Biosorption of heavy metals from aqueous solutions with tobacco dust
B.C. Qi, C. Aldrich
Department of Chemical Engineering, University of Stellenbosch, Stellenbosch, Private Bag X1, Matieland 7602, South Africa
Received 21 November 2006; received in revised form 19 October 2007; accepted 24 October 2007Available online 21 December 2007
A typical lignocellulosic agricultural residue, namely tobacco dust, was investigated for its heavy metal binding efficiency. The tobaccodust exhibited a strong capacity for heavy metals, such as Pb(II), Cu(II), Cd(II), Zn(II) and Ni(II), with respective equilibrium loadingsof 39.6, 36.0, 29.6, 25.1 and 24.5 mg of metal per g of sorbent. Moreover, the heavy metals loaded onto the biosorbent could be releasedeasily with a dilute HCl solution. Zeta potential and surface acidity measurements showed that the tobacco dust was negatively chargedover a wide pH range (pH > 2), with a strong surface acidity and a high OH
adsorption capacity. Changes in the surface morphology of the tobacco dust as visualized by atomic force microscopy suggested that the sorption of heavy metal ions on the tobacco could be asso-ciated with changes in the surface properties of the dust particles. These surface changes appeared to have resulted from a loss of some of the structures on the surface of the particles, owing to leaching in the acid metal ion solution. However, Fourier transform infrared spec-troscopy (FTIR) showed no substantial change in the chemical structure of the tobacco dust subjected to biosorption. The heavy metaluptake by the tobacco dust may be interpreted as metal–H ion exchange or metal ion surface complexation adsorption or both.
 2007 Elsevier Ltd. All rights reserved.
 Biosorption; Heavy metals; Lignocellulose; Tobacco
1. Introduction
The increased attention on the harmful effects of heavymetal ions on human health and the environment overthe past few decades has led to a concomitant focus onimproved water quality. Since industrial effluents in partic-ular are a major contributor to heavy metal contamination,the removal of such metals from these effluents has been apriority in the tightening and enforcement of environmen-tal regulations. Removal of these contaminants can beaccomplished by use of physical or chemical methods,including the use of chemical reagents, ion exchange, acti-vated carbon sorption and membrane technology. Most of these approaches have significant disadvantages. Forexample, chemical reagents are costly (Kadiverlu et al.,2001), as the active agents cannot be recovered for use insuccessive treatment cycles. Moreover, the end product isusually a low volume, highly concentrated metalliferoussludge that may be difficult to dewater and dispose of (San-dau et al., 1996). Likewise, ion exchange and membranesystems could be expensive, especially in small-scale pro-cesses, with the resins or membranes prone to fouling oroxidation. Similarly, activated carbon, the most widelyused adsorbent in the treatment of waste water, is expen-sive and may also require complexing agents to improveits ability to remove inorganic matter (Babel and Kurnia-wan, 2002).Biosorption, which is the ability of certain biomaterialsto bind and concentrate heavy metals from even the mostdilute aqueous solutions, offers a technically feasible andeconomically attractive alternative to the conventionaltechnologies for removal of heavy metal from the contam-inated effluents (Davis et al., 2000; De Carvalho et al.,2001; Esposito et al., 2001; Demir and Arisoy, 2007).Variousbiomaterialsproducedorharvestedfromnaturalresources or agricultural products, mostly in metabolically
0960-8524/$ - see front matter
 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.biortech.2007.10.042
Corresponding author. Tel.: +27 21 808 4485; fax: +27 21 808 2059.
E-mail address: (C. Aldrich).
 Available online at
Bioresource Technology 99 (2008) 5595–5601
inactive states, have been used in disposal of heavy metaleffluents by biosorption. These include microorganisms,such as bacteria, fungi, yeast and algae and lignocellulosebiomaterials, such as peat moss, raw rice bran, rice straw,coconut husks, waste coffee powder, dried plant leaves,etc. (Oliveira et al., 2005; Orhan and Buyukgungor, 1993;Zouboulis et al., 1999; Feng and Aldrich, 2004). Amongthese biomaterials, marine algae and peat moss have beenstudied extensively to sequester heavy metals from contam-inated effluents (Zu¨mriye, 1997; Brown et al., 2000).However, among others, owing to the heterogeneous com-positionand complex chemical structures ofvarious bioma-terials, as well as a wide spectrum of heavy metals, standardcriteria through which biosorption can be industrialized arestill lacking at present. For lignocellulosic biomaterial (e.g.,peat moss and black liquor), it is especially the lignin andhumic acids that are involved in the chemical bonding of heavy metals during biosorption (Brown et al., 2000; Babeland Kurniawan, 2002).Furthermore, besides a strong metallic affinity, thesearch for a low-cost and easily available adsorbent hasled to the investigation of materials of agricultural srcinas potential metal adsorbents (Pagnanelli et al., 2001; Shethand Soni, 2004). These low cost materials have not beenstudied as extensively as other biosorbents, owing to theirlocal, rather than global availability (Babel and Kurnia-wan, 2002) and therefore, in this investigation, the biosorp-tion capacity of a typical lignocellulosic plant biomass anda waste product from the tobacco industry, namely tobaccodust, is considered.
