Film pore diffusion modeling for sorption of azo dye on to exfoliated graphitic nanoplatelets
Indian Journal of Chemical TechnologyVol. 20, January 2013, pp. 7-14 Film-pore diffusion modeling for sorption of azo dye on to exfoliated graphitic nanoplatelets T Maiyalagan1 & S Karthikeyan2,* 1 School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore 639 798 2 Department of Chemistry, Chikkanna Government Arts College, Tirupur 641 602, India Received 8 September 2011; accepted 8 May 2012 Exfoliated graphitic nanoplatelets (xGnPs) have been utilized as a potential adsorbent for toxic textile dye Acid Orange 7 (acid dye). The effects of major variables governing the efficiency of the process, such as temperature, initial dye concentration and pH are studied. The kinetic measurements have been used for determining the specific rate constant, confirming the applicability of pseudo first-order rate expression. Plausible mechanism of ongoing adsorption process involved is obtained by carrying out kinetic measurements. To identify whether the ongoing process is particle diffusion or film diffusion, the treatments given by Boyd and Reichenberg have been employed. The influence of different factors on the adsorption of Acid Orange 7 from solution is explained in terms of electrostatic interaction by considering the dye species and the surface character of the xGnPs. The developed system for the removal of acid dye is found to be very useful, economic, rapid and reproducible. Keywords: Acid Orange 7, Adsorption, Film diffusion, Graphitic Nanoplatelets, Particle diffusionThe adverse effects of discharge of organic pollutants of catalytic degradation and hypo chloride treatment ofdyeing industry waste on health have been already dye waste effluents2,3. Among these approaches,proved. Textile effluents are known toxicants, which adsorption is regarded as an easy and economicinflict acute disorders in aquatic organisms. Uptake of process. This is attributed to its easy availability,textile effluents through food chain in aquatic organisms simplicity of design, ease of operation, variousmay cause various physiological disorders like materials, such as commercial activated carbon, naturalhypertension, sporadic fever, renal damage and cramps1. materials, bio adsorbents and wastes from agriculture,The release of colored waste water into the eco-system have been used for such processes4.is a dramatic source of the aesthetic pollution, The rapid development in nanotechnology shedseutrophication and perturbation in aquatic life. light on the waste water treatment. Nano materials Brightly coloured and water soluble acid dyes, being have been studied for the adsorptions of metalsodium salts of organic sulphonic acids, are composed ions5, dyes6, and antibiotics7. Exfoliated graphiticof ionisable anionic groups such as sulphonates, nanoplatelets (xGnPs) and graphite nano sheets8 havecarboxylates or sulphates. They have direct affinity for been successfully utilized as sorbents to extract oils9polyamide and protein in an acidic bath and hence are and dyes10 from their aqueous solutions. In the presentcommonly used for dying polyamide, as well as nylon, work, exfoliated graphitic nanoplatelets (xGnPs)silk, wool and modified acrylics; also used to some have been used as an adsorbent for Acid Orangeextent for paper , leather and cosmetics. Acid dyes 7 (AO7) removals and the adsorption capacity ofwith higher molecular weight are one of the most xGnPs is regulated by many influencing factors,problematic groups of dyes which tend to pass through such as temperature, pH variations and initialconventional treatment system unaffected. Various dye concentrations.physical and chemical methods of treatment ofindustrial waste water have been suggested. These Experimental Procedureinclude adsorption method, coagulation process, photo xGnPs (with average diameter of 15 µm and average length of <0.01 µm) were procured from——————*Corresponding author. xG Sciences Inc., USA. Detailed information onE-mail: firstname.lastname@example.org fabrication, geometrical and surface characteristics
