Biosorption of basic violet 10 onto activated Gossypium hirsutum seeds:Batch and fixed-bed column studies

N.Sivarajasekar,R.Baskar

Department of Chemical Engineering,Kongu Engineering College,Perundurai,Erode 638052,Tamil Nadu,India

Keywords:Gossypium hirsutum seed Isotherm Kinetic Fixed-bed Mass transfer

ABSTRACT Sulphuric acid activated immature Gossypium hirsutum seed(AIGHS)was prepared to biosorbe basic violet 10(BV10)from aqueous solutions.Methylene blue number,iodine number and Brunauer-Emmett-Teller surface analysis indicated that the AIGHS were hetero-porous.Boehm titrations and Fourier-transform infrared spectra demonstrated the chemical heterogeneity of the AIGHS surface.Batch biosorption studies were used to examine the effects of process parameters in the following range:pH 2-12,temperature 293-313 K,contact time 1-5 h and initial concentration 200-600 mg·L-1.The matching of equilibrium data with the Langmuir-Freundlich form of isotherms indicated that the BV10 was adsorbed via chemisorption and pore diffusion.Kinetic investigation indicated multiple order chemisorption through an Avrami kinetic model.Film diffusion controlled the rate ofBV10 biosorption onto AIGHS.The spontaneous and endothermic nature ofsorption was corroborated by thermodynamic study.Continuous biosorption experiments were performed using a fixed-bed column and the in fluence of operating parameters was explored for different ranges of initial concentration 100-300 mg·L-1,bed height 5-10 cm,and flow rate 2.5-4.5 ml·min-1.A dose response model accurately described the fixed-bed biosorption data.An external mass transfer correlation was formulated explaining BV10-AIGHS sorption.

1.Introduction

Treatment of colour bearing ef fluents is a growing need at the present time.Chemical industries are extensively using more than 10000 synthetic dyes in their processes.Among the variety of dyes used,basic violet 10 was widely employed for acrylic,wool,nylon and silk dyeing because of its favourable characteristics of bright colour,high solubility in water and low-energy consumption[1].This cationic dye decomposes into carcinogenic aromatic amines under anaerobic conditions;therefore discharge ofef fluents bearing this dye into water bodies can cause harmful effects such as allergic dermatitis,skin irritation,mutations and cancer[2-5].Therefore,strict discharge standards are being enforced to controlthe release ofcoloured wastewaterinto the environment.Hence,there is an urgent need for development of effective methods for removal of dyes.Different physicochemical methods including coagulation[6],ozonation[7],chemical oxidation[8],solvent extraction[9],ion exchange[10],photo-catalytic degradation[11],and adsorption[12-14]have been tried by many researchers for the treatment of dye contaminated water.Among the aforementioned methods,adsorption is an effective and eco-friendly process due to its simple design,easy operation and its ef ficacy to remove a wide range of compounds[2,4,5,15,16].Activated carbon is considered to be an effective adsorbent for removal of dyes from water.However,due to its cost,unconventional low-cost bio based adsorbents have attracted the attention of several investigators in recent years[3,4,15].Our search for an ef ficient new biomass to treat BV10 drew attention to immature cotton seeds rejected by seed manufacturers.According to the USA agricultural data base,85%of the world cotton production is shared by 10 countries,namely,China,USA,India,Pakistan,Uzbekistan,Brazil,Turkey,Australia,Greece and Syria.Among these,India is the third largestcountry which produced 5787×103metric tonnes ofun-ginned cotton in the 2012/2013 market year[17].Immature cotton seeds un fit for germination and having poor oil content,are usually discarded as waste and only needs activation to transform them into cheap and high quality adsorbent.Although many agriculturaland waste materialswere used as biosorbents forthe removalofcolour,biosorbentderived fromimmature Gossypium hirsutum seed was never reported to the best of our knowledge.Design of the adsorption process requires insight into adsorption equilibrium,kinetics,rate limiting steps and the thermodynamics of adsorption which can be readily obtained from batch adsorption experiments.However,due to the limitations of batch processing,requiring treatment of small quantities of ef fluent and inconvenience for use on an industrial scale,continuous flow fixed-bed columns are often employed.The design of fixed-bed adsorption processes requires indepth understanding ofmass transferand breakthrough curve dynamics in light of process modelling to facilitate scaling up of the processes.

