zeolitic bagasse fly ash as a low-cost sorbent for the sequestration of p-nitrophenol: equilibrium,...

RESEARCH ARTICLE

Zeolitic bagasse fly ash as a low-cost sorbent for the sequestrationof p-nitrophenol: equilibrium, kinetics, and column studies

Bhavna Shah & Ritesh Tailor & Ajay Shah

Received: 27 May 2011 /Accepted: 1 October 2011 /Published online: 19 October 2011# Springer-Verlag 2011

AbstractPurpose The purpose of the research is to investigate theapplication of bagasse fly ash, a sugar industry solid wastefor the synthesis of zeolites and their behavior for thesorption of p-nitrophenol (p-NP).Methods Zeolitic materials were prepared from bagasse flyash using alkaline hydrothermal (CZBFA) and fusion(FZBFA) treatment. Comparative batch sorption studies ofprepared zeolitic material and virgin material were undertakento determine their capacities for removal of p-nitrophenol.Results PXRD patterns revealed that zeolite P and analcimewere the dominant contents of synthesized zeolitic material.Chemical composition, morphology, and crystalline natureof CZBFA and FZBFA were characterized by XRF, FTIR,and SEM. The Langmuir, Freundlich, Dubinin Redushk-wich, and Temkin sorption isotherms were applied tocompare the sorption nature and capacity of synthesizedCZBFA and FZBFA with virgin BFA. For each sorbent-p-NP system, a pseudo-second-order kinetic model describedthe sorption kinetics accurately. The thermodynamics of thep-NP-sorbent systems exhibit an exothermic sorptionprocess. Intraparticle diffusion model shows that thesorption rate was controlled by film diffusion followed bypore diffusion. Regeneration of sorbents was carried out by

desorption studies with HCl, NaOH, and SDS detergent.The column studies were performed for the practical utilityof sorbents, and breakthrough curve were obtained, whichexhibit higher sorption capacity than batch method.Conclusion The sorption capacities of the synthesizedzeolites had improved sorption capacities for the sequestra-tion of p-NP and can be utilized as low-cost sorbents fortreatment of p-nitrophenolic wastewater.

Keywords p-Nitrophenol . Zeolitic bagasse fly ash .

Sorption . Kinetics . Desorption . Column

1 Introduction

Phenolic derivatives are products and raw materials of themany chemical and allied industries such as petrochemical,oil refinery, plastic, explosives, azo dyes, pigments, leather,paint, pharmaceutical, coking plant, steel, and pesticidesindustries (Mall et al. 2006). The improper discharge ofthese compounds in water bodies over a long period cancause the deleterious effects on water environments, whileits intake by both human and animals causes respiratorytract, liver and kidney damage, central nervous systemimpairment, diarrhea, excretion of dark urine, and maycause inflammation of affected body parts. Among phenolicderivatives, p-nitrophenol (p-NP) is an important finechemical intermediate, serving as a precursor for pesticidesand pharmaceutical products. It had been identified by theEnvironmental Protection Agency as one of the persistent,bioaccumulative and toxic chemicals due to its extensiveimpact on the deterioration of the water environment(Release and pollution prevention report, EPA 2000). Thefrequent applications of this chemical for agriculture,industrial, and defense purposes had lead to its presence

Responsbile editor: Euripides Stephanou

B. Shah (*) : R. TailorDepartment of Chemistry, Veer Narmad South Gujarat University,Surat 395007( Gujarat, Indiae-mail: [email protected]

A. ShahScience and Humanities Department,Vidyabharti Polytechnic trust,Umrakh, Bardoli-394 345,Surat, Gujarat, India

Environ Sci Pollut Res (2012) 19:1171–1186DOI 10.1007/s11356-011-0638-6

in water bodies. According to Toxic Release Inventory (TRI1992), about 3,175 kg p-NP was disposed off by under-ground injection and 113.4 kg disposed off in off-sitelandfills by industries. The minimization and control ofsuch organic pollutants is essential for healthy waterenvironment.

In recent years, a variety of techniques have beenimplemented for the treatment of phenolic wastewater withlimited success (Pandit and Gogate 2004) viz. ozonolysis,photolysis, photocatalytic decomposition, catalytic oxida-tion, coagulation, reverse osmosis by membrane filtration,electrochemical oxidation, ion exchange, biological meth-ods, enzyme treatment, solvent extraction, advanced oxida-tion processes, activated sludge, etc. Conventionally,biological treatment, activated carbon sorption, reverseosmosis, ion exchange, and solvent extraction are the mostwidely used techniques for removing phenols and relatedorganic substances (Dabrowski et al. 2005; Gupta andKarim 2003). Among these methods, sorption is the mostversatile and extensively used method for the removal ofphenolic toxic pollutants (Crini 2006; Wang and Wu 2006;Juang and Lin 2009; Zümriye 2005) and activated carbonhas been the most frequently used sorbent since more thanthree decades (Dabrowski et al. 2005). However, itsutilization as a sorbent in developing countries has beenlimited due to some economic drawbacks, such as highcapital, operational costs, regeneration cost, and problemsassociated with residual disposal. Furthermore, treatmentmethods need to be continuously modified and developedto keep the pollutants below their permissible limits. Theseshortcomings have stimulated interest to investigate thefeasibility of using cheaper materials bagasse fly ash (BFA)for the treatment of organic wastewaters. The removal ofnitro-substituted phenols had also been investigated earlierand the results show that sorption occurs due to the surfaceproperties possessed by the coal fly ash (Singh and Nayak2004). The kinetics of sorption of 1, 2-dihydroxybenzene,1, 3-dihydroxybenzene and 1-hydroxy-4-nitrobenzene ontofly ash where the process was found to consist of bothsurface sorption and pore diffusion with external transportthat mainly governs the rate-limiting process (Sarkar et al.2003).

There has been good research done on conversion ofcoal fly ash into zeolites and its utilization for the metal anddyes removal (Wang and Wu 2006; Wang et al. 2006, 2007)but in none of the open literature could found the study onthe sorption capacity and its application for the removal oforganic compounds by zeolitic bagasse fly ash. It is notunreasonable to expect bagasse fly ash and zeolitic bagassefly ash to behave differently than many of those reported inthe open literature, as far as its sorptive properties areconcerned, because it depends on the origin of thesugarcane bagasse and the conditions of combustion

prevailing during its formation. The major interest of thispaper is to explore the influence of physicochemicalcharacteristics of zeolitic bagasse fly ash (CZBFA andFZBFA), which are produced by different treatments ofBFA and their utilization on the liquid-phase extraction ofp-NP.

The sorption isotherms were expressed by the well-known Langmuir, Freundlich, Temkin, and DubininRedushkwich models and a comparison was made for theapplicability of these models. The effect of temperature onthe p-NP sorption was examined and the thermodynamicdata were also evaluated. The kinetic and dynamic studiesof p-NP sorption were estimated by pseudo-first order,pseudo-second order, external diffusion, intraparticle diffu-sion, mass transfer, and rate expression models. For thepractical utility of the CZBFA and FZBFA, the parametersfor designing the fixed bed column have been calculated.

2 Methods and experimental section

2.1 Materials

All the regents used were of analytical grade. The p-nitrophenol (Merck, India) as sorbate (1,000 mg L−1) wasprepared in double-distilled water, stored in brown-colored glass bottle, and diluted to desired concentrationduring experiments. The sorbent, BFA was obtained froma local sugar mill, Shree Khedut Sahkari Khand UdhyogMandali Ltd., located at Bardoli, Gujarat, India. Theprocured BFA was washed thoroughly with double-distilled water, dried in sunlight for 8 h and then for 4 hin hot air oven at 353±5 K. The dried BFA sieved through75–90μm mesh was used for sorption experiments of p-NP. CZBFA and FZBFA were synthesized using NaOH ofAR grade (Rankem, India).

2.2 Preparation of sorbents

2.2.1 Conventional hydrothermal treatment of BFA(CZBFA)

The dried BFA (200 μm) was suspended into 3 MNaOH solution (10:1 liquid/solid ratio) in a 3-L roundbottom flask. The mixture was refluxed for 72 h withintermittent stirring at 373±5 K. The resultant mixturewas filtered and washed with double-distilled water untilthe excess free sodium hydroxide was removed. Theresultant solid product was dried at 373±10 K in hot airoven. The dried zeolitic material (CZBFA) was sievedthrough 75–90μm mesh size sieve and used for thesorption of p-NP. The yield of the final product (CZBFA)was about 75%±5%.