8 INDIAN J. CHEM TECHNOL., JANUARY 2013of this material is already available11. The textile C.I. Number : 15510dye (AO7) was purchased from Sigma - Aldrich Natural pH : 6.1(Germany), and characterization of the dye are Molecular formula : C16H11N2NaO4Ssummarized in Table 1. All the chemicals used were Molecular weight, g mol-1 : 350.33obtained as research grade chemicals and used pKa : 8.86without purification. Molecular volume, Å3 molecule-1 : 231.95Characterization Molecular dimension, nm : 1.24×0.68×0.22 Morphological structure of as-received xGnPs was Batch adsorption studies were carried out incharacterized by the scanning electron microscopy. 250 mL tight lid glass bottle (Borosil R). StandardThe sample was directly coated on the conductive stock solution (1000 mg/L) containing Acid Orange 7surface and SEM images were obtained with was prepared by dissolving appropriate amount of ita field-emission scanning electron microscope in water. 50 mg of adsorbent was added to 100 mL of(FESEM, JEOL JSM-6700F). BET measurements aqueous dye solution, initial concentration of AO7were performed by using ASAP 2020 volumetric ranging from 20 mg/L to 60 mg/L. The contents of theadsorption analyzer (Micrometrics, USA). The flasks were agitated by placing them in temperaturesurface functional groups on the adsorbents were controlled orbital shaker. The mixture was withdrawnquantitatively measured by Boehm’s titration at specified intervals then centrifuged using electricalmethod12. The Boehm titration is based on the centrifuge (universal make) at 3000 rpm for 10 minprinciple that oxygen groups on graphite surface have and un adsorbed supernatant liquid was analyzeddifferent acidifies being neutralized by bases of for residual dye concentration using Elico make Biodifferent strengths. In our procedure, 20 mg of xGnPs UV-Visible spectrometer (BL-198) at a wave length(as-received) were stirred in 10 mL of 0.05 M base of 484 nm. All the experiments were conducted insolution (NaOH, Na2CO3, and NaHCO3), aqueous duplicate and mean of the two values were taken forsolution under Ar for 48 h, (in order to equilibrate calculation. Maximum deviation is 4%. The amountwith the NaHCO3 solution). The mixtures were of AO7 adsorbed in mg/L at time t was computed byfiltered (on 0.20 µm pore size membrane filters), using the following equation:10 mL volume from each mixture being further C0 − Cttitrated with 0.05 M hydrochloric acid. Three samples qt = ×V … (1) msof each base solution were titrated; a blank samplewithout xGnPs is being titrated with the same where C0 and Ct are the AO7 concentration in mg/Lprocedure. NaOH solution neutralizes all acidic sites initially and at a given time t respectively; V, the(carboxyl, lactonic and phenols) from the surface of volume of the AO7 solutions in mL; and ms, thexGnPs; NaHCO3 neutralizes only carboxyl groups, weight of the xGnPs. The removed AO7 (%) inNa2CO3 reacts with carboxyl, lactonic groups. solution was calculated using the following equation:The quantity of the possible surface groups is C0 − Ctestimated through the difference between the % Removal = ×100 … (2)calculated amount of surface functionality. The pH C0of the point of zero charge (pHpzc) was determined Adsorption dynamics and equilibrium studiesusing the pH drift method13. The Acid Orange 7 dye The study of adsorption dynamics describes thewas used without purification. The characteristics solute uptake rate, and evidently this rate controls theof dye are shown below: residence time of adsorbate uptake at the solid- Table 1 –– Kinetic parameters for the adsorption of AO7 onto xGnPs Concentration Pseudo first-order values Elovich values Pseudo second-order values mg/L -2 kLager × 10 R 2 α Β R 2 qe k2 × 10-2 h R2 min-1 mg/g min g/mg g/mg min 20 1.7306 0.986 0.869 0.606 0.826 18.712 7.458 0.804 0.962 40 1.7186 0.982 0.768 0.271 0.899 19.038 1.438 0.558 0.980 60 1.7094 0.985 0.667 0.199 0.923 46.283 0.656 0.655 0.975
MAIYALAGAN & KARTHIKEYAN: FILM-PORE DIFFUSION MODELING FOR SORPTION OF AZO DYE 9solution interface. The kinetics models of AO7 influenced by the two factors, namely (i) distributionadsorption on the xGnPs were analyzed using pseudo of the dye ionized species in the solution phase, andfirst-order14, pseudo second-order15, kinetic models (ii) overall charge of the adsorbent. Therefore, theand Elovich equation16. The isotherm models of AO7 interaction between dye molecule and adsorbent isadsorption on the xGnPs were analyzed using basically a combined result of charges on the dyeLangmuir and Freundlish equation17. molecules and the surface of the adsorbent18. The effect of pH on the adsorption of AO7 by as-receivedResults and Discussion xGnPs has