistance to the control of the adsorption kinetics

for only slightly porous materials. They indicatedthat the size of adsorbent particles influenced boththe adsorption kinetics and equilibrium for AG 25[81] because of resistance to intraparticle diffusion,but the porosity of the sorbent and its surface areadid not control the adsorption kinetics [82] fornumerous anionic dyes. In the case of RB 5 onchitosan, they observed that the kinetic parametersvaried little and the most significant effect observedwas the decrease in intraparticle diffusivity [80].Juang et al. [93] also reported a greater effect ofparticle size on reactive dye adsorption kinetics bychitosan. These authors indicated that the greaterthe amount and the smaller the size of the chitosanparticles used, the faster the process. Wu et al. [91]found that the adsorption was faster using bead-type chitosan than the flake type.

6.3. Thermochemistry of biosorption

6.3.1. Effect of temperature

Generally speaking, the adsorption of pollutantsincreases with temperature because high tempera-tures provide a faster rate of diffusion of adsorbatemolecules from the solution to the adsorbent [134].However, it well known that temperature plays animportant role in adsorption in activated carbon,generally having a negative influence on the amountadsorbed. The adsorption of organic compounds(including dyes) is an exothermic process and thephysical bonding between the organic compoundsand the active sites of the carbon will weaken withincreasing temperature. Also with the increase oftemperature, the solubility of the dye also increases,the interaction forces between the solute andthe solvent become stronger than those betweensolute and adsorbent, consequently the solute ismore difficult to adsorb. Both of these features areconsistent with the order of Langmuir adsorptioncapacity. The adsorption of dyes by chitosan is alsousually exothermic: an increase in the temperatureleads to an increase in the dye adsorption rate, butdiminishes total adsorption capacity [21,135]. How-ever, these effects are small and normal wastewatertemperature variations do not significantly affectthe overall decolorization performance [21]. Inaddition, the adsorption process is not usuallyoperated at high temperature because this wouldincrease operation costs.

The increase in temperature affects not only thesolubility of the dye molecule (its solubility in-creases) but also the chemical potential of the


material (its potential increases), the potential beinga controlling factor in adsorption. Both effects workin the same direction causing an increase in thebatch system. In general, this could be confirmed bythe thermodynamic parameters. An increase intemperature is also followed by an increase in thediffusivity of the dye molecule, and consequently byan increase in the adsorption rate if diffusion is therate-limiting step. Temperature could also influencethe desorption step and consequently the reversi-bility of the adsorption equilibrium. So, thetemperature (and its variation) is an importantfactor affecting chitosan adsorption and investiga-tions of this parameter offer interesting results,albeit often contradictory.

Annadurai [59], studying BB 9 adsorption onchitosan, found that adsorption increased withtemperature, peaking at 60 1C. Cestari et al. [61]indicated that the adsorption behavior of anionicdyes was directly related to the adsorption temp-erature. They reported that the dimensions ofthe chitosan pores increased with temperature. Thegreater the particle pore sizes, the smaller thecontribution of intraparticle diffusion resistance.So, the increase with the temperature seems todecrease the impact of the boundary-layer effect.However, they concluded that dependencies inrelation to both the chemical structure of the dyemolecules and the temperature were not clearlyidentified. Dutta et al. [75] studying the adsorptionof reactive and direct dyes on chitosan also observedthat as the temperature of the solution increased sodid the extent of adsorption. Uzun and Guzel [114]reported that the adsorption of Rb 5 by chitosanand O II by grafted chitosan must be studied at hightemperatures. They explained their results on thebasis of strong chemical adsorption since the dyeswere more reactive at higher temperatures. Inanother recent work [112], they also indicated thatthe adsorption of RY 2 must be studied at hightemperatures. These authors concluded that theadsorption capacity of chitosan strongly increasedwith increasing temperature. The observed increasein adsorption may be attributed to the fact that onincreasing temperature, a greater number of activesites is generated on the polymer beads because ofan enhanced rate of protonation/deprotonation ofthe functional groups on the beads. The fact thatadsorption of dyes on chitosan increases with highertemperature can be surprising. Temperature is wellknown to play an important role on adsorption inactivated carbon, generally having a negative

influence on the amounts adsorbed. The adsorptionof organic compounds (including dyes) is anexothermic process (negative value of enthalpychange) which is responsible for reduction inadsorption as the temperature is increased. Asmentioned above, the physical bonding betweenthe organic compounds and the active sitesof the carbon will weaken with increasing temp-erature. The fact that an increase in temperature isfollowed by a decrease in adsorption capacitysuggests that adsorption is governed only byphysical phenomena. Also with the increase oftemperature, the solubility of the dyes also in-creases, the interaction forces between the soluteand the solvent become stronger than solute andadsorbent, consequently the solutes are moredifficult to adsorb. Both of these features areconsistent with the order of Langmuir adsorptioncapacity.

Other authors concluded that an increase intemperature leads to a decrease in the amount ofadsorbed dye at equilibrium since adsorption onchitosan is exothermic. Saha et al. [102], studyingthe adsorption of an azo dye onto chitosan flakesnoted that the adsorption capacity was remarkablyreduced with increasing solution temperature. Theyconcluded that the decrease of the equilibriumuptake with the increase in temperature means thatthe dye biosorption process is exothermic. Li et al.[94] reported that the adsorption of RR 189 oncrosslinked chitosan was slightly influenced bytemperature. Thermodynamic parameters such asthe Gibbs free energy change (DG) or enthalpychange (DH), and/or the apparent activation energy(Ea) are often used for the characterization of thetemperature effect. For example, more negativevalues of DG at higher temperatures imply thegreater driving force of adsorption at high tempera-tures than at low. The magnitude of activationenergy gives the type of adsorption, which is mainlyphysical (physisorption) or chemical (chemisorp-tion). The range of 5–40 kJ/mol of activationenergies corresponds to a physisorption mechanismand the range of 40–800 kJ/mol suggests a chemi-sorption mechanism. The values of Ea obtained intwo previous studies [68,94] indicated that theadsorption on crosslinked chitosan had a lowpotential barrier which was assigned to physisorp-tion. These values of heat of adsorption wereestimated from the integrated van’t Hoff equation,which relates the Langmuir equilibrium constant tothe temperature.