1172 Environ Sci Pollut Res (2012) 19:1171–1186

2.2.2 Fusion treatment of BFA (FZBFA)

The dried BFA (200 μm) size was mixed and groundthoroughly with solid sodium hydroxide in predeterminedratio (NaOH/BFA=1.2 w/w) to obtain a homogeneousmixture. The mixture was fused in a stainless steel crucibleat 823±10 K for 1.5 h. The resultant fused product wascooled to room temperature, ground further, and added to128 mL of double-distilled water. The slurry thus obtainedwas agitated mechanically in a conical flask for 12 h atroom temperature. The mixture was then crystallized understatic condition at 363 K for 6 h. The resultant solid wasseparated by filtration using Whatman filter paper no. 42(pore size ca. 2.5 μm). It was then repeatedly washed withdouble-distilled water to remove excess of sodium hydrox-ide and dried at 373±10 K in hot air oven. The driedzeolitic material (FZBFA) was sieved through 75–9μmmesh size sieve. The yield of the final product (FZBFA)was about 82%±5%. The dried BFA and zeolitic materials(CZBFA and FZBFA) were stored in air tight desiccatorstill utilized for the sorption process.

2.3 Characterization of sorbents

The specific surface area and pore volume of the sorbentswere determined using Brunauer–Emmett–Teller (BET) andBJH nitrogen adsorption and desorption methods at 77 Kusing a Micromeritics automatic surface area analyzer(Gemini 2360, Shimadzu, Japan). Chemical compositionof each sorbent was examined by X-ray Fluorescencemethod (X-ray XDL-B, Fischer scope, Japan). The mois-ture contents of the sorbents were determined by KarlFischer (1204R of VMHI) instrument. FTIR spectra wererecorded on Thermo-Nicolet iS-10 Fourier TransformInfrared (FTIR) spectrometer from 4,000 to 400 cm−1.The PXRD patterns of the sorbents were obtained usingPanalytical X-Pert Pro instrument employing nickel-filteredCuKα (1=1.5406 A°) radiations.

2.4 Surface characterization

The surface morphologies of BFA, CZBFA, and FZBFAwere analyzed by scanning electron microscope (SEM).The point of zero charge (pHpzc) values of BFA andsynthesized zeolitic materials were determined by masstitration method (Schwarz and Noh 1990).

2.5 Batch operations

The batch technique was selected for kinetic investigationsbecause of its simplicity. Batch experiments were carriedout for the determination of equilibrium time and thedevelopment of sorption isotherms. All the batch experi-

ments were performed with 25 mL 100 mg L−1 initialconcentration (C0) of p-NP (except concentration study) at303 K temperature (except temperature study) and 75–90 μm particle size of the sorbents (BFA, CZBFA, andFZBFA). The mixture was held in 50 mL airtight stopperconical flasks and agitated at 150 rpm in an incubatorshaker until equilibrium was attained. The effect ofdifferent operational variables viz. pH (2–12), dosage(0.5–10 gL−1), initial p-NP concentration (50–300mg L−1), temperature (303–333 K), and agitation time (1–24 h) were examined for the sorption of p-NP on all threesorbents. After fixed time intervals, the solution was filteredthrough Whatman filter paper no. 42 (pore size ca. 2.5 μm)and concentration of p-NP was measured at 1max 318 nmusing UV-visible spectrophotometer (UV-visible EV 300,Thermo Nicolate). The blank experiments were performedfor each study without the sorbent showed no decline in thesolute concentration, thus nullifying any sorption by eitherthe container or filter material. Each experiment wasrepeated three times (standard deviation value <0.034) andthe mean values were taken. The initial pH (pH0) of the p-NP solutions was adjusted using 1 M solution of eitherNaOH or HCL. The sorption uptake of p-NP at equilibriumqe (in milligrams per gram) was calculated using thefollowing relationship

qe ¼ ðC0 � CeÞm

V ð1Þ

where, C0 is the initial p-NP concentration (in milligramsper liter), Ce is the equilibrium p-NP concentration (inmilligrams per liter), V is the volume of the solution (inliter), and m is the mass of the sorbent (in grams).

2.6 Column study

Glass columns (25×0.5 cm) were filled with 1.0 g of eachsorbent (BFA and CZBFA) having 75–90 mesh size onglass wool support. The weighed material was made slurrywith hot water and fed slowly into the column to avoid theair pocket entrapment. Each column containing sorbent wasloaded by p-NP solution of known concentration andallowed to percolate through the column under gravitationalforce (flow rate of 30 mL h−1). The concentration of p-NPin effluent was measured at the end of the column to studythe amount of p-NP retained on the column. The columnoperation was stopped at about 90% of the sorptioncapacity (C/C0=0.90). The method chosen for the operationof fixed bed sorber is, to a large extent, dependent on theshape of the breakthrough curve obtained by plotting C/C0

versus time or volume. In most of the sorption operationsfor water and wastewater treatment, breakthrough curvesexhibit a characteristic ‘S’ shape with varying degree ofsteepness and position of breakpoint. The breakthrough

Environ Sci Pollut Res (2012) 19:1171–1186 1173

capacity of the column was determined by loading BFA andCZBFA at pH 2.0 with 0.4 mg mL−1 concentration of p-NPusing C/C0 against effluent volume (Ve) plot.

2.7 Desorption studies

Desorption studies help to elucidate the mechanism ofsorption to recover the pollutant from the spent sorbent,apart from assessing economic factor for the feasibility of asorption system and protecting the environment from adisposal problem. In batch desorption study, the loadedsorbents were washed gently with distilled water to removeany unsorbed p-NP. The loaded sorbents (0.5 g) were

resuspended in 50 mL of desorbents (0.5 mol L−1 NaOH,0.5 mol L−1 HCl, 2,000 mg L−1 anionic surfactant SDS)and equilibrated for 24 h.

3 Result and discussion

3.1 Characterization of sorbents

The values of proximate analysis and physicochemicalproperties show gain in amorphous character (Table 1).The amounts of SiO2, Al2O3, and other contents exceptNa2O within zeolites (CZBFA and FZBFA) were de-

Table 1 Physicochemical prop-erties of BFA, CZBFA, andFZBFA

Characteristics Obtained values

BFA CZBFA FZBFA

Proximate analysis

Loss on drying (%) 12.56±0.2 15.63±0.2 18.44±0.2

Moisture content (%) 11.69±0.3 12.11±0.3 16.18±0.3

Ash content (%) 75.84±0.2 70.45±0.2 65.55±0.2

Physico- properties

Specific density 1.93±0.02 2.04±0.02 2.15±0.02

Bulk density (g cc−1) 1.75±0.02 1.98±0.02 2.13±0.02

Dry density (g cc−1) 1.16±0.02 1.31±0.02 1.39±0.02

Void ratio 0.69 0.56 0.54

Porosity (fraction) 0.41 0.36 0.35

pHpzc 8.07±0.05 8.62±0.05 8.67±0.05

N2 adsorption and desorption BET and BJH models

BET

Surface area (m2 g−1) 45.6 368.3 464.7

Adsorption average pore diameter (BET) (nm) 3.38 2.75 2.83

t-Plot

External surface area (m2 g−1) 24.18 16.12 279.01

Micropore area (m2 g−1) 21.42 208.13 185.70

Micropre volume (cm3 g−1) 9.75×10−3 9.56×10−2 8.28×10−2

BJH-pore size distribution

Adsorption cumulative surface area of pores (nm) 18.59 118.00 195.40

Desorption cumulative surface area of pores (nm) 2.35 146.83 261.25

Average adsorption pore diameter (nm) 7.94 5.73 5.50

Average desorption pore diameter (nm) 6.54 4.86 4.57

Total pore volume (cm3 g−1) 3.85×10−2 2.53×10−1 3.29×10−1

Chemical constituents

SiO2% 47.44 43.89 44.87

Al2O3% 20.46 18.75 19.46

Fe2O3% 5.76 3.89 2.89

CaO% 5.12 3.12 3.22

Mgo% 4.59 3.45 4.01

Na2O% 4.66 6.79 7.46

K2O% 3.45 2.46 2.24


creased due to the dissolution (glass phase, aluminumsilicate) into the alkaline solution during alkaline hydro-thermal treatment. The alumina and silicate ions arecondensed to form an alumina–silicate gel, which isprematerial of zeolite crystal covering the outer surfaceof BFA particles. An increase in Na2O contents of CZBFAand FZBFA is caused by capturing of sodium ions toneutralize the negative charge on aluminate in zeoliticframework when zeolite crystal is formed. Alkali fusionenhances the dissolution of fly ash converting most of theash into sodium salts of silicate and aluminate. Thecontent of calcium and magnesium oxides after synthesishas been decreased. BET surface area follow the orderFZBFA (464.7 m2/g) >CZBFA (368.3 m2/g) >BFA(45.6 m2/g). This indicates that the surface area of zeoliticsorbents has been increased significantly after treatments.The average pore diameter of CBZFA (2.75 nm) andFZBFA (2.83 nm) has been found to decrease as comparedto BFA (3.38 nm) after treatment (Table 1). This may bedue to decrease in particle size. The decrease in averageadsorption pore diameter and average desorption porediameter follow the order of BFA>CZBFA>FZBFA whichconfirms the mesoporous structure of the synthesizedsorbents. The pHpzc values obtained by mass titrationmethod are 8.07, 8.61, and 8.67 pH for BFA, CZBFA, andFZBFA, respectively.