2. Experimental work
 2.1. Materials
The heavy metal ions tested in this study includedPb(II), Cu(II), Cd(II), Zn(II) and Ni(II). The metal ionsolutions were made up of atomic absorption standardmetal solution (Merck, 1000 mg/L) by diluting with dis-tilled water to the required concentrations. Analyticalgrades of HCl and NaOH (Merck) solutions were usedfor pH adjustment. The air-dried tobacco dust wasobtained in ground form from the British AmericanTobacco Company in Paarl, South Africa, with a particlesize of 1–2 mm.
 2.2. Methodology
The tobacco dust was protonated with 0.1 M HCl at 1:5(w/v) for 1 h, followed by washing with distilled water untilthe pH of the suspensions reached 4.5–5.0, and dried at105
C. The tobacco dust had the following mass-basedcomposition: 30% lignin, 94.2% total solids, 72.0% volatilesolids, 28.0% ash, 43.0% carbon and 2.37% nitrogen, 1.70%potassium, 4.29% calcium and 0.70% magnesium. It alsocontained 26, 57, 0.57, 761 and 288 mg/kg of B, Cu, Fe,Zn and Na, respectively.Batch equilibrium sorption was performed in 1 L screwcapped plastic bottles on a roller at a speed of 58–60 rpm atambient temperatures (23–25
C) for 24 h. The pH valuesof the final suspensions were measured at equilibrium sorp-tion conditions. The suspension was then filtrated and theindividual metal ion concentrations of the filtrate were ana-lyzed as the equilibrium concentrations of each metal ionspecies. Metal ion loadings on the biosorbent were deter-mined from the difference of metal ion concentrations inthe initial and final suspensions. The capacity of the biosor-bent for the single species sorption system was estimatedfrom the equilibrium metal ion loading values, whichshowed no obvious change with an increase in the metalion loading. However, for the multi-species sorption sys-tem, the equilibrium metal ion loading values were notobserved over the concentration ranges of the multiplemetal ions under the investigation. Thus no sorption capac-ity values could be obtained for the multi-species sorptionsystem. For both sorption systems, the biosorbent load was1.0 g/L metal solution.Batch kinetic tests were performed at 25
C using anoverhead stirrer with a baffle attachment at stirring speedsof 800–1000 rpm. The biosorbent load was 1.0 g/L metalsolution.Desorption experiments were carried out under the sameconditions as the batch sorption experiments, except thatboth the biosorbent and heavy metal ion loadings weredoubled in the sorption stage. After 24 h of contacting,reaction suspensions were filtrated and the filtrated cakeswere washed three times using the same volumes of distilledwater as the sorption suspensions. This was done in orderto remove the physically attached metal ions from theloaded tobacco dust. The washed solids were oven driedovernight at 105
C for subsequent desorption tests.Desorption tests were carried out by magnetically suspen-ding the loaded biosorbents in a series of HCl solutionsat varied concentrations (the suspension contained 1%weight of the metal loaded biosorbent). After the requiredcontact time, the reaction mixture was filtrated and themetal ion content of the filtrate was analyzed. Desorptionefficiency was calculated as a percentage of the amount of the released metal ions relating to the total amount of metal ions which were previously adsorbed on thebiosorbent.The metal ion content in the solution was measuredusing an atomic absorption spectrophotometer (AA 250plus) and ICP spectrometry (Varian Liberty Series IISequential ICP AES).The morphological microstructures of the tobacco dustwere examined by atomic force microscopy (AFM) usinga TMX 2000 Explorer (Topometrix, American, force con-stant of probe, 35–65 N/m). The surface roughness wasestimated from the AFM images and calculated with theAFM surface analysis software (SPMLab V3.0 6.0).The surface acidity of the biosorbent was investigatedusing alkalimetric titration (Huang, 1981). To prepare thesamples for titration, 140 mL of biosorbent suspension
 B.C. Qi, C. Aldrich / Bioresource Technology 99 (2008) 5595–5601
(1.0 g/L) was transferred into each of four beakers. Twowere directly titrated, one with acid (0.1 N HClO