6.3.2. Thermodynamic parameters

The most important features involved in theinvestigation of adsorption phenomenon arethe adsorption isotherm and kinetics, the interfacecharacteristics, the adsorbate–adsorbent interac-tions, and also the thermochemistry of adsorption.In particular, the adsorption characteristics of amaterial can be expressed in thermodynamic para-meters such as DG (Gibbs free energy change),DH (enthalpy change), and DS (entropy change).These parameters can be calculated by using thethermodynamic equilibrium coefficient obtainedat different temperatures and concentrations. Theexpressions reported in Table 11 are used. Evalua-tion of these parameters gives an insight into thepossible mechanisms of adsorption.

The original concepts of thermodynamics as-sumed that in an isolated system where energycannot be gained or lost, the entropy change is thedriving force. In environmental engineering prac-tice, both energy and entropy factors must beconsidered in order to determine what processeswill occur spontaneously. Thermodynamic consid-erations tell us that, at constant temperature andpressure, the DG value is the fundamental criterionof spontaneity, and a negative value for DG standsfor the adsorption to take place, indicating thespontaneity of the reaction. By using the equili-brium constant obtained for each temperature fromthe Langmuir model, DG can be calculated accord-ing to the Gibbs expression (Table 11). It isimportant to note that DG is estimated from theequilibrium adsorption data under the assumptionthat the adsorption of a molecule is reversible andthat an equilibrium condition is established in thebatch system.

Table 11

Thermodynamic equations and their parameters

Expression Linear equation form

Arrhenius ln kads ¼ �Ea

RTþ ln ko

Gibbs DG ¼ �RT ln KL

van’t Hoff ln KL ¼ �DH

RTþ

DS

R

Clausius-Clapeyron DH ¼�RT1T2

T2 � T1ðln C2 � ln C1Þ

The DH and DS changes of an adsorptionreaction can be determined using the van’t Hoffplot (Table 11) and are estimated by determining theisotherm at different temperatures assuming theseparameters to be independent of temperature. Froma more random stage (in solution) to a more orderlystage (on the surface of the adsorbent) for dyemolecules, the entropy change of adsorption (DS)also has a negative sign. The sign of DS wouldindicate the direction, for adsorption (+DS), andfor desorption (�DS). As known from thermody-namics, the negative values of DG and DS require anegative adsorption enthalpy (DH), which in turnimplies that the adsorption phenomenon is exother-mic. The DH value (experimentally measured) canalso be used as a measure of the interaction forcebetween adsorbate and adsorbent, giving an indica-tion of the bonding strength.

Adsorption on solids is classified into physicaladsorption and chemical adsorption, but the divid-ing line between the two is not sharp. However,physical adsorption is non-specific, and the varia-tion of energy for physical adsorption is usuallysubstantially smaller than that of chemical adsorp-tion. Chemical adsorption is similar to ordinarychemical reactions in that it is highly specific.Typically, DH for physical adsorption ranges from�4 to �40 kJ/mol, compared to that of chemicaladsorption ranging from �40 to �800 kJ/mol. Asshown in Table 12, the DH values suggest that theadsorption process might be considered as physicaladsorption in nature. Table 12 also shows that, fordye adsorption on chitosan derivatives, negative DG

values reveal the spontaneity of the process. Theadsorption process is spontaneous in nature andmore favorable at lower concentrations of dye

Parameters

Apparent activation energy

Free energy change

with KL ¼ qe=Ce Enthalpy change

Entropy change

Enthalpy change

ARTICLE IN PRESS

Table 12

Thermodynamics and rate parameters for various dyes using chitosan

Dye Material T (1C) pH DG (kJ/mol) DH (kJ/mol) DS (J/molK) k0 (kg/g min) Ea (kJ/mol) Reference

AB Chitosan 20 3.6 �72.87 [65]

AR 87 30 �5.46 �17.10 �51.8 [83]

Azo 30 �4.97 �2.17 9.25 [102]

BB Chitosan 20 6.9 47.46 [65]

BB 1 Grafted chitosan 30 �11.89 [106]

BB 3 Grafted chitosan 25 3 �6.4 [73]

BB9 Grafted chitosan 25 �22.1 2.47 [58]

BV 3 Grafted chitosan 30 �11.67 [106]

CV Chitosan 20 �10.51 �1.09 32.14 [114]

CV Modified chitosan 20 �11.72 �23.05 �38.66 [114]

CV Modified chitosan 60 �10.17 �23.05 �38.65 [114]

IC 35 6 �9.1 �23.2 �45.8 [63]

IC 25 �2.55 �29.25 90 [101]

O II Chitosan 20 �3.35 �5.68 �7.94 [114]

RR 141 20 11 �7.10 �18.20 �37.88 [109]

RR 189 Crosslinked bead 30 3 �6.6 �52.9 �153.1 8.11� 107 75.7 [68]