The objective ofthis work was to investigate the suitability of immature Gossypium hirsutum seeds as a precursor for biosorbent and to examine the seeds'ability to take up of basic violet 10 from aqueous solution via batch and continuous operation.Batch biosorption studies were carried out to analyse the effect of process parameters,isotherms and kinetics.The rate limiting step,thermodynamics and possible reaction mechanism of the biosorption were also examined.Fixed-bed column experiments were performed to investigate the effect of column parameters.The dynamic response of the fixed-bed column was explored via mathematical modelling and mass transfer correlations.

2.Materials and Methods

2.1.Chemicals

Basic violet 10(also called Rhodamine B,molecular mass=479.02,chemicalformula C28H31ClN2O3andλmax=555 nm)was obtained from S.D.Fine Chemicals Ltd,Mumbai.All other chemicals were obtained from Merck India Ltd,Mumbai.

2.2.Preparation of dye solution

Stock solutions were prepared by dissolving 1 g of BV10 in 1 L of deionised water.All working solutions were prepared by diluting the stock solution with deionised waterto the desired concentration.Before addition ofbiosorbent,the initialpHofthe working solution was adjusted to the desired experimental conditions by mixing appropriate quantities of 0.1 mol·L-1HCl or 0.1 mol·L-1NaOH solutions.The concentration of BV10 in the sample was analysed using a double beam UV-Vis spectrophotometer(ELICO-SL244,India)at a maximum wavelength of 555 nm.

2.3.Activated biomass preparation

Immature Gossypium hirsutum seeds were obtained from seed producers near Attur,Tamil Nadu,India.The seeds were washed thoroughly with deionised water twice to remove impurities and dried at 313 K in a temperature controlled oven for 3 days.The dried seeds were then soaked in concentrated sulphuric acid(98 wt%)with a mass ratio of 1:4.The acid soaked seeds were stirred periodically and kept for 12 h to facilitate effective activation.The resultant slurry was carefully washed with deionised water and 0.1 mol·L-1sodium bicarbonate solution was used to remove traces of acid.The final material,activated immature Gossypium hirsutum seed(AIGHS)was dried at 313 K and finely ground to the size of 0.088 cm,and then stored in anairtightcontainer for biosorption experiments.

2.4.Characterisation of AIGHS

Iodine number and methylene blue number were calculated based on the ASTM 4607-86 standards at 298 K.Speci fic surface area,pore volume and pore diameter were measured based on the adsorptiondesorption isotherm of nitrogen at 77 K using a surface analyser(Micromeritics ASAP 2020)and the BJH method.Surface morphology was examined by using a scanning electron microscope(SEM:JSM-6390LV-JEOL Ltd.,Japan).Surface functional groups were determined based on a Boehm titration method.A solid addition method was employed to determine the zero surface charge(pHpzc).Fourier transmission infra-red(FTIR)analysis was performed using a Thermo Nicolet,Avatar 370 FTIR spectrometer over a spectral range of 4000-400 cm-1at a resolution of 4 cm-1.

2.5.Batch biosorption

Batch biosorption experiments were conducted for the selected parameter ranges such as pH 2-12,temperature 293-313 K,initial concentration 200-600 mg·L-1and contact time 1-5 h.A known amount ofbiosorbentwas added to 200 mlofBV10 solution in Erlenmeyer flasks and agitated at 100 r·min-1by a thermo-regulated shaker.Samples were collected at predetermined time intervals and were centrifuged and analysed for the residual dye concentration using a dual beam UV-Vis spectrophotometer.The percentage of dye removal(R/%)and the dye biosorption capacity(qt/mg·g-1)were calculated using the following equation:

where Ci(mg·L-1)is the initialconcentration ofthe dye,C0(mg·L-1)is the final concentration of the dye,Ct(mg·L-1)is the concentration of BV10 at any time t,V(L)is the volume of the dye solution,and M(g)is the mass of the AIGHS.Desorption studies were carried out using different solvents such as hot deionised water,2 mol·L-1sulphuric acid and 2 mol·L-1acetic acid.