3.2 FTIR analysis

FTIR spectra of BFA, CZBFA, and FZBFA (figures notshown) exhibit a broad band at about 3,400 cm−1

indicating the presence of −OH group of the silanol (Si–OH). The band observed between 1,250–850 and 720–650 cm−1 can be ascribed to asymmetric and symmetricstretching vibrations of internal tetrahedral, TO4 (where T=Si, Al) respectively. The bands between 420 and500 cm−1 are due to bending mode of internal tetrahedral,TO4. The band at 1,096.34 cm−1 (BFA) was shifted to1,022.33 cm−1 (CZBFA), which confirms the tetrahedralcoordination of aluminum in the zeolite framework. InFZBFA, the shift of asymmetric stretching band from1,096.34 to 981.13 cm−1 and symmetric stretching bandfrom 795.07 to 765.92 cm−1 of internal tetrahedral (TO4)due to development of amorphous aluminosilicates by thereaction of dissolved Si+4 and Al+3 confirm the formationof zeolite phases (Vucinic et al. 2003; Wang et al. 2003).The amount of increased tetrahedral sites of the alumino-silicate framework of the zeolite can be enlightened by thedecrease in frequency of asymmetric stretching vibrationof tetrahedral. The band at about 1,651 and 1,441.11 cm−1

are attributed to deformation of −OH vibration of adsorbedwater and bending vibration of interstitial water (Vucinicet al. 2003).

3.3 Powder X-ray diffraction analysis

The identification of corresponding crystalline and mineral-ogical characteristics of sorbents from powder X-ray diffrac-tion (PXRD) patterns were made by comparing the diffractiondata against a database provided by “Joint Committee onPowder Diffraction Standards” and “International Centre forDiffraction Data” (JCPDS 1971; Treacy and Higgins 2001).The PXRD patterns of BFA, CZBFA, and FZBFA whichshowed (Fig. 1) a wide huge hump at lower diffraction anglein BFA diffraction pattern indicates the presence of glassphase (Inada et al. 2005). BFA exhibits the presence of α-quartz peaks (JCPDS 5–490) with several small peaks ofmullite (JCPDS 15–0776), a small peak of stilbite, and otheramorphous materials (Treacy and Higgins 2001).

The PXRD pattern of CZBFA displays comparatively flatand low intense hump. Several new sharp diffraction peaksobtained in CZBFA which were not observed in PXRD ofBFA. The observed newly intense peaks at 2θ=15.97° and2θ=26.13°, are ascribed to formation of zeolitic units inCZBFA. The zeolite P (Phillipsite, JCPDS 39–0219), Anal-cime (JCPDS 76–0901), and Chabazite (JCPDS 12–0194)were positively identified in CZBFA. Phillipsite (P) andanalcime (A) zeolites covers the major part of synthesizedzeolites with d-spacing values of 6.95, 4.22, 3.74, 3.66, 3.32,2.89, 2.68, 1.97, 1.89, 1.80, 1.71, 1.68 and 5.54, 4.82, 3.41,2.27, 2.22, 1.74, respectively (Lin and His 1995).

PXRD pattern of FZBFA shows that the hump at lowerdiffraction angle is still more suppressed as compared toBFA and CZBFA. The crystalline phases identified in theFZBFAwere zeolite P (JCPDS 39–0219), analcime (JCPDS76–0901), zeolite A (JCPDS 14–90), zeolite X (JCPDS 28–1036), and ZSM 12, calcined (JCPDS 15–274). Zeolite Pand analcime were found to be dominant in zeoliteformation during fusion treatment. From the disappearanceof quartz and mullite peaks, it is reasonable to deduce thatboth quartz and mullite in the BFA have reacted withNaOH during fusion to form sodium aluminosilicates. Theresults from the FZBFA are probably related to the fact thatmore aluminosilicates have been dissolved in the solutionsdue to the formation of sodium silicates during fusion andamorphous aluminosilicates during crystallization time.

3.4 Morphology of the sorbents

SEM method was used in understanding the surfacemorphology of BFA, CZBFA, and FZBFA (inset Fig. 1).The SEM image of BFA shows fibrous structure along withlarge irregular shapes and surfaces of silicate masses. Itshows lamellar structure with large but shallow pores andstrands in each fold. The SEM micrographs exhibit theoccurrence of spherical particles in CZBFA and FZBFA asa result of alkaline treatment. It reduces the glass-phase


characteristic as shown in XRD patterns (Treacy andHiggins 2001). The SEM photograph of CZBFA exhibitshoneycomb aperture along hollow particles with interiorvoids, creating more folded strands with deeper pores. SEMmicrograph of FZBFA exhibits a typical size of trapezohe-dral analcime crystals which is approximately 10 μm andthe morphology is similar to those reported by Lin and His(1995). In case of FZBFA, the agglomeration of particlesforming clusters can be expected due to its softer texturehaving pores. The transformation of fly ash to differentphases of zeolite can be demonstrated by SEM.

3.5 Batch sorption studies

3.5.1 Effect of pH

The sorption of p-NP (Fig. 2) decreased dramatically whenthe solution pH varied from acidic to alkaline state for all

the sorbents. The sorption of p-NP by all the sorbents wasnearly equal with increase in pH of the solution from 2.0 to7.0. It may be associated with the fact that the pKa of p-NP is7.15 and below the pH 7.0, p-NP exists in molecular form. Atlow solution pH, the sorption affinity of sorbents toward p-NPis high due to dispersive interactions which are promoted insolutions with pH below the pHpzc of the sorbents (pHpzc=8.0–8.67±0.05), at which repulsive interactions between thecharged surface groups and uncharged molecules are mini-mized. At a pH greater than the pHpzc of BFA, CZBFA, andFZBFA, the surface of the sorbent is negatively charged andpH above the pKa of the solute, p-NP exists mainly in theanionic form and is hydrophilic in nature (Ahmaruzzamanand Laxmi 2010). As the solution pH increases, theconcentration of the anionic form of p-NP and negativecharge of sorbent increases which results in high electrostaticrepulsion between the solute molecules and also between thesolute and sorbent surface. The sorption of p-NP is mainly

Fig. 1 PXRD patterns (X zeo-lite X, Z ZSM 12, L zeolite A, Pphilipsite, A analcime, C chaba-zite, Q α-quartz, M mullite, Sstilbite) and SEM micrographs(×500 magnification) of BFA,CZBFA, and FZBFA (10 μmscale)


due to the molecular form of the solute. The p-NP uptake at2 pH after 360 min is 30.72, 46.79, 49.78 mg g−1 by BFA,CZBFA, and FZBFA, respectively with the sorption efficiencyfollowing the order FZBFA>CZBFA>BFA. The highersorption capacity of FZBFA compared to CZBFA and BFAat acidic solution was due to micropore structure of FZBFA,where p-NP (molecular dimension 0.813 nm) can beintensively sorbed by micropore filling mechanism (Pan etal. 2007).

3.5.2 Effect of dosage and initial p-NP concentration

The initial rise in sorption with sorbent or sorbateconcentration is probably due to a stronger driving force,larger available surface area, and more sorption sites(figures not shown). The rate of sorption increases withdosage of sorbent as it provides a large numbers of vacantand easily accessible sites of the sorbent and sorptioncontinues till the sorbent becomes saturated by the sorbate(p-NP). The removal slowed down as the dose concentra-tion increased and attains a constant value. In initialconcentration study, the uptake of p-NP increases withinitial concentration (C0) for all systems. This is due to highconcentration gradient existing between the solution and

the solid phase. The C0 provides necessary driving force toovercome the resistances to the mass transfer between theaqueous and the solid phases and enhances the interactionbetween p-NP and sorbents (Mall et al. 2006).