) andthe other with base (0.1 N NaOH) (sample titration). Theother two were titrated after removing the biosorbent solidby passing through a 0.45
m filter (blank titration). Theinitial pH value was measured, followed by titration of the solutions under continuous mixing using 0.1 N HClO
to pH 3.0 or 0.1 N NaOH to pH 11.0. The volumes of theacid or base added and the corresponding equilibrium pHvalues were recorded. The pH values were taken every30 min after each addition of acid or base. Pure N
 gaswas continuously purged to exclude CO
 during titration.The change of pH after each acid or base addition wasensured not to exceed 0.3 units.The surface charging characteristics of the biosorbentwere examined by zeta potential measurements (MalvernZetasizer 4) on a multi-species sorption system. Underthe same pH conditions, the biosorbent suspensions inthe absence of heavy metals were used as the blanksamples.Changes in the surface structures of the tobacco dustsubjected to metal sorption and desorption were examinedusing FTIR spectra. The KBr pellets were prepared with amixture of 1.0 mg oven-dried tobacco dust with 300.0 mgKBr (both were dried at 105
C for 48 h). The FTIR spec-tra of the pellets were recorded on a Perkin Elmer 1600 Ser-ies FTIR spectrometer.
3. Results and discussion
3.1. Kinetic experiments
The changes in solution pH and heavy metal removalefficiency in a multi-species sorption system with sorptiontime are shown in Fig. 1. The pH of the sorption systemdecreased steadily during the first 370 min, then stabilizedthroughout the sorption process. This implies that themetal binding to the biosorbent was associated with H
 – metal ion exchange. Almost all the heavy metals achievedtheir sorption equilibria within approximately 90 min atan equilibrium pH of 6.34. Dynamic biosorption selectivityof the tobacco dust for heavy metals was approximately inthe order: Zn(II)
Cu(II) > Cd(II) > Pb(II) > Ni(II).
3.2. Equilibrium experiments
Equilibrium experiments were performed for both sin-gle-ion sorption systems and multi-species sorption sys-tems. The solution pH was presumed to be an importantfactor influencing the heavy metal distributions in theliquid phase, as well as the interactions between the metalions and the biosorbent. Fig. 2 indicates that the influenceof the pH on the biosorption of the heavy metals generallyfollowed the expected metal sorption behaviour, with theremoval efficiency increasing with an increase in the pHof the solution, but decreasing under high equilibriumpH (pH > 9) conditions.Sorption started at a pH of 4–5 and began to decline at apH of 9–10 for most of the heavy metal ions. This suggeststhat sorption from dilute hydrolysable metal ion solutions(such as that of a heavy metal) usually occurs at a pHbelow the region where hydrolysis occurs. This is wheresorption may be driven either by ion exchange or surfacecomplexation, in which metal ions replace surface protons,or as the preferential sorption of the hydrolyzed species(James, 1981). FTIR analysis has suggested the presenceof aromatic structures, nonaromatic double bonds, H-bonded C
O of conjugated ketones and quinines and sym-metric COO- groups (Sparks, 1999) in the tobacco dustbiomass. At a pH of 4–5, the ionic state of ligands of func-tional groups, such as carboxyl, phosphate and amino-groups, which are mostly the chemical structures of the sur-face cell walls of the tobacco dust, would promote a reac-tion with metal ions. This reaction would occur throughelectrostatic attraction between positively charged metalcations and negatively charged binding sites on the tobaccodust (detailed surface charge conditions will be discussedlater). This would result in a rapid rise in the metal removal
01020304050607080901000 2 4 6 8 10 12 14 16 18 20 22 24
Contact time (hour)
   I  o  n  r  e  m  o  v  a   l   (   %   )
Cd CuNi PbZn
Fig. 1. Heavy metal adsorption kinetics of the tobacco dust (pH = 6.34).
01020304050607080901001 2 3 4 5 6 7 8 9 10 11 12
   I  o  n  r  e  m  o  v  a   l   (   %   )
Cd CuNi PbZn
Fig. 2. Changes in heavy metal removal efficiency of the tobacco dust withequilibrium solution pH for single-species sorption.