2.6.Column biosorption

The fixed-bed columns were made ofPVC tubes with 1.5 cm internal diameter and with different heights(5-10 cm).AIGHS was loaded into columns containing glass beads and sieve plates fixed at the top and bottom to enhance uniform inlet flow.Initially,the columns were flushed out with deionised water for 24 h to remove air bubbles and then the bed porosity was measured using tracer experiments.The BV10 solution was pumped into the columns in a downward- flow mode.All of the column experiments were carried out in the ranges of 100-300 mg·L-1,2.5-4.5 ml·min-1,pH 6.8 ± 0.5 and room temperature(301±2)K.Following Eqs.(3)-(5)were utilised to evaluate the fixed-bed column parameters.

where Zm(cm)is the length of the mass transfer zone,mad(mg)is the amount of BV10 sent into the column,Rc(%)is the BV10 removal percentage in the fixed-bed,tb(min)is the breakthrough time,te(min)is the bed exhaustion time,H(cm)is the height of the column and Q is the volumetric flow rate(ml·min-1).

2.7.Nonlinear regression analysis

Nonlinear regression analysis was preferred wherever required to determine the parameters of the equations due to its versatility and its accuracy for equations containing more than two parameters[17,18].The error functions such as ERRSQ,HYBRID,MPSD,ARE,and EABS were selected for nonlinear regression analysis.Statistical-comparison values,such as the coef ficient of determination(R2)and root mean square error(RMSE),were utilised to gauge the goodness of the fit.For a meaningful comparison between parameter sets produced by these five differenterrors,a ‘Sum ofthe Normalised Errors’(SNE)procedure was adopted[17-19].High values of R2and small values of RMSE are the criteria to choose the best fit[17,18].

3.Results and Discussions

3.1.Characterisation of AIGHS

Iodine number and methylene blue are keys to recognise the development of micropores(1.0-2.8 nm)and mesopores(approximately 1.5 nm)on adsorbent surfaces,respectively.Iodine number and methylene blue number of AIGHS were measured to be 510 mg·g-1and 42 mg·g-1,respectively,indicating excellent development of different sizes of pores on the AIGHS surfaces.

The nitrogen adsorption-desorption curve of AIGHS is shown in Fig.1(a).The occurrence of a Type I isotherm at low relative pressure(P/Po＜0.4)and a combination of type IV isotherm with an open knee hysteresis at higher relative pressure(P/Po＞0.4)indicates structural heterogeneity via development of slit-shaped micropores[19]and a signi ficant number of mesopores[20].SEM micrographs[Fig.1(b)]and BET surface area analysis further support these observations.The higher the pore volume,along with the combination of a variety of pores,the higher is the sorption capacity[21].The AIGHS contains with both meso-pores(pore volume:0.302 cm3·g-1)and micropores(pore volume:0.111 cm3·g-1)with a BET surface area of 495.96 m2·g-1.

In addition to the porosity,the reactive functionalgroups on the surface strongly in fluence biosorption behaviour of the AIGHS.Boehm titration provides evidence for the existence of surface functional groups such as carboxylic,lactonic,phenolic and carbonyl contributing to surface acidity[22].The surface of the AIGHS was found to be chemically heterogeneous,and the basicity(1.24 mmol·g-1)was less compared to the surface acidity(5.37 mmol·g-1).The point of zero charge(pHpzc)at which the pH of AIGHS surface has net electrical neutrality was evaluated to be 6.4 from Fig.1(c).The surface pHpzcvalue agreed with Boehm titration results and con firmed the dominance of acidic groups on the surface of the AIGHS.