been evaluated in pH range 2-11. WhichCharacterization of xGnPs reveals that the removal of dye slightly decreases, The dye adsorption on xGnPs depends on many when the pH is increased from 2-4 and then remainfactors, such as surface functional groups, specific almost constant up to pH 8. A large decrease insurface area and composition of the solution, the most adsorption capacity for this dye is observed as theimportant factor being the surface area. BET surface pH approaches pKa of AO7 under basic condition.area measurement of xGnPs asserts the large When the solution pH is above the pKa of dyehysteresis area of N2 adsorption-desorption isotherm (pKa for AO7 is 8.86), the adsorption decreases due(Fig.1), suggesting the wide distribution of pores. The to the electrostatic repulsion between dissociatedspecific surface area of xGnPs calculated using BET adsorbate and adsorbent surface. Below the isoelectricequation is found to be 112.67 m2/g. Large hysteresis point (pHpzc of the xGnPs is 8.1), surface of adsorbentarea indicates a near uniform distribution of pores and may acquire positive charge leading to an increasedlarge surface area of xGnPs, suggesting the high anionic dye adsorption due to electrostaticquality of graphene sheets. The SEM image reveals attraction19.that the flat surface of xGnPs is homogeneous Effect of initial concentration of dye solution(Fig. 2), the property being responsible for this The initial concentration of AO7 solution wasselectivity. The large layer of carbon containing varied (20 40 and 60 ppm) and batch adsorptiondelocalized π electrons can explain the different experiments were carried out with 100 mg ofretention mechanisms such as electron transfer, the adsorbent at 30oC and pH 7. An increasedion-pairing and hydrophobic interaction. The surface percentage removal of AO7 from 75 to 90 is observedarea and pH pzc are consistent with the literature18. The with 100 mg of the adsorbent, when the initialproperties of xGnPs specific surface area (SBET), concentration of the AO7 solution is increasedpHpzc, carboxylic acid, lactone groups, and phenolic from 20 ppm. The higher uptake of AO7 at lowgroups values are 112.67 m2 g-1, 8.1, 0.25 meq g-1 concentration may be attributed to the availability0.48 meq g-1 and 0.36 meq g-1 respectively. of more active centers on the surface of the adsorbentEffect of pH for lesser number of adsorbate species. The pH value of the solution being an importantcontrolling parameter in adsorption is mainly Fig. 1 –– Nitrogen adsorption–desorption isotherm of xGnPs Fig. 2 –– SEM image of exfoliated graphitic nanoplatelets (xGnPs)
10 INDIAN J. CHEM TECHNOL., JANUARY 2013Effect of temperature it is thus necessary to perform multiple regressions on Temperature influences the AO7 adsorption different ranges of the data. The kinetics could not beproperties on xGnPs. The temperature effect on the approximated using Elovich model.sorption capacity of xGnPs was examined at 30, 45 Pseudo second-order modeland 60°C using initial dye concentration of 20 mg/L The same data are shown as pseudo second-orderat pH 7. The adsorption capacity of the xGnPs equations in Fig. 5. These plots show that the data fitsincreases with decreasing temperatures from 60°C has good correlation coefficients (>0.962) when theto 30°C, which indicates that the adsorption process pseudo second-order equation is employed. It isis exothermic. The optimum temperature for dye possible to ascertain from them whether the rateadsorption of the adsorbent, within the temperature determining process is a chemical reaction. Thus,range studied, is found to be 30°C. increasing the initial dye concentration from 20 mg/LKinetic modeling to 60 mg/L the AO7 sorbed at any contact time increases. This is obvious for higher initialPseudo first-order model concentration values, as a more efficient utilization of Figure 3 shows a plot of pseudo first-order the sorption capacities of the adsorbent would beequation for the results of adsorption of AO7 from expected due to greater sorption driving force.20 mg/L to 60 mg/L between log (qe-qt) and agitationtime over whole sorption period with high correlation Isothermal modelingcoefficient (>0.986) for all the lines (Table 1). It is The Langmuir adsorption isotherm obtained inclear that the pseudo first-order equation may be used 160 min of agitation time is shown in Fig. 6.to describe the kinetics of