3.5.3 Effect of temperature

The temperature has pronounced effect on the sorption of p-NP because the rise in temperature causes more dissociationof the p-NP molecules (figure not shown). The sorption ofp-NP on BFA, CZBFA, and FZBFA increases with increasein temperature showing endothermic nature of sorption(Tutem et al. 1998). The p-NP possesses appreciablesolubility (11.6 gL−1 at 20°C) due to hydrogen bonding inaqueous solution. These hydrogen bonds get broken withrise in temperature resulting p-NP to be less soluble.Therefore, it exhibits a higher tendency for sorption onthe surface rather than remaining in the solution (Bhatnagar2007). Among three studied sorbents, FZBFA exhibitshigher sorption capacity than CZBFA and BFA due tolarger surface area and pore volume. The efficiency ofsorbents for the removal of p-NP is FZBFA>CZBFA>BFA.

3.5.4 Evolution of thermodynamic parameters

Thermodynamic parameters were evaluated from Van’t Hoffplots at different temperatures (303 K, 313 K, 323 K, and333 K). The change of enthalpy (ΔH0) and change of entropy(ΔS0) evaluated from the slope and intercept of Eq. 2 and thevalues obtained are given in Table 2.

R lnKd ¼ �ΔG0

T¼ ΔS0 � ΔH0

Tð2Þ

where, Kd is the distribution coefficient, ΔH0 is the enthalpychange (in kilojoules per mole), ΔS0 is the entropy change(in joule per mole Kelvin), R is the gas constant (8.314 Jmol−1 K−1), and T (in Kelvin).

The negative values of ΔG0 indicate the sorption processis spontaneous without any induction period and morefavorable at higher temperatures. The positive values ofΔH0 were in the range of 5.79–17.86 kJ mol−1<20 kJ mol−1

which confirms the endothermic nature of the overallsorption process and also the sorption to be physical rather

Fig. 2 Effect of pH and time on the uptake of p-NP by blacksquares BFA, black circles CZBFA, and black triangles FZBFA,(temperature =30°C, C0=100 mg L−1, dose=2 gL−1)

Table 2 Thermodynamicparameters for sorption of p-NPon BFA, CZBFA, and FZBFA

Sorbents –ΔG0 (kJ mol−1) ΔH0 (kJ mol−1) ΔS0 (J K−1 mol−1)

303 313 323 333

BFA 5.79 6.22 6.66 7.03 6.82 41.64

CZBFA 10.98 12.05 13.06 13.97 19.29 99.99

FZBFA 14.42 15.69 16.96 17.86 20.72 116.22


than chemical involving weak attraction forces (Tutem et al.1998; Bhatnagar 2007).

The positive value of ΔS0 corresponds to an increase inthe degree of freedom of the sorbed species, suggestingweak interaction between p-NP and sorbents which isthermodynamically favorable (Tutem et al. 1998). Duringsorption, it can be assumed that the p-NP molecules on thesorbent surface were more chaotically arranged comparedto its arrangement in the aqueous solution.

Generally, sorption is an exothermic process, it is expectedthat an increase in temperature of the sorbate–sorbent systemwould result in decreased sorption capacity. However, if thesorption process is influenced by diffusion processes, thesorption capacity is expected to rise with temperature (Mall etal. 2006) as shown in the present case, because diffusion isan endothermic process. Furthermore, it may be due tochanges in sorbent pore structure or increase of theequilibrium constant with rise in temperature.

3.5.5 Effect of contact time on sorption

The amount of p-NP sorbed increases with contact timeand attains equilibrium at about 6 h for all the sorbents(inset Fig. 2), although no significant variation in residualp-NP concentration was detected after 250 min. Initiallyfor 150 min, the uptake of p-NP was fast due toavailability of a large number of vacant sorbent sites andhigh concentration gradient between solution and solidinterphases. After a time lapse, the remaining vacantsurface sites are difficult to be occupied by the solute dueto repulsive forces between the solute molecules on thesolid surface and solution phase and thus uptake ratebecomes slow near equilibrium. The sorption equilibriumis attained where the sorption rate equals the desorptionrate after which the rate of p-NP uptake remains almostasymptotic with time. The equilibrium sorptive removalfollows the order: FZBFA (49.93 mg g−1)>CZBFA(47.25 mg g−1)>BFA (31.94 mg g−1) indicating thatFZBFA has more potential to sorb p-NP than CZBFAand BFA.

3.6 Kinetics of the sorption process

The sorption of p-NP from liquid to solid phase can beconsidered as a reversible reaction with equilibrium estab-lished between two phases. The specific rate constant of thesorption for sorbate and sorbent was determined from thepseudo-first-order rate and pseudo-second-order rate expres-sions. The pseudo-first-order equation is (Ahmaruzzamanand Laxmi 2010; Shah et al. 2009):

logðqe � qtÞ ¼ log qe � ð kf2:303

Þ � t ð3Þ

where, qt (in milligrams per gram) is the amount of p-NPsorbed at time t (minutes), kf is the rate constant of pseudo-first-order sorption (min-1).

The linearity of log (qe−qt) against t plots (not shown)with good correlation coefficients confirmed its applicabil-ity for p-NP sorption on the sorbents. The values of first-order rate constant, kf and qe evaluated from the slope andintercept of the plots are given in Table 3. These valuesindicate faster removal of p-NP on the sorbents (FZBFA,CZBFA, and BFA). However, the equilibrium sorptioncapacity, qe, obtained using the intercept of the pseudo-first-order model did not yield reasonable value of qe. Thevalues were well below the monolayer capacities (BFA55.04 mg g−1, CZBFA 82.71 mg g−1, FZBFA91.99 mg g−1) found by Langmuir equilibrium isothermmodel, suggesting the sorption process is not a true first-order reaction. Furthermore, the linear plots drawn accord-ing to pseudo-first-order equation do not strictly passthrough the origin possibly due to the dependence of qevalues on the initial p-NP concentration (C0). The depen-dence of qe and kf values on C0 designates more complexsorption mechanism (Tutem et al. 1998).

3.6.1 Pseudo-second-order kinetics model

The pseudo-second-order equation represented by followinglinear Eq. 4 (Ahmaruzzaman and Laxmi 2010; Shah et al.2009)

t

qt¼ 1

ks � q2e� � þ 1

qe

t ð4Þ

where, ks is the corresponding kinetic constant (grams permilligram per minute).

The plot of t/qt versus t (Fig. 3) gives the values of initialsorption rate (h=1/ks qe

2) and equilibrium sorption capacity(qe) from the intercept and slope, respectively (Table 3).The values of correlation coefficients (Table 3) were higherand much closer to unity for pseudo-second-order com-pared to that of the pseudo-first-order kinetic model. Theresults obtained reveal that the initial sorption rate (h value)was highest for sorption of p-NP by FZBFA than that byBFA and CZBFA. It can be concluded that sorption of p-NPonto the sorbents follows pseudo-second-order kinetics anda somewhat complex mechanism of sorption instead ofsingle-step process. It may indicate that physical interactionare involved in the sorption process, which may be partlydue to H-bonding between the hydroxyl group of p-NP andsome active functional groups on sorbent surface. Accord-ing to these results, sorption can be more appropriatelydescribed by the pseudo-second-order kinetic model.Similar results were also obtained for the sorption of p-NP on activated tea (Ahmaruzzaman and Laxmi 2010) andactivated carbon fiber (Zheng et al. 2007).


It is common to assume in batch system under rapidstirring that overall rate of binding depends primarily on thediffusivity of the sorbate: diffusion through the boundarylayer of the fluid immediately adjacent to the externalsurface of the sorbent particle (film diffusion) and diffusionthrough the sorbent particles (intraparticle diffusion). So,the overall sorption process may be controlled either by oneor more steps, e.g., film or external diffusion, porediffusion, surface diffusion, and sorption on the poresurface, or a combination of more than one step.

3.6.2 External diffusion study

To distinguish between film and intraparticle diffusion, datawere fitted to the external diffusion model. The plot of ln(Ct/C0) versus time has been commonly used to describewhether the sorption process is controlled by diffusion inthe sorbent particles and consecutive diffusion in the bulkof the solution (Shah et al. 2009).

lnCt

C0

� �¼ �ked � A

V

� �� t ð5Þ

where, ked is external diffusion rate constant (centimetersper minute), Ct is the concentration at time t, A/V is externalsorption area to the total solution volume and t is sorptiontime.

