Modeling of Single and Competitive Metal Adsorption onto a Natural

Environ. Sci. Technol. 2002, 36, 2242-2248

Modeling of Single and Competitive Metal Adsorption onto a Natural Polysaccharide ZACARIA REDDAD, CLAIRE GERENTE,* YVES ANDRES, AND PIERRE LE CLOIREC Ecole des Mines de Nantes, GEPEA, 4 rue Alfred Kastler, BP 20722, 44307 Nantes Cedex 3, France

Sugar beet pulp, a common agricultural waste, was studied in the removal of metal ions from aqueous solutions. Potentiometric titrations were used to characterize the surface acidity of the polysaccharide. The acid properties of the material can be described by invoking three distinct types of surface functional groups with the intrinsic acidity constants (pKaint) values 3.43 ( 0.1, 6.05 ( 0.05, and 7.89 ( 0.1, respectively. The contents of each functional group (i.e., the carboxyl and phenol moieties) were also determined. Then, a simple surface complexation model with the diffuse layer model successfully described the sorption of several metal ions (Cu2+, Zn2+, Cd2+, and Ni2+) onto the polysaccharide under various experimental conditions: pH ranging from 2 to 5.5, ionic strength from 0.01 to 0.1 M, metal concentration between 10-4 and 10-3 M, for a constant sorbent concentration equal to 2.5 g‚L-1. It was observed experimentally that the affinity of the polysaccharide was in the sequence of Cu2+ > Zn2+ > Cd2+ > Ni2+. Predictions of sorption in binary-metal systems based on single-metal data fits represented competitive sorption data reasonably well.

Introduction Because of its harmful effects, metal pollution is considered to be among the most serious threats to the environment. Natural waters can easily be polluted by metal ions as a result of their release by industrial plants or mining activities. Therefore, alternative treatment methods have been investigated, involving the sorption of metal ions onto low-cost materials such as fly ash (1, 2), peat (3, 4), microbial biomass (5), and numerous agricultural byproducts (6, 7). Sugar beet pulp, a common waste from sugar refining industry, competes both on a cost and performance basis because this cheap and available adsorbent has been found to be efficient in the removal of metal ions from aqueous solutions (8, 9). Intensive studies have been performed to define the interaction phenomena occurring at the solid/water interface. Many batch experiments for the sorption of metal ions have been reported, and different models have been successfully used to describe these experiments. However, most of them fail to reproduce the data over a wide range of experimental conditions. For example, the adsorption isotherms, such as Freundlich and Langmuir equations, cannot be used to describe the metal adsorption in a wide pH range, as the models parameters are normally pH-dependent. Moreover, the empirical Freundlich model cannot be easily interpreted * Corresponding author phone: 33 (0) 2 51 85 82 85; fax: 33 (0) 2 51 85 82 99; e-mail: [email protected]. 2242

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 10, 2002

in mechanistic terms. Because pH is one of the key parameters in biosorption, it is desirable to use isotherm equations that can accommodate pH and ionic strength as model variables. The surface complexation models, which were developed to describe the interaction of metal ions in solution with natural oxides (10-15), are able to take into account the effects of these factors. In the case of metal oxides, these mechanistic models were successfully used to simulate the sorption of mercury(II) onto iron oxide and amorphous silica (16, 17) and under a wide range of pH. Over the past few years, these models have been extended to various adsorbents. The surface complexation models showed encouraging success at describing metal sorption onto activated carbon in single and competitive systems (18-21). By using a discrete site model, it has been shown that this approach was also able to describe well the metal sorption onto heterogeneous organic materials such as humic acids (22), bacteria (23), and some agricultural byproducts (24, 25). One of the problem facing the application of these models to heterogeneous organic materials is the difficulty in characterizing each individual moieties involved in the metal uptake. In this work, the use of sugar beet pulp in the removal of toxic heavy metals from aqueous solutions has been investigated. First, acid and base titration experiments were performed to determine the intrinsic acidity constants and the concentrations of the different surface functional groups. The sorption of Cu2+, Zn2+, Cd2+, and Ni2+ cations as a function of pH, ionic strength, and metal concentration has been performed. Then, a simple and discrete site model with the diffuse layer model was applied to describe the metal sorption as a function of these latter parameters. Given the adjusted intrinsic adsorption constants, predicting competitive fixation in binary metal systems validated the model proposed.