RR 189 Crosslinked bead 30 3 43 [94]


compared with higher concentrations, as reportedby Chatterjee et al. [83]. The appreciably low freeenergy values indicated saturation of the processand the enthalpy values suggested that the reactionwas exothermic, and especially favorable at lowtemperature. DG was more negative with decreasingtemperature, which suggested that lower tempera-ture makes the adsorption easier, as observed byUzun and Guzel [114]. However, the authors [114]concluded that the dye adsorption by chitosan mustbe studied at high temperatures. Saha et al. [102],studying the adsorption of an azo dye onto chitosanflakes, also reported that the Gibbs free energydemonstrated that the adsorption was favorableand the pronounced chitosan–dye interaction wasreflected in the values of enthalpy. Prado et al. [100]concluded similar observations. Indigo carmine/chitosan interaction showed favorable enthalpicand entropic processes, reflecting thermodynamicstability of the complex formed (Scheme 1), whilethe dye/chitin interaction showed an exothermicenthalpy and a highly unfavorable entropic effect,resulting in a non-spontaneous thermodynamicsystem. Other observations were also found: thepositive values of DH [58] and DS [102,110,114]suggested the endothermic nature of adsorption andincreased randomness at the solid/solution interfaceduring the adsorption of dye on chitosan deriva-tives. A low value of DS indicated that noremarkable changes in entropy occur [114]. Chenet al. [65], studying the adsorption of acid (AB) andbasic (BB) dyes on chitosan reported that the

negative values of DH for AB dye indicated thatheat was released during the adsorption process andthe positive value for BB indicated that heat wasabstracted from the surroundings They concludedthat the effect of temperature on the adsorption ofcationic dye was peculiar. There is an sharp increasein equilibrium adsorption with increased tempera-ture, which was thought to be due to enhanced dyemobility and a temperature-induced swelling effectwithin the internal structure of the chitosan,allowing the large dye ions to penetrate into theparticles. In addition, for AB, the adsorption rateand intraparticle diffusion coefficient were muchlarger than for BB. In other words, the intraparticlediffusion of AB was more rapid. The diffusioncoefficient decreased with increasing temperature.This could be explained by the fact that withincreasing temperature, the amount of dye adsorbedon the active sites increased in the early stages,leading to a decrease in the mobility of the diffusionmolecules to pass through for adsorption in thelong-term stage. Thus, the diffusion rate of dyes inthe intraparticle diffusion process decreased withincreasing temperature.

In general, in the external mass transport process,the values of the diffusion coefficient increase as thetemperature of adsorption increases. When thetemperature increases, the thickness of the bound-ary layer surrounding the adsorbent and the masstransport resistance of the adsorbate in the bound-ary layer decreases. Thus, the diffusion rate of dyesin the external mass transport process increases with


temperature, while in intraparticle diffusion, thecoefficient of diffusion values decrease with increas-ing temperature. At low temperature, the diffusioncoefficient of external mass transport is slightlylower than the diffusion coefficient of intraparticlediffusion. So, at low temperature, dye adsorption islimited by the external mass transport. With theincrease of temperature external mass transportbegins to play a major role in dye adsorption bychitosan. One of the reasons for the positive changesof the enthalpy and entropy could be the release ofnumerous water molecules. The adsorption of thehydrated (poly)anions onto a hydrophilic polymernetwork inevitably disturbs the order of watermolecules in the nearest environment and releasesthem to the external liquid. In other words,adsorbed molecules are attracted probably due tolong-distance electrostatic interactions between op-positely charge groups. During the formation of theionic bonds between the dye and the polymer, thecounterions should gain a higher degree of freedomand increase the entropy.

7. Economic aspects

Research is mainly focused on the technicalperformances of chitosan derivatives, while theireconomic aspect is usually neglected. Cost isactually an important parameter for comparingadsorbent materials. According to Bailey et al.[136], a sorbent can be considered low cost if itrequires little processing, is abundant in nature, or isa by-product or waste material from anotherindustry. Chitosan-based materials display econom-ic advantages:

�
Chitin is a material obtained from natural rawresources. It is only commercially extracted fromcrustaceans which are conveniently available aswaste from processing shellfish. The wastesconsists of chitin (20–30%), proteins (20–40%),salts (mainly carbonate and phosphate, 30–60%)and lipids (0–14%) [35]. These proportions varywith species and season. Several countries possesslarge unexploited crustacean resources, especiallyin Asia. � Chitin and chitosan are now produced commer-
cially at low cost and their production is alsoeconomically interesting, especially if it includesthe recovery of carotenoids. A prerequisite forthe greater use of chitin in industry is cheapmanufacturing processes and/or the development

of profitable processes to recover chitin and by-products such as proteins and pigments. It isimportant to note that the recovery of theseproducts from waste is an additional source ofrevenue Crustacean shells contains considerablequantities of carotenoids which so far have notbeen synthesized, and which are marketed as afish food additive in aquaculture, mainly forsalmon [21]. In addition, calcium carbonatewhich is another major component of crab shells,is converted to calcium oxide and sodiumcarbonate [12]. Pigments may be also recoveredas high value side products.
� The production of the chitosan-based materials is
economically feasible because they are easy toprepare with relatively inexpensive chemicalreagents under mild conditions. The proceduresalso require relatively harmless chemicals.

However, the industrial isolation of the polymersis restricted due to problems of environmentalpollution. The traditional method of extractioncreates its own environmental problems as itgenerates large quantities of concentrated effluentcontaining polluting bases and degradation pro-ducts and presenting inconsistent physicochemicalproperties. At the same time, the conversion tochitosan at high temperature with strong alkali cancause variability of product properties and chitosanquality, and can also increase the processing costs.This also appears to have limited potential forindustrial acceptance. Recently, some other sourcessuch as yeast and fungi (zygomycetes) have begun tobe employed to obtain chitosan. They can be readilycultured in simple nutrients and used as analternative source of chitosan. With advances infermentation technology chitosan preparation fromfungal cell walls will become an alternative route forthe production of this polymer via an eco-friendlypathway.