sorption of AO7 on toxGnPs. Although the pseudo first-order equation doesnot provide any mechanistic evidence, it has beenproved suitable for highly heterogeneous systems ofwhich the adsorption of AO7 onto xGnPs isundoubtedly such a case.Elovich model The results of the sorption of AO7 on to xGnPshave been represented in the form of Elovich equationin Fig. 4 at various initial dye concentrations (20, 40,and 60 mg/L). From the plot a linear relationshipbetween the amount of AO7 adsorbed, qt and ln(t) areestablished. These plots show different distinct linearregions within individual sets of data. In these cases, Fig. 4 –– Elovich plot for adsorption of AO7 onto xGnPsFig. 3 –– Pseudo first-order plot for adsorption of AO7 onto xGnPs Fig. 5 –– Pseudo second-order plot for adsorption of AO7 onto xGnPs
MAIYALAGAN & KARTHIKEYAN: FILM-PORE DIFFUSION MODELING FOR SORPTION OF AZO DYE 11The values of RL obtained in this study lie within best-fit description for the sorption of AO7 on tothe range 0.127-0.151 indicating the favorable case xGnPs relative toFreundlish isotherm model.of adsorption for the present adsorbent-adsorbate Table 3 shows the adsorption capacity of somesystem. The Freundlich adsorption isotherm obtained of the adsorbents used for the adsorption of Acidin 160 min of agitation is shown in Fig. 7. The values Orange 720-25. It is observed that the adsorptionof absorption intensity 1/n<<1 reveal the applicability capacity of exfoliated graphitic nanoplatelets isof the Freundlich adsorption isotherm in Fig. 7. comparatively good when compared with some of theThe values of 1/n and kf are given in the Table 2. adsorbents already reported in the literature for theThe study of temperature effects on the Freundlich adsorption of Acid Orange 7 in aqueous solution.parameters reveals a decreasing trend in the The differences in maximum adsorption efficienciesadsorption capacity with increase in temperature. of various sorbents might be due to differentHowever, the variation in the adsorption intensity is structures and sorption mechanisms of variousnegligible. These data are useful for practical design sorbents and experimental conditions.purposes. Langmuir adsorption isotherm provides a Mechanism for sorption of AO7 onto xGnPs Because of the high correlation coefficients obtained using pseudo first-order, pseudo second- order and Elovich kinetic models, it is impossible to conclude which adsorption mechanism actually occur and is responsible for the ability of adsorbent to review other sources of information in an attempt to identify the specific adsorption mechanism. In adsorption process of dye on the solid surface, the dye species migrate towards the surface of the adsorbent. This type of migration proceeds till the concentration of the adsorbate species, adsorbed on to the surface of the adsorbent. Once equilibrium is Fig. 6 –– Langmuir plot for adsorption of AO7 onto xGnPs attained, the migration of the solute species from the solution stops. Under this situation, it is possible to measure the magnitude of the distribution of the solute species between the liquid and the solid phases. The magnitude of this kind of distribution is a measure of the efficiency of the chosen adsorbent and the adsorbate species. When graphene platelets are made to contact with a solution containing dyes, the dyes first migrate from the bulk solution to the surface of the liquid film. This surface exerts a diffusion barrier. This barrier may be very significant or less significant. The involvement of a significant quantum of diffusion barrier indicates the dominant role taken up by the Fig. 7 –– Freundlich plot for adsorption of AO7 onto xGnPs film diffusion in the adsorption process. Furthermore, Table 2 –– Parameters of Langmuir and Freundlich adsorption isotherm models for AO7 on xGnPs Temperature Langmuir isotherm Freundlich isotherm o C b Q0 RL R 2 1/n n kf R2 L/mg mg / g 30 0.06826 85.172 0.151 0.987 0.5939 1.485 1.484 0.897 45 0.06637 92.132 0.121 0.979 0.6381 1.694 4.132 0.902 60 0.06571 117.332 0.127 0.981 0.6524 1.7034 6.345 0.887