Experimental Section Adsorbent. The sugar beet pulp was obtained from Lyven (Cagny, France). It was ground using a hammer mill and then sieved. The samples of particle size smaller than 250 µm were discarded, as they probably contained soil and sand particles (26). Only material with a particle size between 250 and 500 µm was used for the metal sorption experiments. This fraction was washed with deionized water (20 g‚L-1) for 16 h, filtered off, and air-dried at 40 °C. In a previous study (28), a complete characterization of the material revealed that the polysaccharides accounted for 72.5% of the dry matter. The sugar beet pulp contained three major components (i.e., glucose (21.2% of the dry matter), arabinose (19.9%), and galacturonic acid (20.0%)). The specific surface area (374 m2‚g-1) was measured by BET H2O adsorption at 30 °C in order to better approximate the wet-sorbent conditions (24). This method was representative of the swelling of the polysaccharide observed when immersed in water (8, 27). Potentiometric Titrations. The potentiometric measurements were carried out at 20 ( 0.5 °C, using a glass electrode and under a nitrogen stream in order to avoid the dissolution of carbon dioxide in the solution. The equipment used consisted of a titrimeter (TIM 900 radiometer) with an automatic buret. The beet pulp (1 g) was first stirred in 100 mL of 0.001, 0.01, and 0.1 M NaNO3 solutions until the pH remained constant (1.5 h). The titrations were then conducted by adding 0.1 M HNO3 or 0.1 M NaOH. Between each titrant addition, the equilibrium was considered reached when the drift readings were less than 0.1 mV for 10 s. Last, the same titrations were performed on a control (i.e., NaNO3) solutions 10.1021/es010237a CCC: $22.00

 2002 American Chemical Society Published on Web 03/27/2002

without beet pulp. The intrinsic acidity constants Ka1int, Ka2int, and Ka3int were calculated by using the single extrapolation method (14, 15, 29). Boehm’s Method. Potentiometric titrations were also performed according to Boehm’s method (30) by adding 0.1 M base solutions of increasing strength (NaHCO3, Na2CO3, NaOH, and NaOC2H5) to the protonated sugar beet pulp. The acidity constants of carboxyl groups, lactones, and phenols differ over several orders of magnitude, and this method consists of differentiating these various types of groups by their neutralization behavior. The experimental methods and functional group identification results have been reported in a previous study (28). Adsorption of Metals Influenced by pH. All of the experimental conditions were fixed so that chemical precipitation would be avoided. First, aqueous solutions (400 mL) of 0.1 or 0.01 M NaNO3 were mixed with 1 g of sugar beet pulp for 1.5 h in order to hydrate the polysaccharide and to keep a constant ionic strength during the experiments. Once the equilibrium pH was reached (around 5.5), the pH values were fixed with HNO3 1 M. The solution pH ranged from 2 to 5.5 and the initial metal concentrations between 10-4 and 10-3 M. The experiments were performed in 0.5 L closed glass flasks at room temperature (20 ( 0.5 °C). The time to reach equilibrium depended on the nature of the metals and on the pH. It ranged from 0.5 to 1 h, according to our previous studies (28, 32). The final pH was measured, and the aqueous samples from adsorption experiments were filtered through 0.45 µm cellulose acetate membrane and then analyzed with an atomic absorption spectrophotometer (Perkin Elmer 2280). Modeling. Once the intrinsic acidity constants and the total number of surface sites were determined experimentally, the adsorption constants were evaluated for all of the metals studied. The functional groups identified with Boehm’s titrations were expected to have the most significant role in the fixation of the metal ions. Given the heterogeneous composition of the polysaccharide, each functional group was considered in the model. Hence, the surface acidity of the polysaccharide was described by a 3-pK heterogeneous model (23). According to this model, three surface acidity constants are defined by mass action laws, including an electrostatic correction term due to the surface charge

≡COOH T ≡COO- + H+ FΨ0 {≡COO-}{H + } Kaint ) (1) exp RT {≡COOH}

(

)

where Kaint is the intrinsic surface acidity constant, R is the molar gas constant (8.314 J‚mol-1‚K-1), F is the Faraday constant (96 485 C‚mol-1), T is the absolute temperature (K), and Ψ0 is the potential at the surface (V). (≡COOH) refers to the surface sites and {} represents the activity. Adsorption equilibrium can be described by the formation of inner-sphere surface complexes which are characterized, on one hand, by the formation of bonds between a free metal cation and a functional group and, on the other, by the release of protons from the surface. Hence, a single-metal surface complexation reaction was employed for each functional group 2+

≡COOH + Me

+

+

T ≡COOMe + H FΨ0 {≡COOMe+}{H+} exp ) (2) 2+ RT {≡COOH}{Me }

int K≡COOMe +

( )

The diffuse layer model (DLM) was used to consider the change in the surface charge and to compute the relationship between surface potential Ψ0 (V) and surface charge σ0 (C‚m-2). DLM was selected because it presents the advantages

FIGURE 1. Potentiometric titrations of the protonated polysaccharide by the addition of base solution (0.1 M) of increasing strength.