The surface functional groups of AIGHS were identi fied using FTIR[Fig.1(d)].The peaks at 3419 cm-1can be assigned to O-H stretching mode of hydroxyl groups and adsorbed water.The position and asymmetry ofthis band atlowerwave numbers indicated a presence ofstrong hydrogen bonds relevant to carboxyls,phenols or alcohols(C-O,C=O and O-H).The bands at 2921 cm-1and 2852 cm-1can be attributed to aliphatic(C-H)asymmetric and symmetric stretch of CH2.The peaks approximately 1704 cm-1are characteristic of stretching vibrations of C=O in ketones,aldehyde,lactone and carboxyl groups.A very weak peak approximately 1625 cm-1and a shoulder approximately 1452 cm-1can be attributed to C=C stretching of aromatic rings and C=Ostretching ofethers,respectively.A shoulder at1380 cm-1,smaller peaks at 1162 cm-1and 1029 cm-1can be assigned to stretching vibrations of S=O in sulfoxides and sulfonates.The band observed at 793 cm-1is attributed to stretching vibrations ofS-O.The shoulders visible approximately 714 cm-1and 608 cm-1may be due to the C-Hbend of alkenes and acetylenic groups.In general,FTIR spectra helped identify the various surface acidic and basic functional groups that resulted on the AIGHS from surface oxidation.

Fig.1.(a)Adsorption and desorption of N2 onto AIGHS;(b)SEM micrographs of AIGHS;(c)Determination of pHzpc for AIGHS surface;(d)FTIR spectra of AIGHS.

3.2.Batch biosorption studies

3.2.1.Effect of pH

Fig.2(a)presents the effect of pH on the biosorption of BV10 at a temperature 313 K,initial concentration of 150 mg·L-1and contact time of3 h.The degree ofdye sorption onto AIGHS showed a marked increase from 82.89%to 100%as the pH changed from 2 to 12.The maximum adsorption was found to be 100%at pH 12.A drastic increase noted at pH greater than pHpzcwas due to the favourability to cationic adsorption through an increase in negative charge density,reduction in H+ions concentration on sorption sites and electrostatic attraction between solute and biosorbent[23].Whereas,at pH lower than pHpzc,high electrostatic repulsion between H+and BV10 cations resulted in the relatively poor biosorption[24-26].Cationic BV10 molecules and the surface active sites can dissociate into water as follows:

Dissociation of BV10 molecule:

At pH＜pHpzc,BV10 cations reacted with the reduced OH functional group that resulted in biosorption of BV10 onto AIGHS.Whereas at pH＞pHpzc,BV10 cations reacted with OH,COOHand SO3Hsurface functional groups.

Fig.2.(a)Effectofsolution pH;(b)effectofsolution temperature;(c)effectof contacttime and initialBV10 concentration;(d)regeneration cycles of BV10 loaded AIGHS using acetic acid.

3.2.2.Effect of temperature

Fig.2(b)shows the effectofsolution temperature on the biosorption ofBV10 ata pH 12,initialconcentration of150 mg·L-1and contacttime of 3 h.An increasing trend in percentage of dye removal from 89.21%to 100%was noticed while changing the solution temperature from 293 K to 313 K.This indicated that the biosorption of BV10 was endothermic.i.e.,the rise in solution temperature promoted the surface activity and kinetic energy of BV10 which in turn increased the film diffusivity and pore diffusivity of BV10[4,13-16].

3.2.3.Effect of initial concentration and contact time

The sorption of BV10 onto AIGHS was observed as a function of time until the amount of BV10 adsorbed became constant.Fig.2(c)elucidates the effect of initial concentration and contact time on the biosorption of BV10 at a pH 12 and a temperature 313 K.Spontaneous sorption was observed initially(within 30 min)followed by a gradual increase(until 180 min),and after that there was only a slight difference in the percentage BV10 removal.Hence the equilibrium time was considered to be 180 min for all of the experiments,independent of the initial BV10 concentrations[25,26].The percentage BV10 removal onto AIGHS decreased with the increase of initial BV10 concentrations due to the unavailability of active sites and pores for BV10 molecules on AIGHS particles[4,13-16].