8. Concluding remarks

The state-of-the-art in the field of biosorption ofdyes by chitosan using batch systems is reviewed inthis paper, based on a substantial number ofrelevant references published recently. Of course,this is an ambitious project since a direct compar-ison of data obtained using different materials isdifficult to make. The experimental conditions usedin the batch system are also not systematically the


same. Nevertheless, the following conclusions maybe reached:

�
The works reviewed above indicate that bioad-sorption onto chitosan is becoming a promisingalternative to replace conventional adsorbentsused for decolorization purposes. Outstandingprogress has been made, demonstrating theapplication of chitosan and crosslinked chitosanin dye bioadsorption. These materials areefficient in dye removal with the additionaladvantage of being cheap, non-toxic and bio-compatible. � There is abundant literature concerning the
evaluation of adsorption performances of chit-osan, especially in terms of adsorption capacity(amount of dye adsorbed). At least 100 dyeshave been studied so far. All the studies showedthat chitosan had an extremely high affinity formany classes of dyes. In particular, it hasdemonstrated outstanding removal capacitiesfor acid, reactive and direct dyes. However,dependencies in relation to the chemical structureof the dyes were not clearly identified and thereis, as yet, little information in the literature onthis topic.
� Chitosan is characterized by its easy dissolution
in many dilute mineral acids, with the remarkableexception of sulfuric acid. It is thus necessary tostabilize it chemically for the recovery of dyes inacidic solutions. Several methods have beendeveloped to reinforce chitosan stability. Theadvantage of chitosan over other polysaccharidesis that its polymeric structure allows specificmodifications without too many difficulties. Thechemical derivatization of the polymer by graft-ing new functional groups onto the chitosanbackbone may be used to increase the adsorptionefficiency, to improve adsorption selectivity, andalso to decrease the sensitivity of adsorptionenvironmental conditions.
� It is interesting to note the relationships between
physicochemical properties and/or sources ofchitosan and the dye-binding properties. Mostof the properties and potential of chitosan asadsorbent can be related to its cationic nature,which is unique among abundant polysacchar-ides and natural polymers, and its high chargedensity in solution.
� While adsorption on activated carbons is largely
independent of the pH, the adsorption of dyes onchitosan is controlled by the acidity of the

solution in the case of anionic dyes. It isimportant to indicate that a source of discrepan-cies in published studies may be related tomisunderstanding the impact of pH variationon the adsorption performance.
� However, which adsorbent is better: chitosan
(raw material, preconditioned chitosan, graftedor crosslinked chitosans) or CAC? There is nodirect answer to this question because the bestchoice depends on the dye and it is impossible todetermine a correlation between the chemicalstructure of the dye and its affinity for eithercarbon or chitosan. Each product has advantagesand drawbacks. In addition, comparisons aredifficult because of the scarcity of informationand also inconsistencies in data presentation.

Although extensive work has been done, futureresearch needs to look into some of the followingaspects:

�
The biosorbent and the dye structure: It isnecessary to continue to search for and selectthe most promising types of chitosan. To date,there is no systematic and comparative studytaking into account the physicochemical proper-ties of the different kind of dyes. A more detailedstudy appears to be necessary to show how thechemical structure of the dyes affects not only theadsorption capacities but also the understandingof adsorption phenomenon involved in theuptake of a given dye. Recently, some investiga-tors have focused on studying the influence of thechemical structure of dyes on adsorption capa-city. These studies would help in optimizing thetype and amount of chitosan. The developmentof mechanistic and mathematical models in orderto simulate the adsorption process and tocharacterize the interaction between the surfaceof the chitosan and the adsorbed species are alsoimportant aspects in future biosorption studies,and should be developed. � Real effluent: The experimental conditions should
be chosen to simulate real wastewater on thebasis of thermodynamics and reaction kineticsstudies;
� Large-scale experiment: Biosorption processes
are basically at the stage of laboratory-scalestudy in spite of unquestionable progress. Muchwork in this area is necessary to demonstrate thepossibilities on an industrial scale.


Acknowledgements

The authors wish to thank Nadia Morin-Criniand Brigitte Jolibois (LBE, University of Franche-Comte) and gratefully acknowledge the financialsupport of the OSEO ANVAR of Franche-Comte.

References

[1] Ali M, Sreekrishnan TR. Aquatic toxicity from pulp and

paper mill effluents: a review. Adv Environ Res 2001;5:

175–96.

[2] Pokhrel D, Viraraghavan T. Treatment of pulp and paper

mill wastewater—a review. Sci Total Environ 2004;333:

37–58.

[3] Thompson G, Swain J, Kay M, Forster CF. The treatment

of pulp and paper mill effluent: a review. Bioresour Technol

2001;77:275–86.

[4] Robinson T, McMullan G, Marchant R, Nigam P.

Remediation of dyes in textile effluent: a critical review

on current treatment technologies with a proposed alter-

native. Bioresour Technol 2001;77:247–55.

[5] Banat IM, Nigam P, Singh D, Marchant R. Microbial

decolorization of textile-dye-containing effluents: a review.

Bioresour Technol 1996;58:217–27.

[6] Crini G. Non-conventional low-cost adsorbents for dye

removal: a review. Bioresour Technol 2006;60:67–75.

[7] Aksu Z. Application of biosorption for the removal of

organic pollutants: a review. Process Biochem 2005;40:

997–1026.

[8] Forgacs E, Cserhati T, Oros G. Removal of synthetic dyes

from wastewaters: a review. Environ Int 2004;30:953–71.

[9] Synowiecki J, Al-Khateeb NA. Production, properties, and

some new applications of chitin and its derivatives. Crit

Rev Food Sci Nutrition 2003;43:145–71.

[10] Struszczyk MH. Chitin and chitosan—part I. Properties

and production. Polimery 2002;47:316–25.

[11] Rinaudo M. Chitin and chitosan: properties and applica-

tions. Prog Polym Sci 2006;31:603–32.

[12] Peter MG. Application and environmental aspects of chitin

and chitosan. J M S Pure Appl Chem 1995;32:629–40.

[13] Cheng WP, Chi FH, Yu RF, Lee YC. Using chitosan as a

coagulant in recovery of organic matters from the mash

and lauter wastewater of brewery. J Polym Environ 2005;

13:383–8.

[14] Meyssami B, Kasaeian AB. Use of coagulants in treatment

of olive oil wastewater model solutions by induced air

flotation. Bioresour Technol 2005;96:303–7.