TABLE 1. Estimation of the Acidic Functional Groups onto the Sugar Beet Pulp concentration (µeq‚g-1) strong carboxyl groups weak carboxyl groups very weak phenolic groups estimation of the CEC

246 ( 3 220 ( 5 109 ( 5 575 ( 13

pKaint 3.43 ( 0.10 6.05 ( 0.05 7.89 ( 0.10

of simplicity because only a few parameters have to be optimized (14, 15, 24)

σ0 ) x8RT0I × 103 sinh

ZΨ0F 2RT

(3)

where  is the dielectric constant of water (78.5 at 25 °C), 0 is the permittivity of free space (8.854 10-12 C‚V-1‚m-1), Z is the valence of the electrolyte, and I is the ionic strength (M). The mechanisms presented in Table 2 can be applied to all of the active binding sites of the sugar beet pulp. In addition to the free metal ion Me2+, the inorganic species Me(OH)+ and Me2(OH)22+ were included in the equilibrium calculations, using the formation constants compiled by Smith and Martell (33). The intrinsic adsorption constants values of the surface species were fitted using the program FITEQL (34). All surface complexation reactions used in modeling are summarized in Table 2. Last, the chemical speciation of the soluble metallic species was calculated for a total metal ion concentration equal to 10-3 M in aqueous solution (0.1 M NaNO3) with hydrolysis equilibrium constants given by ref 33 (see Supporting Information).

Results and Discussion Surface Acidity. Figure 1 shows the potentiometric titration curves of the protonated material according to Boehm’s method. It is obvious that different moieties with distinct acidities were present according to the shapes of the titration curves. The quantitative and qualitative estimations of the functional groups are summarized in Table 1. The titration curves obtained by the addition of Na2CO3 solution have always presented two distinct waves, showing the presence of strong and weaker acidic groups. These two acidities predominated in the sugar beet pulp (466 ( 8 µeq‚g-1) and VOL. 36, NO. 10, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2243

FIGURE 2. Single extrapolation method applied to the polysaccharide titration data (10 g‚L-1 adsorbent; I ) 0.1 M NaNO3). The respective intrinsic acidity constants are deduced from extrapolation of the three linear portion curves.

FIGURE 3. Adsorption of Cu2+, Zn2+, Cd2+, and Ni2+ ions as a function of pH and for 0.01 M ionic strength (2.5 g‚L-1 adsorbent and 2 × 10-4 M metal concentration).

represented 81% of the total acidity, whereas the very weak acidity, probably of phenolic type from lignin, ferulic acid, and some charged groups from proteins, was less abundant (109 ( 5 µeq‚g-1) (28). Each of these moieties is a potential ligand for metal ions. These results are of great importance because this difference in acidity of the carboxyl groups will certainly induce a difference in metal ion reactivity. Evaluation of the Intrinsic Acidity Constants. The single extrapolation method consists of plotting the variation of pKa versus the surface charge Q (mmol‚g-1) (14, 29). Figure 2 illustrates the plots obtained for 0.1 M ionic strength. Three mean surface acidity constants of pKa1int ) 3.43 ( 0.10, pKa2int ) 6.05 ( 0.05, and pKa3int ) 7.89 ( 0.10 were determined from the acidimetric-alkalimetric titrations performed at 0.1, 0.01, and 0.001 M ionic strengths. The intrinsic acidity constants values did not appear to vary greatly with changes in ionic strength of the solution (32). The pKa1int and pKa2int values correspond to the two distinct carboxylic groups that were identified by Boehm’s titrations. By potentiometric titration, galacturonic acid was previously found to have an intrinsic acidity constant of 3.44 ( 0.02 (27). Concerning the sugar beet pulp, Dronnet (27) found a value of 3.85 that was evaluated with Lifson and Katchalsky theory (35). Last, the acidity constant value pKa3int ) 7.89 ( 0.10 is consistent with that of a phenolic group (22, 24, 36, 37), which was also revealed by Boehm titrations. Hence, the intrinsic acidity constants determined were in close agreement with previously published values. The respective concentrations and acidity constants of the functional groups onto the polysaccharide were first required to model the metal adsorption experiments as a function of pH. Adsorption of Metals Influenced by pH. The sorption of Cu2+, Zn2+, Cd2+, and Ni2+ ions was studied as a function of pH and for different total metal concentrations and ionic strengths. All of the experiments were carried out at maximum pH values of 5.5 in order to avoid metal hydroxide precipitation. It was also verified that all of the metal ions studied were in their divalent ionic form Me2+ in solution up to pH 6 (see Supporting Information). The nitrate complex Me(NO3)+ was always present in a low amount, around 10% of the total metal concentration. Although the interaction of this positive nitrate complex with surfaces sites of the polysaccharide cannot be excluded, it can be concluded that the removal of all metals mainly consisted in adsorption of free divalent cations by surface sites.