3.2.4.Isotherms

Isotherms provide better insight into biosorbent-dye binding and the nature of the sorption system[17,18].Batch equilibrium data obtained for five different initial concentrations(50-250 mg·L-1)at pH 12,temperature 313 K and contact time 3 h were fitted to various isotherms such as:Langmuir,Freundlich,Redlich Peterson,Sips,Toth,Hill,Khan,Fritz-Schluender,and Marczewski-Jaroniec.The nonlinear minimum SNE,error function selected,optimum parameter set for each isotherm model,R2and RMSE values are displayed in Table 1.The goodness of the models used to explain the equilibrium data was in the following order:Marczewski-Jaroniec＞Redlich-Peterson＞Fritz-Schluender＞Khan＞Sips＞Hill＞Langmuir＞Toth＞Freundlich.

Good fitting to the Marczewski-Jaroniec isotherm indicated that the BV10 biosorption on AIGHS was homogeneous with a distribution of adsorption energies in active sites[15-18].Correlation to Redlich-Peterson and Fritz-Schluender elucidated the combined nature of physisorption and chemisorption.The values of the Redlich-Peterson model exponent βRP(0.80＜1),Sips model exponent nS(1.46＞1)and Khan model exponent aK(0.76＜1)indicate that the biosorption of BV10 happened via surface functional groups and into pores of AIGHS[15-18].The maximum adsorption capacity of BV10 onto AIGHS is compared in Table 2.

Table 2 Maximum adsorption capacity values of BV10 for different adsorbents

3.2.5.Kinetics

In order to obtain clear insight into BV10 biosorption onto AIGHS,various kinetic equations including Ho,Sobkowsk-Czerwi,Blanchard,Elovich,and Avrami were employed for testing the batch kinetic data.Batch experiments were carried out for differential concentrations 50-250 mg·L-1with pH 12,temperature 313 K and contact time 3 h.The nonlinear minimum SNE that was determined,the error function selected,the calculated optimum parameter set,R2and RMSE values for each kinetic model are presented in Table 3.

Among all of the kinetic models,the Avrami model provided a proper representation of the experimental data.Equilibrium capacities(38.3883-97.6830 mg·g-1)of the Avrami model showed an increasing trend during an increase in initial concentration but the model found an opposite trend in the case of kinetic rate constants(0.2821-0.1891 min-1).The Avrami kinetic model indicated that the mechanismof biosorption followed multiple kinetic orders which changed during contactofthe BV10 with AIGHS.Additionally, fitto Sobkowsk-Czerwiindicated that higher initial dye uptake at small time duration(＜30 min)was due to chemisorption and atlater times was by pore diffusion.Therefore,kinetic study also indicates biosorption of BV10 onto AIGHS was a combination of chemisorption and physisorption[15-18].

Table 1 Statistical values and optimum parameters for isotherms

Table 3 Statistical values and optimum parameters for kinetic models at different initial concentrations

3.2.6.Rate determination step

The rate-limiting step depends on the physical and chemical characteristics of the biosorbent as well as on the mass transfer process.For a solid-liquid sorption process,solute transfer is characterised by external mass transfer or intraparticle diffusion,or both.The rate-controlling parameter is distributed between intraparticle and film diffusion mechanisms.Irrespective ofthe case externaldiffusion willbe involved in the sorption process.The most common technique used for identifying the rate determining step is described by Weber and Morris[15,33]via fitting a plot for the following Equation:

where Kidis the intraparticle diffusion rate constant(mg·g-1·min-0.5).The two stages in the Weberand Morris plot[Fig.2(d)]corresponded to surface sorption and intraparticle diffusion,respectively.Therefore,to understand the diffusion mechanism quantitatively,it is a prerequisite to calculate their coef ficients.By assuming the AIGHS particle to be a sphere of radius ‘r’and that diffusion follows the Fick's law,according to Reichenberg and Helfferich[15,34]:

Eq.(7)predicts the film diffusion coef ficient Df(cm2·s-1)atsmaller times and pore diffusion coef ficient Dp(cm2·s-1)at larger times.From Fig.2(e),the mean Dfand Dpvalues calculated over the initial concentration range(200-600 mg·L-1)were 3.97 × 10-8cm2·s-1and 7.61 × 10-6cm2·s-1,respectively.The mean film diffusion coef ficient was on the order of 10-8cm2·s-1and con firmed that the biosorption was controlled by film diffusion in the concentration range that was investigated[15,35].