[15] Huang C, Chen S, Pan JR. Optimal condition for

modification of chitosan: a biopolymer for coagulation of

colloidal particles. Water Res 2000;34:1057–62.

[16] Franks GV. Stimulant sensitive flocculation and consolida-

tion for improved solid/liquid separation. J Colloid Int Sci

2005;292:598–603.

[17] Chen L, Chen D, Wu C. A new approach for the

flocculation mechanism of chitosan. J Polym Environ 2003;

11:87–92.

[18] Guibal E. Interactions of metal ions with chitosan-based

adsorbents: a review. Sep Purif Technol 2004;38:43–74.

[19] Varma AJ, Deshpande SV, Kennedy JF. Metal complexa-

tion by chitosan and its derivatives: a review. Carbohydr

Polym 2004;55:77–93.

[20] Struszczyk MH. Chitin and chitosan—part II. Applications

of chitosan. Polimery 2002;47:396–403.

[21] Ravi Kumar MNV. A review of chitin and chitosan

applications. React Funct Polym 2000;46:1–27.

[22] No HK, Meyers SP. Application of chitosan for treatment

of wastewaters. Rev Environ Contam Toxicol 2000;163:

1–28.

[23] Jayakumar R, Prabaharan M, Reis RL, Mano JF. Graft

copolymerized chitosan—present and status applications.

Carbohydr Polym 2002;62:142–58.

[24] Muzzarelli RAA, Muzzarelli C. Chitosan chemistry:

relevance to the biomedical sciences. Adv Polym Sci 2005;

186:151–209.

[25] Sashiwa H, Aiba S. Chemically modified chitin and

chitosan as biomaterials. Prog Polym Sci 2004;29:

887–908.

[26] Berger J, Reist M, Mayer JM, Felt O, Peppas NA, Gurny

R. Structure and interactions in covalently and ionically

crosslinked chitosan hydrogels for biomedical applications.

Eur J Pharm Biopharm 2004;57:19–34.

[27] Kato Y, Onishi H, Machida Y. N-succinyl-chitosan as a

drug carrier: water-insoluble and water-soluble conjugates.

Biomaterials 2004;25:907–15.

[28] Khor E, Lim LY. Implantable applications of chitin and

chitosan. Biomaterials 2003;24:2339–49.

[29] Senel S, McClure SJ. Potential applications of chitosan in

veterinary medicine. Adv Drug Delivery Rev 2004;56:

1467–80.

[30] Krajewska B. Applications of chitin- and chitosan-based

materials for enzyme immobilizations. Enzyme Microbial

Technol 2004;35:126–39.

[31] Macquarrie DJ, Hardy JJE. Applications of functionalised

chitosane in catalysis. Ind Eng Chem Res 2005;44:

8499–520.

[32] Guibal E. Heterogeneous catalysis on chitosan-based

materials: a review. Prog Polym Sci 2005;30:71–109.

[33] Krajewska B. Membrane-based processes with use of

chitin/chitosan materials. Sep Purif Technol 2005;41:

305–12.

[34] Cagri A, Ustunol Z, Ryser ET. Antimicrobial edible films

and coatings. J Food Protect 2004;67:833–48.

[35] Agullo E, Rodriguez MS, Ramos V, Albertengo L. Present

and future role of chitin and chitosan in food. Macromol

Biosci 2003;3:521–30.

[36] Prabaharan M, Mano JF. Chitosan-based particles as

controlled drug delivery systems. Drug Deliv 2005;12:

41–57.

[37] Ravi Kumar MNV, Muzzarelli RAA, Muzzarelli C,

Sashiwa H, Domb AJ. Chitosan chemistry and pharma-

ceutical perspectives. Chem Rev 2004;104:6017–84.

[38] Giri Dev VR, Neelakandan R, Sudha S, Shamugasundram

OL, Nadaraj RN. Chitosan—a polymer with wider

applications. Textile Mag 2005;46:83–6.

[39] Bornet A, Teissedre PL. Applications and interest of chitin,

chitosan and their derivatives in enology. J Int Sci Vigne

et Vin 2005;39:199–207.

[40] Mishra G, Tripathy M. A critical review of the treatment

for decolorization of textile effluent. Colourage 1993;40:

35–8.


[41] Prabaharan M, Mano JF. Chitosan derivatives bearing

cyclodextrin cavities as novel adsorbent matrices. Carbo-

hydr Polym 2006;63:153–66.

[42] Crini G. Recent developments in polysaccharide-based

materials used as adsorbents in wastewater treatment. Prog

Polym Sci 2005;30:38–70.

[43] Giles CH, Hassan ASA, Subramanian RVR. Adsorption of

organic surfaces. IV. Adsorption of sulphonated azo dyes

by chitin from aqueous solutions. J Soc Dyers Colour

1958;74:682–8.

[44] McKay G, Blair HS, Gardner JR. Adsorption of dyes on

chitin. I. Equilibrium studies. J Appl Polym Sci 1982;27:

3043–57.

[45] McKay G, Blair HS, Gardner JR. The adsorption of dyes

on chitin. III. Intraparticle diffusion processes. J Appl

Polym Sci 1983;28:1767–78.

[46] McKay G, Blair HS, Gardner JR. The adsorption of dyes

onto chitin in fixed bed columns and batch adsorbers.

J Appl Polym Sci 1984;29:1499–514.

[47] McKay G, Blair HS, Gardner JR, McConvey IF. Two-

resistance mass transport model for the adsorption of

various dyestuffs onto chitin. J Appl Polym Sci 1985;30:

4325–35.

[48] McKay G, Blair HS, Gardner JR. Two-resistance mass

transport model for the adsorption of acid dye onto chitin

in fixed beds. J Appl Polym Sci 1987;33:1249–57.

[49] Yamamoto H. Chiral interaction of chitosan with azo dyes.