The determination coefficient and external diffusioncoefficient, ked calculated from the plots (figure not shown)were presented in Table 3. The determination coefficient,R2 values >0.9 for all the studied systems would indicatethat the sorption is probably a surface phenomenon,occurring on the exterior of the sorbent particle. Thus,external diffusion plays important role in the sorption. Thevalue of R2 was comparatively low than intraparticlediffusion, suggesting that both film and intraparticlediffusion play important role in the sorption process.

3.6.3 Intraparticle diffusion study

The possibility of intraparticle diffusion (pore diffusion)was explored by using Weber–Morris intraparticle diffusionmodel (Ahmaruzzaman and Laxmi 2010; Shah et al. 2009;Weber et al. 1963)

qt ¼ kid � t1=2 þ I ð6Þwhere, kid is the intraparticle diffusion rate constant (milli-grams per gram per minute) and ‘I’ is the intercept(milligrams per gram).

Plot of qt versus t1/2 should be straight line passingthrough the origin when sorption mechanism follows theintraparticle diffusion process only. However, the dataexhibit a multilinear plot, indicating that more than oneT

able

3Kinetic

parametersforthesorptio

nof

p-NPby

BFA

,CZBFA

,andFZBFA

Pseud

o-firstorder

Pseud

o-second

order

Externaldiffusion

Intraparticle

diffusion

q e(m

gg−

1)

k f(m

in−1)

R2

q e (mgg−

1)

k s (gmg−

1min

−1)

h (mgg−

1min

−1)

R2

k ed

(cm

min

−1)

R2

I 1 (mgg−

1)

k id,1

(mgg−

1min

−1/2)

R2

I 2 (mgg−

1)

k id,2

(mgg−

1min

−1/2)

R2

BFA

11.34

9.74

×10

−30.98

7333

.99

1.53

×10

−31.76

0.99

419.76

×10

−40.92

9620

.09

0.80

0.97

6129

.82

0.11

0.98

32

CZBFA

10.64

9.72

×10

−30.98

1249

.33

1.61

×10

−33.92

0.99

624.25

×10

−40.97

0036

.51

0.73

0.97

8045

.89

0.07

0.94

28

FZBFA

7.17

1.61

×10

−20.97

3151

.41

2.42

×10

−36.42

0.99

758.63

×10

−40.97

3142

.89

0.49

0.93

7449

.15

0.04

0.99

27


step influences the sorption process. The first sharp portionindicates external surface sorption, while the second curvedportion is the gradual sorption stage, where intraparticlediffusion is the rate-limiting step. This final linear portion isthe equilibrium stage where sorption slows down due to anextremely low solute concentration. The shape of theWeber–Morris plot (Fig. 3) reveals that the sorption of p-NP onto sorbents is controlled by external mass transferfollowed by a gradual sorption stage with intraparticlediffusion dominating with macropore/mesopore diffusionand micropore diffusion (Fierro et al. 2008) on BFA,CZBFA, and FZBFA. In the present study, the plots do notpass through the origin and have intercepts, I, a measure ofthe boundary layer thickness, i.e., the larger the value ofintercept, the greater the boundary layer effect, which is inthe range of 20.09–49.15 mg g−1. The values of rateparameter, kid, calculated from the slope of linear portion ofthe plot ranges from 0.49 to 0.80 mg g−1 min−1/2 for firstportion, while for the second portion, it is in 0.04–0.11 mg g−1 min−1/2 range (Table 3). Though the coefficientvalues (Table 3) for the intraparticle diffusion model arelower than those obtained for the pseudo-second-ordermodel, the values are somewhat close to each other, whichsignifies that intraparticle diffusion is not the only ratecontrolling step (Mall et al. 2006). It is reasonable beconcluded that surface sorption and intraparticle diffusionoperate concurrently during the sorbate–sorbent interac-tions. The lower values of kid, 2 than kid, 1 suggest that p-NPdiffuses into the pores of the sorbents. As the diffusionresistance increases with time, the diffusion rate decreases,thus, sorption is a multi-step process involving transport of

p-NP to the surface of the sorbents followed by diffusioninto the interior of the pores.

3.6.4 Mass transfer study

In order to evaluate the contribution of the bulk transportand particle diffusion to the overall sorption process, themass transfer model suggested by McKay was utilized forthe determination of surface mass transfer coefficient, βL(Shah et al. 2009).

lnCt

C0

� �� 1

ð1þ mkÞ� ��

¼ lnmk

ð1þ mkÞ� �

� ð1þ mkÞmk

� �� bL � S � t

ð7Þ

where, m is the mass of sorbent per unit volume of particlefree sorbate solution (grams per liter), k is the sorptionequilibrium constant (liters per gram) obtained by multi-plying the Langmuir constants, sorption capacity (qe), andsorption energy (b), βL is the mass transfer coefficient(centimeters per minute), and S is the outer surface of thesorbent per unit volume of the particle free slurry (percentimeter).

The straight line obtained for the ln Ct=C0ð Þ�½1= 1þ mkð Þf g� versus t plots for different initial concen-trations of p-NP and at different temperatures (figure notshown) confirms the validity of McKay equation for thepresent systems. The βL values were affected by the initialconcentration as well as change in temperature (Table 4).The values of βL decreased with increase in initial p-NPconcentration, which could be the reason for the decreasingtrend of uptake of p-NP by BFA, CZBFA, and FZBFAwitha gradual increase in concentration. This may be due toincrease in saturation of sorbent surface with increase ininitial concentration. The increase in βL value with rise intemperature indicates that the sorption is faster at highertemperature. The high value of βL values suggests that thevelocities of mass transfer of p-NP from solution to sorbentphase are quite rapid (Sarkar et al. 2003). The step of bulktransport cannot represent the rate-limiting step. Therefore,the remaining sorption step, i.e., diffusion step, is the keystep in determining the rate of sorption. Also the solution isdynamically agitated throughout the interaction period; it isplausible to assume that the rate is not only limited by masstransfer from bulk liquid to the particle external surface orboundary film surrounding the sorbent surface (Sarkar et al.2003; Anirudhan et al. 2009).

3.6.5 Rate expression diffusion model

The diffusion process may be the key factor in determiningthe rate of overall process. The kinetic data have beenfurther treated by Boyedetal and Reichenberg models using

Fig. 3 Intraparticle diffusion plots (qt vs. t1/2) and pseudo-second-

order kinetic model (t/qt vs. time) of p-NP on black squares BFA,black circles CZBFA, and black triangles FZBFA (temperature=30°C,C0=100 mg L−1, dose=2 gL−1)


reported equations (Shah et al. 2009; Sarkar et al. 2003) forestablishing the practical utility of sorption process anddesigning the sorption reactor.

F ¼ 1� 6

p2X1n¼1

1

n2� exp �n2Bt

� ð8Þ

Bt ¼ �0:4977� ln 1� Fð Þ ð9Þ

Bt ¼ p2Di

r2ð10Þ

where, F is the fractional attainment of equilibrium at timet, Di is the effective diffusion coefficient of p-NP in thesorbent phase, r is the radius of sorbent particle assumed tobe spherical, n is integer (1, 2, 3, ..) defining series obtainedfor a Fourier-type analysis and B is time constant.

The linearity test of Bt versus t plot (figure not shown) atdifferent initial p-NP concentrations and different temper-atures have been incorporated to distinguish between thefilm and particle diffusion contribution in sorption process.The plots were linear but did not pass through the originwhich envisages that the external transport mainly governsthe rate-limiting process in the studied solute concentrationrange (Shah et al. 2009; Sarkar et al. 2003). To confirm thisassumption, McKay plots, log (1-F) versus t (figure notshown) at different initial p-NP concentrations and differenttemperatures have been drawn. The plots were linearindicating a purely film diffusion process governs thesorption process for all the studied p-NP-sorbent system.

The values of Di were calculated using the followingequation:

lnDi ¼ lnD0 � Ea

RTð11Þ

From the slope and the intercept of the linear plots of logDi

versus 1/T, the values of activation energy, Ea, and the pre-exponential constant (analogous to Arrhenius frequencyfactor), D0, were calculated for the sorption process. Theentropy of the activation, ΔS# was computed usingobtained D0 values (Table 5).

lnD0 ¼ ln 2:72þ ln d2 þ lnkBT

h

� �þ ΔS#

R

� ð12Þ

where, kB is Boltzman constant (1.38×10−23 J K−1), h is thePlanks constant (6.63×10−34 J s), R is the gas constant(8.3145 JK−1 mol−1), T is the temperature (Kelvin), d is thedistance between the active site of the sorbent (0.5×10−8 cm).