B.C. Qi, C. Aldrich / Bioresource Technology 99 (2008) 5595–5601
efficiency at a pH of 4–5. As the pH is lowered, the mea-sured surface charge on the tobacco would become morepositive, leading to a weakening of the electrostatic interac-tion of metal ions with the tobacco dust.Boththesingleandmultiplespeciessorptionisothermsof thetobaccodustarepresentedinFig.3.Thesorptioncapac-ities for Pb(II), Cu(II), Cd(II), Zn(II) and Ni(II) of thetobacco dust were approximately 39.6, 36.0, 29.6, 25.1 and24.5 mg/g sorbent, respectively, in a single species sorptionsystem. In contrast, under the same equilibrium metal con-centrations, the heavy metal sorption isotherms in multi-species sorption system were different from those in a singlespecies sorption system. The selectivity of the metal bindingcapacities of the tobacco dust also differed in the single andmultiple species systems. In a multi-species sorption systemit was in the order of Cu(II) > Pb(II) > Zn(II) > Cd(II) >Ni(II), while in the single species sorption system it was inthe order of Pb(II) > Cu(II) > Cd(II) > Ni(II)
Zn(II).In general, three types of interactive sorption behaviourcould be displayed by a mixture of heavy metals, viz. syn-ergism, antagonism or non-interaction. With synergism,the effect of the mixture is greater than the sum of eachof the individual effects of the constituents in the mixture.With antagonism, the effect of the mixture is less than thatof the sum of the individual effects of the constituents in themixture. With non-interaction, the effect of the mixture isequivalent to the sum of each of the individual effects of the constituents in the mixture (Zu¨mriye, 1997).The sorption competition among these heavy metal ions,especially among Cu(II), Cd(II) and Ni(II), was evidentfrom comparison of the metal uptake in the single and mul-tiple species systems. Under the same initial metal concen-trations, the metal loading of Cu(II) was generally higher inthe multi-species sorption system than in the single speciessorption system. Conversely, the metal loadings of Cd(II)and Ni(II) were generally lower in the multi-species sorp-tion system than in the single species sorption system. Onthe whole, the metal loadings of Pb(II) and Zn(II) appearedsimilar in both multi-species and single species sorptionsystems. Therefore, synergism (for Cu(II)), antagonism(for Cd(II) and Ni(II)) and non-interaction (for Pb(II)and Zn(II)) were present simultaneously in the multi-spe-cies sorption system, depending on the species of heavymetals. When the combined effects of more than one metalion on the same biomass were found to be antagonistic,such as the interactions between Cd(II) and Cu(II) or Ni(ii)and Cu(II), it can be postulated that the effective number of binding sites available for the uptake of a single metal spe-cies was reduced, depending on the sorption equilibria of the other cations. The extent of this reduction for the indi-vidual species is possibly related to the electro-negativitiesof the cations. In contrast, screening or competition forthe binding sites on the biomass surface could result inthe metal ions mutually improving their individual affinityfor the biomass, leading to a synergistic effect. A thermo-dynamic model could be developed by extracting the activemetal ion absorption sites on the surfaces of the tobaccodust to prove this postulation. However, such a develop-ment falls beyond the scope of this paper.
3.3. Desorption experiments
Desorption of heavy metal ions from the loaded tobaccodust with different concentrations of HCl solutions(desorption time 6 h) is shown in Fig. 4. The changes of desorption efficiency with time is given in Fig. 5, for a0.09 mol/L HCl solution.Fig. 4 shows that except for the Pb(II), almost all of the heavy metals adsorbed on the tobacco dust couldbe desorbed easily by a dilute HCl solution (less than0.01 mol/L). The desorption efficiency was in the order:Ni(II) > Zn(II) > Cd(II) > Cu(II) > Pb(II). This confirmedthat the tobacco dust had higher binding capacities forPb(II) and Cu(II) than for Ni(II) in particular, which could
04812162024283236400 2 4 6 8 10 12 14 16 18
Equilibrium concentration of heavy metals (mg/l)
   M  e   t  a   l   l  o  a   d   i  n  g   (  m  g   /  g   )
Cd CuNi PbZn
0246810121416182022242628300 1 2 3 4 5 6 7 8 9 10
Equilibrium concentrations of heavy metals (mg/l)
Fig. 3. Heavy metal sorption isotherms of the tobacco dust at pH 6.5–7.2 (left: single-species sorption, right: multi-species sorption).5598
 B.C. Qi, C. Aldrich / Bioresource Technology 99 (2008) 5595–5601
even be desorbed partially by distilled water (data notshown).Fig. 5 shows that desorption of heavy metals with HClsolutions (pH < 2) from the tobacco dust is a rapid process,which can be accomplished in less than 60 min. This sug-gests that H
has a higher competitive capacity than theheavy metal ions under low pH condition (pH < 2).