Makromol Chem 1984;185:1613–21.

[50] Maghami GG, Roberts GA. Studies on the interaction of

anionic dyes on chitosan. Makromol Chem 1988;189:

2239–43.

[51] Seo T, Hagura S, Kanbara T, Iijima T. Interaction of dyes

with chitosan derivatives. J Appl Polym Sci 1989;37:

3011–27.

[52] Smith B, Koonce T, Hudson S. Decolorizing dye

wastewater using chitosan. Am Dyest Rep 1993;82:

18–36.

[53] Stefancich S, Delben F, Muzzarelli RAA. Interactions of

soluble chitosans with dyes in water. I. Optical evidence.

Carbohydr Polym 1994;24:17–23.

[54] Delben F, Gabrielli P, Muzzarelli RAA, Stefancich S.

Interactions of soluble chitosans with dyes in water. II.

Thermodynamic data. Carbohydr Polym 1994;24:25–30.

[55] Shimizu Y, Kono K, Kim IS, Takagishi T. Effects of added

metal ions on the interaction of chitin and partially

deacetylated chitin with an azo dye carrying hydroxyl

groups. J Appl Polym Sci 1995;55:255–61.

[56] Park RD, Cho YY, Kim KY, Bom HS, Oh CS, Lee HC.

Adsorption of toluidine Blue O onto chitosan. Agric Chem

Biotechnol 1995;38:447–52.

[57] Park RD, Cho YY, La YG, Kim CS. Application of

chitosan as an adsorbent of dyes in wastewater from

dyeworks. Agric Chem Biotechnol 1995;38:452–4.

[58] Lima IS, Ribeiro ES, Airoldi C. The use of chemically

modified chitosan with succinic anhydride in the methylene

blue adsorption. Quim Nova 2006;29:501–6.

[59] Annadurai G. Adsorption of basic dye on strongly

chelating polymer: batch kinetics studies. Iran Polym J

2002;11:237–44.

[60] Annadurai G. Design of optimum response surface

experiments for adsorption of direct dye on chitosan.

Bioproc Eng 2000;23:451–5.

[61] Cestari AR, Vieira EFS, Pinto AA, Lopes ECN. Multiple

adsorption of anionic dyes on silica/chitosan hybrid 1.

Comparative kinetic data from liquid- and solid-phase

models. J Colloid Int Sci 2005;292:363–72.

[62] Cestari AR, Vieira EFS, dos Santos AGP, Mota JA, de

Almeida VP. Adsorption on of anionic dyes on chitosan

beads. 1. The influence of the chemical structures of dyes

and temperature on the adsorption kinetics. J Colloid Int

Sci 2004;280:380–6.

[63] Dos Anjos FSC, Vieira EFS, Cestari AR. Interaction of

indigo carmine dye with chitosan evaluated by adsorption

and thermochemical data. J Colloid Int Sci 2002;253:

243–6.

[64] Chang YC, Chen DH. Adsorption kinetics and thermo-

dynamics of acid dyes on a carboxymethylated chitosan-

conjugated magnetic nano-adsorbent. Macromol Biosci

2005;5:254–61.

[65] Chen L, Chen DH, Gao L. Kinetics of dyes adsorbed by

chitosan. J Dong Hua Univ (Eng Ed) 2002;19:80–3.

[66] Chiou MS, Chuang GS. Competitive adsorption of dye

metanil yellow and RB15 in acid solutions on chemically

cross-linked chitosan beads. Chemosphere 2006;62:731–40.

[67] Chiou MS, Ho PY, Li HY. Adsorption of anionic dye in

acid solutions using chemically cross-linked chitosan beads.

Dyes Pigm 2004;60:69–84.

[68] Chiou MS, Li HY. Adsorption behavior of reactive dye in

aqueous solution on chemical cross-linked chitosan beads.

Chemosphere 2003;50:1095–105.

[69] Chiou MS, Kuo WS, Li HY. Removal of reactive dye from

wastewater by adsorption using ECH cross-linked chitosan

beads as medium. J Environ Sci Health A Toxic/Hazard

Substances Environ Eng 2003;38:2621–31.

[70] Chiou MS, Ho PY, Li HY. Adsorption behavior of dye

AAVN and RB4 in acid solutions on chemically cross-

linked chitosan beads. J Chin Inst Chem Eng 2003;34:

625–34.

[71] Kim TY, Cho SY. Adsorption equilibria of reactive dye

onto highly polyaminated porous chitosan beads. Korean J

Chem Eng 2005;22:691–6.

[72] Crini G, Martel B, Torri G. Adsorption of C.I. Basic Blue 9

on chitosan-based materials. Int J Environ Pollut 2008;

33(1–4) in press.

[73] Crini G, Robert C, Gimbert F, Martel B, Adam O, De

Giorgi F, et al. The removal of Basic Blue 3 from aqueous

solutions by chitosan-based adsorbent: batch studies.

J Hazard Mater doi:10.1016/j.jhazmat.2007.08.025.

[74] Kimura IY, Laranjeira MCM, de Favere VT, Furlan L.

The interaction between reactive dye containing vinylsul-

fone group and chitosan microspheres. Int J Polym Mater

2002;51:759–68.

[75] Dutta PK, Durga Bhavani K, Sharma N. Adsorption for

dyehouse effluent by low cost adsorbent (chitosan). Asian

Textile J 2001;10:57–63.

[76] El-Tahlawy KF, Gaffar MA, El-Rafie S. Novel method for

preparation of b-cyclodextrin/grafted chitosan and it’s

application. Carbohydr Polym 2006;63:385–92.

[77] Gaffar MA, El-Rafie SM, El-Tahlawy KF. Preparation

and utilization of ionic exchange resin via graft copolymer-

ization of b-CD itaconate with chitosan. Carbohydr Polym

2005;56:387–96.