12 INDIAN J. CHEM TECHNOL., JANUARY 2013 Table 3 –– Comparison of maximum adsorption capacity for (iii) adsorptions of the ingoing ion (adsorbate) on the Acid orange 7 on other different adsorbents. interior surface of adsorbent.Adsorbent Adsorption capacity Reference Out of these three processes the third process is mg/g considered to be not the limiting step in the uptake ofCanola stalks 25.06 20 dyes on to xGnPs28. The remaining two steps impartBeech wood sawdust 5.06 21 the following three possibilities:Spent brewery grains 30.5 22 Case I − External transport < internal transport, whereSoil 3.47 23 rate is governed by particle diffusion.Waste Brewery’s yeast 3.56 24 Case II − External transport > internal transport,Untreated S. marginatum 35.62 25Exfoliated graphitic 85.172 Present study where rate is governed by external diffusion.nanoplatelets Case III − External transport ≈ internal transport, where the transport of the adsorbate ions to thethe rate of an adsorption process is controlled either boundary may not be possible with significant rate,by external diffusion, internal diffusion or by both this may result into a possibility of formation of atypes of diffusion. liquid film surrounded by the adsorbent particles with The adsorption of adsorbent on graphene layer is a proper concentration gradient.remarkably different from other conventional porouscarbons in several aspects. First, due to their two- In the present study, the quantitative treatment ofdimensional nano structure, the external surface the sorption dynamic is found in accordance with theavailable for adsorption is considerably larger than the observation of Reichenberge29, as described by thesurface area arising from inner cavities. The following equation:predominance of outer cavity surface area to inner cavity 6 ∞ 1surface area determines the adsorption characteristics of F =1− π 2 ∑n N −1 2 exp [ − n 2 β t ] … (3)dyes on xGnPs. The adsorption on the external surfaceof graphene nanoplatelets is more important than the where F is the fractional attainment of equilibrium atadsorption inside micro/mesoporous cavities. Another time t; and n, the constant30.noteworthy difference should be ascribed to the Qtinterstitial space between individual graphene sheets. F= … (4)The dimension of this space is determined by the Q∞relative positions among individual graphene sheets. where Q1 and Q∞ are the amounts adsorbed after time In the batch mode contact time adsorption t and after infinite time respectively.experiments, rapid stirring is maintained. This inducesAO7 from the solution to the external surface of π 2 Di B= = time constant … (5)the adsorbent material and this step may control the r02rate of the adsorption process26. To interpret the where Di is the effective diffusion coefficient ofexperimental data it is necessary to recognize the adsorbate in the adsorbent phase; and ro, the radius ofsteps involved in the process of adsorption that adsorbent particles.govern the overall rate of removal of dye. The For energy observed values of F, correspondingingenious mathematical treatments recommended by values of Bt are derived from Reichenberg̕ s table38. InBoyd et al.27 have been applied. These mathematical each case the plot of Bt vs time distinguishes betweentreatments are found to be useful to distinguish the processes involved film diffusion and particles-between particles diffusion and film diffusion. diffusion controlled rate of adsorption.The successive steps in the adsorption dyes by Typical Bt vs time plots at the concentrationadsorbents are: 20 mg/L of AO7 adsorbed on xGnPs at different(i) transport of adsorbates to the external surface of temperature are represented in Fig. 8. It is found to be adsorbent (film diffusion); non-linear throughout the temperature 30, 45 and(ii) transport of adsorbates within the pores of the 60°C, thus the process involved can be represented as adsorbent, except for a small amount of film diffusion. At 30°C the adsorbent exhibits linearity adsorption, which occurs on the external surface in Bt vs time plots in the entire concentration range, (particle diffusion); and but the straight lines obtained do not pass through