3.4. Surface morphology changes of the tobacco dust owing to biosorption and desorption
Atomic force microscopic images have shown that thesurfaces of the tobacco dust changed from fine homoge-neous particulate textures to structures dominated bybump-like aggregates, with some areas being significantlysmoother than others, following the sorption of heavy met-als and the desorption of hydrochloric acid. The presenceof smooth surface areas and channel-like configurationsindicated that certain components of the tobacco dustcould be leached off the surface by the HCl or heavy metalsolutions. This also suggests that the sorption of heavymetal ions on the tobacco dust could be associated withchanges in the local surface properties owing to the lossof some of the structural components of the tobacco dust.
3.5. Surface acidity of tobacco
Surface acidity is an important property of hydrous sol-ids which are closely related to the mode and extent of interfacial reactions such as adsorption and coagulation.As mentioned previously, surface adsorption was consid-ered to be an important mechanism by which heavy metalions may be bound to the tobacco dust.As a typical natural lignocellulosic biomass, tobaccodust is a complex material with lignin, cellulose and hemi-cellulose as its major constituents. These constituents gen-erally have polar functional groups, such as alcohols,carboxylic acids and phenolic hydroxides. The hetero-geneous functional groups on the surface structure of thetobacco dust make it difficult to distinguish, by conven-tional analytical means, between dissociation of protonsfrom the surface and adsorption of OH
ions onto the sur-face. However, simply recording the pH values during thecourse of titration of a suspension is sufficient to interpretthe surface sorption property of the biomass in terms of adsorption of hydrogen or hydroxide ions or their com-plexes (Huang, 1981).Titration analysis has shown that the pH of the suspen-sion (supernatant) decreased or increased sharply corre-sponding to the acid or base added, since the free H
ions released from the biomass surface were initiallyneutralized by the added base or acid. In contrast, thepH plateau on the titration curves resulting from furtheraddition of base or acid, represents a surface adsorptionor complexation of H
and or OH
with the solubleorganic components of the tobacco dust, which were previ-ously leached into the solution.Operationally, the amount of adsorption is related tothe number of appropriate ions that cannot be accountedfor in a mass balance between the titration curve of sus-pension and that of the supernatant. Ions not in solutionhave to be on the surface of the biomass, so that adecreased rate of increase in the pH of the suspensioncompared to that of supernatant with the same dosageof the base indicates the adsorption or complexation(neutralization) of OH
on the solid surface of thetobacco dust, possibly accompanied by proton release.Conversely, an increased rate of increase in the pH of the suspension compared to that of the supernatant withthe same dosage of the acid indicates the adsorption orcomplexation of H
on the solid surface of the tobaccodust. This also confirms the amphoteric property of thetobacco dust, which can adsorb both OH
and H
.The dosage of base or acid that cannot be accountedfor by the change of the pH of the supernatant, suggeststhe complexation of H
and or OH
with the organiccompounds in the supernatant, confirming the leachabil-ity of the organic component of the tobacco dust.In addition, titration analysis has shown that a largeramount of base was needed for the suspension than forthe supernatant to reach an equilibrium pH value of 11.This indicated that the solid surface of the tobacco dust
01020304050607080901000 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
HCl concentration (mol/l)
   D  e  s  o  r  p   t   i  o  n  e   f   f   i   t   i  e  n  c  y   (   %   )
Pb Zn CdCu Ni
Fig. 4. Desorption of heavy metal ions from the tobacco dust withdifferent concentrations of HCl solution.
901000 4 8 10 12 14 16 18 20 22 24
Time (hour)
   D  e  s  o  r  p   t   i  o  n  e   f   f   i  c   i  e  n  c  y   (   %   )
Pb Zn CdCu Ni
2 6
Fig. 5. Desorption of heavy metal ions from the tobacco dust with0.09 mol/L HCl solution.
B.C. Qi, C. Aldrich / Bioresource Technology 99 (2008) 5595–5601
of 7