[78] Fahmy HM, Mohamed ZE, Abo-Shosha MH, Ibrahim

NA. Thermosole cross-linking of chitosan and utilization in

dx.doi.org/10.1016/j.jhazmat.2007.08.025


the removal of some dyes from aqueous solutions. Polym

Plastics Technol Eng 2004;43:445–62.

[79] Guibal E, Touraud E, Roussy J. Chitosan interactions with

metal ions and dyes: dissolved-state vs. solid-state applica-

tion. World J Microbiol Biotechnol 2005;21:913–20.

[80] Gibbs G, Tobin JM, Guibal E. Influence of chitosan

preprotonation on reactive black 5 adsorption isotherms

and kinetics. Ind Eng Chem Res 2004;43:1–11.

[81] Gibbs G, Tobin JM, Guibal E. Adsorption of acid green 25

on chitosan: influence of experimental parameters on

uptake kinetics and adsorption isotherms. J Appl Polym

Sci 2003;90:1073–80.

[82] Guibal E, McCarrick P, Tobin JM. Comparison of the

adsorption of anionic dyes on activated carbon and

chitosan derivatives from dilute solutions. Sep Sci Technol

2003;38:3049–73.

[83] Chatterjee S, Chatterjee S, Chatterjee BP, Das AR, Guha

AK. Adsorption of a model anionic dye, eosin Y, from

aqueous solution by chitosan hydrobeads. J Colloid Int Sci

2005;288:30–5.

[84] Hebeish A, Rafei R, El-Shafei A. The crosslinking of

chitosan with glutaraldehyde for the removal of dyes and

heavy metals ions from aqueous solutions. Tinctoria

2004;101:28–34.

[85] Hebeish A, Rafei R, El-Shafei A. Crosslinking of chitosan

with glutaraldehyde for removal of dyes and heavy metals

ions from aqueous solutions. Egypt J Chem 2004;47:65–79.

[86] Chang MY, Juang RS. Equilibrium and kinetic studies on

the adsorption of surfactant, organic acids and dyes from

water onto natural biopolymers. Colloid Surf A: Physico-

chem Eng Aspects 2005;269:35–46.

[87] Chang MY, Juang RS. Adsorption of tannic acid, humic

acid, and dyes from water using the composite of chitosan

and activated clay. J Colloids Int Sci 2004;278:18–25.

[88] Juang RS, Wu FC, Tseng RL. Use of chemically modified

chitosan beads for adsorption and enzyme immobilization.

Adv Environ Res 2002;6:171–7.

[89] Wu FC, Tseng RL, Juang RS. Kinetic modeling of liquid-

phase adsorption of reactive dyes and metal ions on

chitosan. Water Res 2001;35:613–8.

[90] Wu FC, Tseng RL, Juang RS. Enhanced abilities of highly

swollen chitosan beads for color removal and tyrosinase

immobilization. J Hazard Mater 2001;B81:167–77.

[91] Wu FC, Tseng RL, Juang RS. Comparative adsorption of

metal and dye on flake- and bead-types of chitosans

prepared from fishery wastes. J Hazard Mater 2000;B73:

63–75.

[92] Annadurai G, Lee DJ, Juang RS. Box–Behnken studies on

dye removal from water using chitosan and activated

carbon adsorbents. J Chin Inst Chem Eng 2000;31:609–15.

[93] Juang RS, Tseng RL, Wu FC, Lee SH. Adsorption

behavior of reactive dyes from aqueous solutions on

chitosan. J Chem Tech Biotechnol 1997;70:391–9.

[94] Chiou MS, Li HY. Equilibrium and kinetic modeling of

adsorption of reactive dye on cross-linked chitosan beads.

J Hazard Mater 2002;B93:233–48.

[95] Martel B, Devassine M, Crini G, Weltrowski M, Bour-

donneau M, Morcellet M. Preparation and adsorption

properties of a beta-cyclodextrin-linked chitosan derivative.

J Polym Sci A: Polym Chem 2001;39:169–76.

[96] Paneva D, Stoilova O, Manolova N. Magnetic hydrogel

beads based on chitosan. e-Polymers 2004;60:1–11.

[97] Wong YC, Szeto YS, Cheung WH, McKay G. Adsorption

of acid dye on chitosan—equilibrium isotherm analyses.

Proc Biochem 2004;39:693–702.

[98] Wong YC, Szeto YS, CheungWH, McKay G. Pseudo-first-

order kinetic studies of the adsorption of acid dyes onto

chitosan. J Appl Polym Sci 2004;92:1633–45.

[99] Wong YC, Szeto YS, Cheung WH, McKay G. Equilibrium

studies for acid dye adsorption onto chitosan. Langmuir

2003;19:7888–94.

[100] Miyata K, Jin C, Maekawa M. Adsorption removal of

anionic dyes in waste water from dyeing by chitosan. Sen’I

Gakkaishi 2002;58:462–5.

[101] Prado AGS, Torres JD, Faria EA, Dias SCL. Comparative

adsorption studies of indigo carmine dye on chitin and

chitosan. J Colloid Int Sci 2004;277:43–7.

[102] Saha TK, Karmaker S, Ichikawa H, Fukumori Y.

Mechanisms and kinetics of trisodium 2-hydroxy-1,

1’-azonaphthalene-3,4’,6-trisulfonate adsorption onto chit-

osan. J Colloid Int Sci 2005;286:433–9.

[103] Shimizu Y, Tanigawa S, Saito Y, Nakamura T. Synthesis

of chemically modified chitosans with a higher fatty acid

glycidyl and their adsorption abilities for anionic and

cationic dyes. J Appl Polym Sci 2005;96:2423–8.

[104] Shimizu Y, Tominaga T, Saito Y. Diethylaminoethylation

of chitin and the adsorption of acid dyes onto the resulting

polymer. Adsorpt Sci Technol 2004;22:427–37.