3.2.7.Thermodynamics

Thermodynamic parameters including the changes in Gibbs free energy ΔG0(kJ·mol-1),Heat of adsorption or enthalpy ΔH0(kJ·mol-1),entropy ΔS0(kJ·mol-1·K-1)and apparent isosteric heat of adsorption ΔHis(kJ·kg-1)were calculated from Eqs.(8)-(10)to access the spontaneity of the biosorption process.

where R is the gas constant(8.314 J·mol-1·K-1)and T is the absolute temperature(K).Biosorption of BV10 by AIGHS was spontaneous and realistic with negative values of mean ΔG0(-95,837.17 J·mol-1).The positive larger value of mean ΔH0(1358.91 J·kg-1)indicates that the overall biosorption process was endothermic and chemisorptive in nature.A positive mean ΔS0(323.45 J·mol-1·K-1)value indicated increased disorder and some structural changes at the AIGHS surface as well as of the BV10 during biosorption.The positive value of the mean ΔHis(10,701 J·kg-1)further con firmed the endothermic nature[15,17,18].

3.2.8.Desorption study

A 38%desorption of BV10 with hot deionised water evidenced physisorption to some degree involving weak bonds of attachment.The possibility of biosorption due to irreversible ion exchange was less likely because desorption ofBV10 with 2 mol·L-1sulphuric acid wasinsigni ficant(9%).Nonetheless,a remarkable percentage of elution was noted with acetic acid(89%),and this suggested that chemisorption was responsible for biosorption of BV10 over and above physisorption[17,27,35].Further from Fig.2(d),it can be seen that the BV10 adsorbed AIGHS could be regenerated effectively twice by acetic acid.

3.3.Column biosorption studies

3.3.1.Effect of flow rate

The effects of flow rate on column performance were evaluated at three different flow rates(2.5,3.0 and 3.5 ml·min-1)using a 5 cm deep bed at 25 mg·L-1[Fig.3(a)].While the flow rate changed from 2.5 ml·min-1to 3.5 ml·min-1,the percentage of BV10 removal(35.1%to 33.5%),breakthrough time(65 min to 15 min)and bed exhaustion time(230 min to 150 min)values decreased.i.e.,at high flow rates,BV10 percolated and reacted with the functional groups of AIGHS for insuf ficient time and this resulted in improper utilisation of the column.

3.3.2.Effect of in fluent concentration

Fig.3(b)shows the effect of the initial concentration of BV10 studied at 100-200 mg·L-1,2.5 ml·min-1and 5 cm bed height.The breakthrough time(65 min to 22 min)and bed exhaustion time(230 min to 130 min)values decreased with the increase of initial BV10 concentration.In contrast,the percentage of BV10 removal(35.1%to 46.4%)and mass transfer zone(3.59 cm to 4.15 cm)values increased with increasing initial BV10 concentration.As was observed,higher concentration provided a stronger concentration gradient that caused a strong driving force for mass transport and higher uptake.

3.3.3.Effect of bed height

The biosorption performance was tested at various bed heights(5-10 cm)at 2.5 ml·min-1and 200 mg·L-1.Fig.3(c)shows the breakthrough pro files of BV10 biosorption at various bed heights.An increase in the bed height(5 cm to 7.5 cm)increased the percentage BV10 removal(35.1%to 62.7%),breakthrough time(65 min to 205 min),bed exhaustion time(230 min to 650 min)and mass transfer zone(3.59-6.85 cm)values.The increase in bed height increased the mass of AIGHS and surface area available for BV10 binding,which led to better performance of the column.