The values of Di are presented in Table 5. The resultsrevealed that the value of Di decreased significantly withincrease in initial p-NP concentrations due to reduction in thediffusion of p-NP into the boundary layer from the solutionand enhancement into the solid from boundary layer.

The value of Di (Table 5) increased with rise intemperature due to increased mobility of molecules anddecreased in retarding force on the solute particles. Asdiffusion is an endothermic process, the increase in sorptionwith temperature suggests that a large number of p-NPmolecules acquire sufficient energy to undergo an interac-tion with active sites at surface. The swelling effect withinthe internal structure of the sorbents with temperature mayalso enable large number of p-NP molecules to penetratefurther into the pores (Mall et al. 2006; Sarkar et al. 2003;Gupta et al. 1998). Thus, effective diffusion constant, Di inthe sorption comprises of two components—the diffusionwithin the pores of wider widths and weaker retarding forcesof electrostatic interaction accounts for the faster and the onewithin the pores of narrower mesh widths and strongerretarding forces accounts for the slower component of Di.The small negative value of ΔS# reflects that no significantchange occurs in the internal structure of sorbents during

Table 4 Mass transfer coeffi-cients (βL) for phenols sorptionby BFA, CZBFA, and FZBFA atdifferent initial p-NP concentra-tions and at different temperature

Sorbent Conc. (mg L−1) βL (cm min−1) R2 Temp. (K) βL (cm min−1) R2

BFA 100 6.68×10−7 0.9179 303 6.27×10−7 0.9833

200 2.38×10−7 0.9883 313 6.76×10−7 0.9906

300 1.69×10−7 0.9871 323 6.93×10−7 0.9864

333 7.21×10−7 0.9899

CZBFA 100 1.20×10−6 0.9761 303 1.25×10−6 0.9706

200 3.09×10−7 0.9841 313 1.64×10−6 0.9898

300 2.08×10−7 0.9974 323 2.17×10−6 0.9683

333 2.42×10−6 0.9762

FZBFA 100 3.53×10−6 0.9794 303 1.62×10−6 0.9617

200 4.92×10−7 0.9887 313 1.83×10−6 0.9637

300 2.54×10−7 0.9957 323 2.21×10−6 0.9468

333 2.60×10−6 0.9511


sorption process. These studies indicate that the diffusionplays an important role in the sorption of p-NP.

3.7 Adsorption isotherms and modeling

The fitting of equilibrium data to different isotherm modelsis an imperative step in determining a suitable model thatcan be used to describe the sorption process and relatedsorption capacity of a sorbent. The sorption equilibriumdata were fitted to the linearized equations of the Langmuir(1916), Freundlich (1906), Dubinin–Redushkwich (D–R)(1947), and Temkin (Sathishkumar et al. 2008) isothermmodels. The models for characterization of equilibrium

distribution relate the quantity qe (in milligrams per gram)as a function of concentration at a fixed temperature.

The linear regression lines obtained for Langmuirisotherm graphs of Ce/qe against Ce, give highly significantregression coefficient values closer to unity (Table 6) andsmall relative standard error of the goodness-of-fit of themodels than the other isotherm equations indicating thatsorption data better fits the Langmuir isotherm model. Thevalue of dimensionless parameter, RL, is less than unity(Table 6), which manifest that the sorption is favorableunder the applied conditions. The comparatively smallervalue of RL for sorption by FZBFA than CZBFA and BFAindicates sorption to be more feasible. According to the

Table 5 Effective diffusion coefficients Di for the sorption of p-NP by sorbents at different initial p-NP concentration and different temperature

Sorbent Conc.(mg L−1)

Di×10−8

(cm2 min−1)R2 B×10−2

(L min−1)Temp.(K)

Di×10−8

(cm2 min−1)R2 D0×10

−6

(cm2 min−1)Ea

(kJ mol−1)−ΔS#

(J K−1 mol−1)

BFA 100 4.52 0.9824 2.20 303 3.54 0.9553 0.187 4.18 64.32

200 3.87 0.9825 1.89 313 3.79 0.9784

300 3.54 0.9599 1.72 323 3.93 0.9707

333 4.13 0.9236

CZBFA 100 4.05 0.9912 1.97 303 2.62 0.9632 1.91 10.79 45.02

200 3.58 0.9922 1.74 313 3.05 0.9923

300 2.86 0.9905 1.39 323 3.39 0.9913

333 3.88 0.9934

FZBFA 100 4.88 0.9943 2.38 303 2.54 0.9926 4.61 13.04 37.70

200 3.32 0.9938 1.62 313 3.15 0.9921

300 2.44 0.9949 1.19 323 3.64 0.9935

333 4.07 0.9940

Table 6 Langmuir, Freundlich,D-R, and Temkin isothermsparameters for the sorption of p-NP on BFA, CZBFA, andFZBFA

Sorbents Parameter values Relative standard error

Langmuir qm (mg g−1) b (L mg−1) R2 RL

BFA 55.04 0.049 0.9890 0.064 0.107

CZBFA 82.71 0.177 0.9999 0.018 0.043

FZBFA 91.99 0.705 0.9992 0.005 0.008

Freundlich Kf (L g−1) n R2

BFA 12.89 3.79 0.9680 0.740

CZBFA 22.94 3.55 0.9350 0.712

FZBFA 57.38 9.79 0.9823 0.375

D-R Xm (mg g−1) β (mol2 kJ−2) R2 E (kJ mol−1)

BFA 26.97 3.23×10−5 0.9842 1.24 0.457

CZBFA 45.97 1.18×10−7 0.9131 2.05 0.420

FZBFA 48.87 3.65×10−9 0.9276 1.17 0.972

Temkin KT (L mg−1) B1 R2

BFA 4.92 8.98 0.9385 0.901

CZBFA 5.18 13.71 0.9560 0.934

FZBFA 2.54 6.97 0.9678 0.465


Langmuir assumption, the monolayer capacity of thesorbents follows the order: FZBFA (91.99 mg g−1)>CZBFA(82.71 mg g−1)>BFA (55.04 mg g−1). These values arehigher than those obtained for fly ash (7.80–9.68 mg g−1)(Singh and Nayak 2004), rice husk char (39.21 mg g−1)(Ahmaruzzaman and Sharma 2005), Samla coal(51.54 mg g−1) (Crini et al. 2002). Sorption capacity ofCZBFA and FZBFA is also higher than activated carbonprepared from coconut coir pith (72.33 mg g−1) (Anirudhanet al. 2009) and residual Samla coal (86.95 mg g−1) (Basuet al. 2001). The Freundlich isotherm plot of lnqe versuslnCe presents straight line with intercept of multilayersorption capacity, Kf, was in the range of 12.89–57.38 Lg−1

(Table 6). The values of heterogeneity factor, n, obtainedfrom the slope were greater than one which confirmsmultilayer formation at the sorbent surface. The value of nfor the sorption of p-NP is higher for FZBFA than BFA, andCZBFA demonstrate the higher sorption of p-NP on FZBFAwith multifaceted sorption process, but the regressioncoefficient is lower than Langmuir model. This observationdemonstrates the molecular interaction of the sorbate specieswith subsequent aggregation in the surface monolayer.

The Dubinin Redushkwich isotherm model is applied totest a pore-filling mechanism in micropores of the sorbent,rather than layer-by-layer formation of a film on the wallsof the pores. The value of constant related to energy, β, andsorption capacity, Xm, were obtained from the slope andintercept of the linear plot lnqe versus ε2, respectively(Table 6). The sorption capacity Xm is found to be less thanthat for the Langmuir isotherm for all the p-NP-sorbentsystems. The sorption energy, E, of the process wascalculated using the value of β, from that it can be deducedthat the sorption mechanism is either ion exchange orphysical in nature. The sorption process follows an ionexchange process when the magnitude of E is between8 and 16 kJ mol−1, while it is of a physical nature if thevalues of E<8 kJ mol−1 (Ozcan Safa et al. 2005). Theobserved values of sorption energy, E, are <8 kJ mol−1

(Table 6) suggestive of physisorption. The values of KT andB1 obtained from the Temkin plots of qe versus lnCe, areshown in Table 6. Temkin equation represents the poorestfit of experimental data than Langmuir equation. Theregression constant is higher for FZBFA than BFA andCZBFA suggesting that heat of sorption of all the moleculesin the layer decreases linearly with coverage due tosorbent–sorbate interactions. So, the sorption of p-NP onFZBFA can be characterized by a uniform distribution ofthe binding energies.