[105] Shimizu Y, Taga A, Yamaoka H. Synthesis of novel

crosslinked chitosans with a higher fatty diacid diglycidyl

and their adsorption abilities toward acid dyes. Adsorpt Sci

Technol 2003;21:439–49.

[106] Chao AC, Shyu SS, Lin YC, Mi FL. Enzymatic grafting of

carboxyl groups on to chitosan—to confer on chitosan the

property of a cationic dye adsorbent. Bioresour Technol

2004;91:157–62.

[107] Trung TS, Thein-Han WW, Qui NT, Ng CH, Stevens WF.

Functional characteristics of shrimp chitosan and its

membranes as affected by the degree of deacetylation.

Bioresour Technol 2006;97:659–63.

[108] Trung TS, Ng CH, Stevens WF. Characterization of

decrystallized chitosan and its application in biosorption

of textile dyes. Biotechnol Lett 2003;25:1185–90.

[109] Sakkayawong N, Thiravetyan P, Nakbanpote W. Adsorp-

tion mechanism of synthetic reactive dye wastewater by

chitosan. J Colloid Int Sci 2005;286:36–42.

[110] Hu ZG, Zhang J, Chan WL, Szeto YS. The adsorption of

acid dye onto chitosan nanoparticles. Polymer 2006;47:

5831–7.

[111] Hu ZG, Chan WL, Szeto YS. Dyeing silk with chitosan

nanoparticles. ATA J 2006;17:38–40.

[112] Uzun I. Kinetics of the adsorption of reactive dyes by

chitosan. Dyes Pigm 2006;70:76–83.

[113] Uzun I, Guzel F. Rate studies on the adsorption of some

dyestuffs and p-nitrophenol by chitosan and monocarbox-

ymethylated(MCM)-chitosan from aqueous solution. Dyes

Pigm 2005;118:141–54.

[114] Uzun I, Guzel F. Kinetics and thermodynamics of the

adsorption of some dyestuffs and p-nitrophenol by chitosan

and MCM-chitosan from aqueous solution. J Colloid Int

Sci 2004;274:398–412.

[115] Uzun I, Guzel F. External mass transfer studies during the

adsorptions of some dyestuffs and p-nitrophenol onto chitosan

from aqueous solution. Turk J Chem 2004;28:731–40.


[116] Wen YZ, Liu WQ, Fang ZH, Liu WP. Effects

of adsorption interferents on removal of reactive red 195

dye in wastewater chitosan. J Environ Sci China 2005;17:

766–9.

[117] Kofuji K, Qian CJ, Nishimura M, Sugiyama I, Murata Y,

Kawashima S. Relationship between physicochemical

characteristics and functional properties of chitosan. Eur

Polym J 2005;41:2784–91.

[118] Chang YC, Chang SW, Chen DH. Magnetic chitosan

nanoparticles: studies on chitosan binding and adsorption

of Co(II) ions. React Funct Polym 2006;66:335–41.

[119] Bodnar M, Hartmann JF, Borbely J. Preparation and

characterization of chitosan-based nanoparticles. Bioma-

cromolecules 2005;6:2521–7.

[120] Wan Ngah WS, Fatinathan S. Chitosan flakes and

chitosan-GLA beads for adsorption of p-nitrophenol in

aqueous solution. Colloid Surf A Physicochem Eng Aspects

2006;277:214–22.

[121] Knorr D. Dye binding properties of chitin and chitosan.

J Food Sci 1983;48:36–41.

[122] Tien C. Adsorption calculations and modeling. Boston,

Newton, MA: Butterworth-Heinemann; 1994. 244pp.

[123] Langmuir I. The constitution and fundamental properties

of solids and liquids. JACS 1916;38:2221–95.

[124] Freundlich HMF. Uber die adsorption in losungen. Z Phys

Chem 1906;57:385–471.

[125] Giles CH, Smith D, Huitson A. A general treatment and

classification of the solute adsorption isotherm. I. Theore-

tical. J Colloid Int Sci 1974;47:755–65.

[126] Giles CH, D’Silva AP, Easton IA. A general treatment and

classification of the solute adsorption isotherm. II. Experi-

mental interpretation. J Colloid Int Sci 1974;47:766–78.

[127] Lagergren S. Zur theorie der sogenannten adsorption

geloster stoffe. Kungliga Svenska Vetenskapsakademiens.

Handlingar 1898;24:1–39.

[128] Ho YS, McKay G. Kinetic models for the adsorption of

dye from aqueous solution by wood. Trans Chem Eng.

1998;76:183–91.

[129] Weber WJ, Morris JC. Kinetics of adsorption on carbon

solution. J Sanitary Eng Div Am Soc Civ Eng 1963;89:

31–59.

[130] Crank J. The mathematics of diffusion. 2nd ed. Oxford:

Clarendon Press; 1975. 414pp.

[131] Findon A, McKay G, Blair HS. Transport studies for the

adsorption of copper ions by chitosan. J Environ Sci

Health 1993;A28:173–85.

[132] McKay G, Blair HS, Findon A. Adsorption of metal ions

by chitosan. In: Eccles H, Hunt S, editors. Immobilisation

of ions by bio-adsorption. Chichester, UK: Ellis Horwood;

1986. p. 59–69.

[133] Ho YS. Review of second-order models for adsorption

systems. J Hazard Mater 2006;B136:681–9.

[134] Bernardin Jr, FE. Experimental design and testing of

adsorption and adsorbates. In: Slejko FL, editor. Adsorp-

tion technology: a step-by-step approach to process

evaluation and application. New York, 1985. p. 37–90

[chapter 2].

[135] Harry S. The theory of coloration of textiles. In: Johnson

A, editor. Thermodynamics of dye adsorption. 2nd ed.

West Yorkshire, UK: Society of Dyers and Colorists; 1989.

p. 255.

[136] Bailey SE, Olin TJ, Bricka M, Adrian DD. A review of

potentially low-cost adsorbents for heavy metals. Water

Res 1999;33:2469–79.