Fig.3.(a)Effect of in fluent flowrate Ct/Ci;(b)effect of initial BV10 concentration;(c)effect of bed height;(d)effect of flow rate over mass transfer resistance and axial dispersion.

3.3.4.Modelling

The practical applicability of a column biosorption process requires the analysis of breakthrough curves with various mathematical models[17].Hence the dynamic response of the fixed-bed column was veri fied with models such as Thomas,Adams-Bohart,Yoon-Nelson,Bed Depth Service Time(BDST),and dose response.The minimum SNE,error functions selected,optimum parameter set,R2and RMSE values for each model at different bed heights(5-10 cm)are presented in Table 4.

As a whole,all of the selected models could be employed to predict the theoretical breakthrough curves.By comparing the RMSE values of all of the models,the smaller values of RMSE for the dose response model demonstrated its accuracy over the others to predict the breakthrough curves.Same kind of results was obtained while adsorbing methylene blue onto peanut husk[36].

3.3.5.Mass transfer in fixed-bed column

Insight into the mass transfer can be used to enhance the design of the fixed-bed column adsorber to meet speci fied requirements.The film mass transfer coef ficient(kf,cm·min-1)quanti fies the solute transfer between solid and liquid phases and it can be estimated from different Nshcorrelations relating hydrodynamics and structural properties.The Chilton-Colburn methodology[17,37]which successfully extends the Reynolds'analogy to liquid phase adsorption was adopted to formulate an empirical correlation for the BV10-IGHS fixed-bed sorption system.The Chilton-Coburn factor(JD)equation is

where NRe(=Qdpρ/μ)is the Reynolds number,ρ (g·cm-3)is the density and μ (g·cm-1·min-1)is the viscosity of water,respectively.Non-linear regression analysis based on the minimum SNE procedure,was utilised to correlate JDwith NRefor the BV10-IGHS fixed-bed sorption system.

where NScis the Schmidt number:is the molecular diffusivity.Table 5 presents the various empirical correlations and mean values of external mass transfer coef ficient calculated for the BV10-IGHS fixed-bed sorption system with a 5 cm bed height over the flow rate(2.5-5.5 ml·min-1).Based on comparison of the mean kfvalues,it appears that the existing correlations were not compatible with the current fixed-bed system.The resistance to mass transfer(1/kf,min·cm-1)was estimated and the axial dispersion coef ficientDax(cm2·min-1)was evaluated from the correlation proposed by Ruthvan[38]

Table 5Empirical mass transfer correlations for BV10 adsorption by AIGHS

Table 4 Parameters of various packed-bed models at different bed heights

where v(cm·min-1)is the interstitial velocity of the aqueous phase.Fig.3(d)shows the variation of the mass transfer resistance and Daxvalues against flow rate(25-55 ml·min-1)in a 5 cm bed.As observed,while increasing flow rate,the resistance to mass transfer values decreased and Daxvalues increased due to better mass transfer at the liquid-solid interface.In another view,a higher flow rate forced more solute to be exposed to the active sites ofthe AIGHSin less time by overcoming boundary effects,which favoured the mass flux as well as diffusion in the axial direction[17].

4.Conclusions

The possibility of using immature cotton seeds as a precursor for the preparation of biosorbent was investigated.The surface morphology,surface chemistry,physical and chemical characteristics of AIGHS were characterised.Biosorption equilibrium data were in good agreement with Redlich-Peterson and Fritz-Schluender Isotherms.The Avrami second-order model described the batch biosorption kinetic data well.Film diffusion was the rate determining step for BV10 biosorption.The sorption thermodynamics illustrated the endothermic and spontaneous nature ofbiosorption ofBV10 by AIGHS.The BV10 was biosorbed onto AIGHS initially via multiple order chemisorption followed by slow pore diffusion.The fixed-bed column biosorption data were successfully modelled by a dose-response model.An empirical external mass transfer correlation was formulated for the current BV10-AIGHS sorption system.

Nomenclature