The sorption equilibrium data of p-NP on BFA followsthe order of Langmuir≥D-R>Freundlich>Temkin whilesorption data of p-NP on CZBFA follows the order ofLangmuir>Temkin>Freundlich and on FZBFA the order isLangmuir>Freundlich>Temkin>D-R.

3.8 Designing of fixed bed sorption column

Column-type continuous flow operation appears to have adistinct advantage over batch-type operation because therate of sorption depends on the concentration of solute inthe solution being treated. For column operation, thesorbents are continuously in contact with a fresh solution.A column study for the sorption of p-NP using FZBFAwasnot feasible as the column got blocked within 30 min. Thuscolumn studies were carried out only with BFA andCZBFA. The breakthrough point was selected arbitrarilyat some low-value Cb (break point concentration ofeffluent) and Cx (exhaustion point concentration) closelyapproaching C0 (influent concentration), at which thesorbent is considered to be essentially exhausted. Massunit for C and Ve are used to illustrate the concept of massbalance in the sorption system (Sarkar et al. 2005).

The breakthrough curve of p-NP sorption on BFA andCZBFA is expressed in terms of C/C0 against total quantityof p-NP effluent solution, Ve, which passes through thecolumn (Fig. 4). The values of Cx, Cb, Vb (effluent volumecorresponding to Cb), Vx (volume of effluent correspondingto Cx) obtained from this graph were used to determineprimary sorption zone (PSZ=δ), total time taken for theestablishment of PSZ (tx), time required for initial forma-tion of PSZ (tf), time required for the downward movementof PSZ (tδ), fractional capacity (f), mass flow rate (Fm) andpercentage saturation at breakpoint (Table 7) with the aid ofreported equations (Sarkar et al. 2005; Gupta et al. 2000).The percentage saturation of the column at break point isobtained by Eq. 13. The obtained parameters give an ideaof the time required for breakthrough to occur and howmuch additional solution loaded per unit cross sectionalarea of the sorber would result in complete exhaustion ofthe sorbent column. If the technique is applied on largescale, these data can be useful for the design of fixed bedsorbers for the treatment of known p-NP concentrations.

% saturation ¼ Dþ dðf � 1ÞD

� 100 ð13Þ

The value of Vx−Vb was 285.35 and 305.73 mg cm−2 forBFA and CZBFA respectively, indicative of the additionalquantity of sorbate loaded per unit cross-sectional area thatwill result in complete exhaustion of sorbent capacity in agiven bed. The total time required for the primary sorptionzone, traveling to the end of the column and out of the bedwere 820 and 900 min on BFA and CZBFA, respectively.The time taken for initial formation of the primary sorptionzone (tf) was 2.33 and 2.5 h for BFA and CZBFA,respectively. The fractional capacity ‘f’ of the column inthe sorption zone at breakpoint to continue to remove solutefrom solution was 0.5 on both the sorbents. The percentage


saturation at breakpoint is 79.41 for BFA and 80.00 forCZBFA. From these observations a direct relationshipbetween the length of the sorption zone (δ) and percentagesaturation at break point can be developed. The smaller thelength of the sorption zone, the higher is the percentagesaturation in contrast with results obtained by Gupta et al.(2000). The breakthrough capacities and exhaustion capac-ities are higher than the batch sorption capacities. Thehigher capacities of column operations are due to contin-uously large concentration gradient at the interface zone asit passes through the column, while in the batch isothermtest the concentration gradient decreased with time (Guptaet al. 2004; Bhatnagar 2007). The degree of columnutilization (DCU) is 85.29% and 85.31% for BFA andCZBFA, respectively. Thus, these results show that thesesorbents can be effectively used in columns to remove p-NPfrom wastewaters.

3.9 Desorption studies

Desorption data show that HCl and SDS have a lowertendency to desorb p-NP, with 22.81%, 17.93%, and14.98% desorbed by 0.5 mol L−1 HCl and 22.98%,20.45%, and 17.78% by SDS from BFA, CZBFA, andFZBFA, respectively. The maximum desorption of 81.98%,79.98%, and 79.12% is obtained with 0.5 mol L−1 NaOHfrom BFA, CZBFA, and FZBFA, respectively, due toformation of highly water-soluble sodium phenolates whichare readily desorbed from the sorbents. Similar observationsare reported by Anirudhan et al. (2009) for desorption ofphenol, p-chlorophenol, and p-nitrophenol from activatedcarbons.

Desorption of p-NP from saturated fixed bed columnwas carried out with 0.5 mol L−1 NaOH (flow rate of1 mL min−1) at room temperature. For that, the exhaustedcolumn was washed with double-distilled water (25 mL) totake out unsorbed p-NP. The negligible amount of p-NPwas found in washing effluent. After washing, 0.5 mol L−1

NaOH was percolated through the column, collected in10 mL fractions from which desorbed amount of p-NP wasmeasured. The efficiency of solute recovery was calculatedfrom the breakthrough and recovery curves (Fig. 4). Fromthe desorption plots, it is clear that the first 10 mL aliquotelutes more than 50% of p-NP and the rest required furthertreatment of nine fraction of 10 mL which gives about 98%overall percentage recovery from BFA and CZBFA col-umns. These desorption studies indicates that about 120 mLof 0.5 mol L−1 NaOH is sufficient for almost completedesorption of p-NP from the sorbent column used.

3.10 Economic evaluation

Cost analysis is an important parameter in determining thecriteria for applicability of the sorbent and selection of thetreatment process in industrial use for environmentalprotection. The cost of the sorption process is mainlydependent on the cost of the sorbent used for the removal ofthe organic compound from wastewater. In developingcountries, low-cost materials and their economic feasibilityare more important compared to that of commercialactivated carbon (Ahmaruzzaman and Laxmi 2010). InIndia, the cost of cheapest variety commercial activatedcarbon is ≈US $2,000 per tons, generally used for effluenttreatment (Bhatnagar 2007). The bagasse fly ash is

Fig. 4 Breakthrough curves(black circles BFA and blackcircles CZBFA) and desorptionplots (white squares BFA andwhite circles CZBFA) of p-NP


available for US $25 per tones. The cost of synthesizedzeolite CZBFA and FZBA was estimated to be US $150–180 per tons including the cost of purchase, transport,chemicals, electrical energy, and labor required which islower compared to that of commercial activated carbonavailable in the market. Since the raw material of thesorbent is available easily, the cost of CBZFA and FZBFAis lesser as compared to activated carbons of cheapestvariety, and its utilization for the removal of p-NP is quitejustified for the treatment of phenolic wastewaters.

4 Conclusion

In this study, CZBFA and FZBFA prepared from BFA byhydrothermal and fusion methods were characterized byFTIR, PXRD, SEM, and XRF techniques and possessedimproved morphological and sorption properties. Batch andcolumn studies were performed for the removal of p-NPfrom simulated polluted water. The obtained thermodynam-ic data confirmed endothermic physisorption and kineticdata represented to fit well a pseudo-second-order kineticmodel than pseudo-first-order model. The uptake of p-NPcan be explained by external diffusion followed by intra-particle diffusion process. Batch experimental data fits theLangmuir isotherm, indicating the monolayer coverage ofp-NP molecules at the outer surface of the sorbents.CZBFA and FZBFA showed higher sorption capacity forp-NP as compared to sorption capacity of original BFA.The mass transfer and rate expression studies confirm thatthe rate of sorption is associated with mass transfer frombulk liquid to the particle external surface and intraparticlediffusion from the sorbent surfaces. The fine particle size ofFZBFA prevented the column study with this sorbent.Column studies of BFA and CZBFA revealed that thebreakthrough capacity was higher than the batch capacity.The column parameters obtained in this study can facilitatethe treatment of p-NP containing wastewater of knownconcentration by fixed bed column. The sorbed p-NP wasquantitatively desorbed with 0.5 mol L−1 NaOH. Thecomparative data showed that used materials, though notvery good as commercial activated carbon, has quite highsorption capacity for p-NP compared to the rest of thesorbents and may be used effectively for the removal of p-NP from aqueous p-NP-contaminated streams. The resultsdemonstrated that BFA, which has a very low economicvalue, may be used effectively for the synthesis of zeoliticmaterials and can be effectively used for the removal of p-NP from aqueous systems.