MAIYALAGAN & KARTHIKEYAN: FILM-PORE DIFFUSION MODELING FOR SORPTION OF AZO DYE 13 values of ∆ S# reflect that no significant change occurs in the internal structure of chosen adsorbent using the adsorption process. Conclusion The study shows that xGnPs is an effective adsorbent for the removal of AO7 from aqueous solution. The adsorption of AO7 is dependent on the initial concentration and agitation time. Equilibrium of AO7 adsorption reaches at 160 min. The pseudo first- and second-order equations provide a best fit description for the sorption of AO7 onto xGnPs related to Elovich model, but theFig. 8 –– Time vs Bt plots at different temperature of AO7 - xGnPs pseudo first-order correlation coefficient has betteradsorption correlation value than pseudo second-order equation, Table 4 –– Values of energy of activation (Ea), entropy of Pseudo first-order equation is consider to be the activation ( S#), effective diffusion coefficient (Di) and most appropriate due to high correlation coefficient pre-exponential factor (Do) when compared to pseudo second-order equation, and Parameter Value adsorption takes place via film diffusion process. Di, cm2s-1 Langmuir and Freundlich adsorption isotherms 30° C 1.4687 × 10-11 correlate the equilibrium adsorption data. The 45° C 1.313 × 10-11 adsorption of AO7 onto xGnPs is an exothermic 60° C 1.093 × 10-11 reaction based on enthalpy change values. Ea, kJmol-1 -9.7153 S#, JK-1mol-1 -179.53 Acknowledgement Do, cm2s-1 9.4932 × 10-12 The authors acknowledge with thanks the supportorigin, revealing thereby that the rate-determining of Department of Chemistry, Chikkanna Govt.process is film diffusion at this temperature for Arts College, Tirupur and Sophisticated Analyticalchosen adsorbent. Instrument Facility, Indian Institute of Technology, The Di values were also calculated for each Madras for characterization process.adsorbent material at the three different temperatures References(30, 45 and 60°C) using Eq. (6), and the values 1 Kaviyarasan V, Mohan N, Kannan V, Ebenezer P &observed specify that Di increases within increasing Karthikeyan S, Poll Res, 22 (1) (2003) 77.temperature. This may be due to the increased 2 Behnajady M A, Modirshhla N, Daneshvar N & Rabbani M,mobility of ions and decreased retarding forces acting Chem Eng J, 127 (2007) 167.on diffusing ion. The energy of activation (Ea), 3 Gupta V K & Suhas, J Environ Manage, 90 (2009) 2313.entropy of activation ( S#), and pre-exponential 4 Rafatullah M, Sulaiman O, Hashim R & Ahmad A, J Hazard Mater, 177 (2010) 70.constant (Do) analogous to the Arrhenius frequency 5 Rao G P, Lu C & Su F, Sep Purif Technol, 58 (2007) 224.factor are evaluated indicating no significant change 6 Yao Y, Xu F, Chen M, Xu Z & Zhu Z, Bioresour Technol,in the internal structure of xGnPs during the 101 (2010) 3040.adsorption, as shown below: 7 Ji L L, Chen W, Zheng S R, Xu Z Y & Zhu D Q, Langmuir, 25 (2009) 11608.Di = Do exp [− E a / RT ] … (6) 8 Li X & Gen G, Mater Lett, 63 (2009) 930. 9 Vieira F, Cisneros I, Rosa N G, Trindade G M & MohallemDo = (2.72d 2 kT / h) exp ∆S # / R … (7) N D S, Carbon, 44 (12) (2006) 2590. 10 Tryba B, Morawski W A, Kalenczuk R J & Inagaki M, Spillwhere d is the average distance between the Sci Technol B, 8 (5-6) (2003) 569.successive exchange sites and is taken as 5 Å; and R, 11 Kalaitzidou K, Fukushima H & Drzal L T, Composites, 38 (2007) 1675.h and k are the gas, plank and Boltzmann constants 12 Zu S Z & Han B H, J Phys Chem C, 113 (31) (2009) 13651.respectively. The values of Ea, Di, Do, ∆ S# and other 13 Lopez-Ramon M V, Stoeckli F, Moreno-Castilla C ¶meters are given in the Table 4. The negative Carrasco-Marin F, Carbon, 37 (1999) 1215.
14 INDIAN J. CHEM TECHNOL., JANUARY 201314 Lagergren S, Handlingar, 24 (4) (1898) 1. 23 Smaranda C, Bulgariu D & Gavrilescu M, Environ Eng15 Ho Y S, McKay G, Wase D A J & Forster C F, Adsorp Sci Manag, 8 (2009) 1391. Technol, 18 (2000) 639. 24 Wu Y, Hu Y, Xie Z, Feng S, Li B & Mi X, Appl Biochem16 Chien S H & Clayton W R, Soil Sci Soc Am J, 44 (1980) 265. Biotechnol, 163 (2011) 882.17 Freundlich H M F, Z Phys Chem, 57 (1906) 385. 25 Kousha M, Daneshvar E, Sohrabi M S, Jokar M & Bhatnagar A, Chem Eng J, 192 (2012) 67.18 Luo P, Zhao Y F, Zhang B, Liu J D, Yang Y & Liu J F, Water Res, 44 (2010) 1489. 26 Weber Jr W J & Morris J C, J Sanit Eng Div, Am Soc Civil Engineers, 89 (1963) 31.19 Peng Y, Fu D, Liu R, Zhang F & Liang X, Chemosphere, 71 (2008) 990. 27 Boyd G E, Adamson A W & Meyers L S (Jr), J Am Chem Soc, 69 (1947) 2836.20 Hamzeh Y, Izadyar S, Azadeh E, Abyaz A & Asadollahi Y, Iran J Health Environ, 4 (2011) 49. 28 Bhattacharya A K &Venkobachar C, Adsorption J Environ Eng Div Am Soc Civil Engineers, 110 (1984) 110.21 Izadyar S & Rahimi M, Pak J Biol Sci, 10 (2) (2007) 287.22 Safarik I, Horska K & Safarikova M, J Cereal Sci, 53 29 Reichenberg D, J Am Chem Soc, 75 (1953) 589. (2011) 78. 30 Gupta V K & Ali I, J Colloid Interface Sci, 271 (2004) 321.