Acknowledgments The authors are grateful to the Special Assis-tance Programme meritorious fellowship, UGC, New Delhi, India.T

able

7Param

etersof

fixedbedsorber

Sorbent

C0

(mgcm

−3)

Cx

(mgcm

−3)

Cb

(mgcm

−3)

Vx

(mgcm

−2)

Vb

(mgcm

−2)

Vx−V

b

(mgcm

−2)

Fm

(mgcm

−2min

−1)

D (cm)

t x (min.)

t δ (min.)

t f (min.)

δf

% Saturation

Breakthroug

hcapacity

(mgg−

1)

Exh

austion

capacity

(mgg−

1)

DCU%

BFA

0.4

0.39

0.01

783

5.67

550.32

285.35

1.01

912

820

280

140

4.94

0.50

079

.41

105.63

123.84

85.29

CZBFA

0.4

0.39

0.01

991

7.19

611.46

305.73

1.01

912

900

300

150

4.80

0.50

080

.00

118.33

135.68

85.31


References

Ahmaruzzaman M, Laxmi GS (2010) Activated tea waste as apotential low-cost adsorbent for the removal of p-nitrophenolfrom wastewater. J Chem Eng Data 55:4614–4623

Ahmaruzzaman M, Sharma DK (2005) Adsorption of phenols fromwastewater. J Colloid Interface Sci 287:14–24

Anirudhan TS, Sreekumari SS, Brigle CD (2009) Removal of phenolsfrom water and petroleum industry refinery effluents by activatedcarbon obtained from coconut coir pith. Adsorption 15:439–451

Basu JK, Dutta S, Ghar RN (2001) Studies on adsorption of p-nitrophenol on charred saw-dust. Sep Purif Technol 21:227–235

Bhatnagar A (2007) Removal of bromophenol from water usingindustrial wastes as low cost adsorbents. J Hazard MaterB139:93–102

Crini G (2006) Non-conventional low-cost adsorbents for dyeremoval: a review. Bioresour Technol 97:1061–1085

Crini G, Morin N, Rouland JC, Janus L, Morcellet M, Bertini S (2002)Adsorption de beta-napthol sur des gels de cyclodextrine-carboxym-ethyl-cellulose reticules. Eur Polym J 38:1095–1103

Dabrowski A, Podkoscielny P, Hubicki M, Barczak M (2005)Adsorption of phenolic compounds by activated carbons acritical review. Chemosphere 58:1049–1070

Dubinin MM, Radushkevich LV (1947) The equation of thecharacteristic curve of the activated charcoal. Proc Acad SciUSSR Phys Chem Sec 55:331–337

Fierro V, Torné-Fernández V, Montané D, Celzard A (2008)Adsorption of phenol onto activated carbons having differenttextural and surface properties. Microporous Mesoporous Mater111:276–284

Freundlich HMF (1906) Over the adsorption in solution. J Phys Chem57:385–470

Gupta SK, Karim K (2003) Continuous biotransformation andremoval of nitrophenols under denitrifying conditions. WaterRes 37:2953–2959

Gupta VK, Sharma S, Yadav IS, Mohan D (1998) Utilization ofbagasse fly ash generated in the sugar industry for the removaland recovery of phenol and p-nitrophenol from wastewater. JChem Technol Biotechnol 71:180–186

Gupta VK, Srivastava SK, Tyagi R (2000) Design parameters for thetreatment of phenolic wastes by carbon columns (obtained fromfertilizer waste material). Water Res 34:1543–1550

Gupta VK, Ali I, Saini VK (2004) Removal of chlorophenols fromwastewater using red mud: an aluminum industry waste. EnvironSci Technol 38:4012–4018

Inada M, Tsujimoto H, Eguchi Y, Enomoto N, Hojo J (2005)Microwave-assisted zeolite syntheses from coal fly ash inhydrothermal process. Fuel 84:1482–1486

Joint committee on powder diffraction standards, JCPDS (1971),Index (Inorganic to the powder diffraction file), PublicationPDIS-211, Newton square, PA

Juang RS, Lin SH (2009) Adsorption of phenol and its derivativesfrom water using synthetic resins and low-cost natural adsorb-ents: a review. J Environ Manage 90:1336–1349

Langmuir I (1916) The constitution and fundamental properties ofsolids and liquids. J Am Chem Soc 38:2221–2295

Lin C-F, His H-C (1995) Resource recovery of waste fly ash: synthesis ofzeolite-like materials. Environ Sci Technol 29:1109–1117

Mall ID, Srivastava VC, Swamy MM, Prasad B, Mishra IM (2006)Adsorptive removal of phenol by bagasse fly ash and activatedcarbon: equilibrium, kinetic and thermodynamics. Colloid Surf272:89–104

Ozcan Safa A, Erdem B, Ozcan A (2005) Adsorption of acid blue 193from aqueous solutions onto BTMA-bentonite. Colloid Surf A266:73–81

Pan B, Du W, Zhang W, Zhang X, Zhang Q, Pan B, Lv L, Zhange Qu,Chen J (2007) Improved adsorption of 4-nitrophenol onto a novelhyper-cross-linked polymer. Environ Sci Technol 41:5057–5062

Pandit AB, Gogate PR (2004) A review of imperative technologies forwastewater treatment II: hybrid methods. Adv Environ Res8:553–597

Release and pollution prevention report (2000) US EnvironmentalProtection Agency; Washington, DC

Sarkar M, Acharya PK, Bhattacharya B (2003) Modelling theadsorption kinetics of some priority organic pollutants in waterfrom diffusion and activated energy parameters. J ColloidInterface Sci 266:28–32

Sarkar M, Acharya P, Bhattacharya B (2005) Removal characteristicsof some priority organic pollutants from water in a fixed bed flyash column. J Chem Technol Biotechnol 80:1349–1355

Sathishkumar M, Binupriya AR, Kavitha D, Selvakumar R, SheemaKK, Choi JG, Yun SE (2008) Organic micro-pollutant removal inliquid-phase using carbonized silk cotton hull. J Environ Sci 20(9):1046–1054

Schwarz JA, Noh JS (1990) Effect of HNO3 treatment on the surfaceacidity of activated carbons. Carbon 28:675–682

Shah BA, Shah AV, Singh RR, Patel NB (2009) Sorptive removal ofnickel onto weathered basaltic andesite products: kinetics andisotherms. J Environ Sci Health Part A 44:880–895

Singh BK, Nayak PS (2004) Sorption equilibrium studies of toxicnitro-substituted phenols on fly ash. Adsorpt Sci Technol22:295–310

Toxic Release Inventory (1992) Toxicological profile for nitrophenols:2-nitrophenols and 4-nitrophenols. Office of toxic substances,Washington

Treacy MMJ, Higgins JB (2001) Collection of simulated XRD powderpatterns for zeolites. Fourth Revised Edition Published on behalfof the Structure Commission of the ‘International ZeoliteAssociation’, Elsevier, Amsterdam, London, New York, Oxford,Paris, Shannon, Tokyo

Tutem E, Apak R, Unal CF (1998) Adsorptive removal of chlor-ophenols from water by bituminous shale. Water Res 32:2315–2324

Vucinic D, Miljanovic I, Rosic A, Lazic P (2003) Effect of Na2O/SiO2

mole ratio on the crystal type of zeolite synthesized from coal flyash. J Serb Chem Soc 68:471–478

Wang S, Wu H (2006) Review environmental-benign utilisation of flyash as low-cost adsorbents. J Hazard Mater B 136:482–501

Wang Y, Guo Y, Yang Z, Cai H, Xavier Q (2003) Synthesis of zeolitesusing fly ash and their application in removing heavy metalsfrom water. Sci China Ser: D 46:967–976

Wang S, Soudi M, Li L, Zhu ZH (2006) Coal ash conversion intoeffective adsorbents for removal of heavy metals and dyes fromwastewater. J Hazard Mater B 133:243–251

Wang S, Li L, Zhu ZH (2007) Solid-state conversion of fly ash toeffective adsorbents for Cu removal from wastewater. J HazardMater B 139(2):254–259

Weber WJJr, Asce AM, Morris JC (1963) Kinetics of adsorption oncarbon from solution. J Sanit Eng Di VAm Soc Ci V Eng 89:31–59

Zheng Z, Tang D, Lin K, Luan J, Zhang J (2007) Adsorption of p-nitrophenol from aqueous solutions onto activated carbon fiber. JHazard Mater 143:49–56

Zümriye A (2005) Application of biosorption for the removal oforganic pollutants: a review. Process Biochem 40:997–1026


zeolitic bagasse fly ash as a low-cost sorbent for the sequestration of p-nitrophenol: equilibrium,...

Documents