TABLE 2. Intrinsic Stability Constants and Surface Complexation Reactions Used for Modeling Sorption of Metals onto Sugar Beet Pulpa

2244

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 10, 2002

log Kint

surface acidity carboxylic groups s≡COOH T s≡COO- + H+ w≡COOH T w≡COO- + H+ phenolic groups ≡COH T ≡CO- + H+

-3.43b -6.05b -7.89b

metal surface complexation s≡COOH

s≡COOMe+

Cu

Ni

Zn

-4.3c

-1.3c

-1.1c

Cd

+ T + -1.1c + Me2+ T w≡COOMe+ + H+ -5.0c -4.8c -4.8c -5.1c ≡COH + Me2+ T ≡COMe+ + H+ 0.1c -1.9c -2.4c -3.1c w≡COOH

Me2+

H+

dissolved species

Cu

Ni

Zn

Cd

Me2+ + OH- T Me(OH)+ 2Me2+ + 2OH- T Me2(OH)22+

6.3d 17.7d

4.1d

5d

3.9d

a (s≡COOH) strong carboxyl groups; (w≡COOH) weak carboxyl groups, (≡COH) phenolic groups. b Calculated from potentiometric titrations. c Stability constants were determined with FITEQL 4.0 (34). d Reference 33, stability constants computed for 0 M ionic strength.

The results concerning the removal percentages of Cu2+, Zn2+, Cd2+, and Ni2+ ions for 2 × 10-4 M metal concentration and 2.5 g‚L-1 beet pulp concentration are reported in Figure 3. The curves present a form commonly observed in previous studies (14-25). For all metal ions, the removal efficiency never reached 100%. Metal removal increased in a narrow pH range (2-3 pH units). By increasing the pH, the sorption edges were reached for Cu2+, Zn2+, Cd2+, and Ni2+ in this sequence, the maximum sorption being reached between pH 5 and 5.5. However, nickel element did not really reach a plateau like the other metals because its adsorption occurred in a wider pH range. According to the surface complexation theory, the increase in metal removal as pH increases can be explained on the basis of a decrease in competition between proton and metal species for the surface sites and by the decrease in positive surface charge (point of zero net proton charge equal to 5, according to ref 32), which results in a lower Coulombic repulsion toward the metal. Among the metal cations studied, copper showed the strongest affinity for the polysaccharide. The preference of

FIGURE 4. Adsorption of (a) Zn2+, (b) Cd2+, and (c) Ni2+ ions as a function of pH and for two different initial concentrations (2.5 g‚L-1 adsorbent). The lines represent modeled sorption curves derived from the stability constants given in Table 2. several adsorbents for metals has been related to the equilibrium constant of the first metal hydrolysis reaction (18, 20, 38). The reported effect is a stronger attraction for the metal showing a higher first hydrolysis equilibrium constant (Table 2). Therefore, the tendency observed for this material was in good agreement with those previously reported for other sorbents such as activated carbon (18), probably because similar functional groups are involved in the metal sorption. Figure 4, parts a-c, represent the respective adsorption of Zn2+, Cd2+, and Ni2+ cations for two different metal concentrations and 0.01 M ionic strength. The efficiency of metal removal was affected by the initial metal concentration (i.e., an improvement of the removal percentages was observed with decreasing initial metal concentration at constant pH). At the same time, the pH-metal removal curves were shifted to more acidic regions. Such a behavior in metal removal could be attributed to an increase in available surface sites with the decrease in the total metal concentration. Adjusting Model Parameters in Single Metal Systems. In the proposed surface complexation model, only monodentate complexes were used to describe the cation binding onto the polysaccharide. As their concentration was negligible (see Supporting Information), the sorption of the nitrate metal complexes onto the polysaccharide was not taken into

account. The assumption of three discrete adsorption sites on the adsorbent was based on the characterization of the surface acidity. The surface site concentration and the acidity constants obtained by potentiometric titrations were used and considered as fixed parameters. The metal complexation constants were fitted for single-metal experiments. The estimated values are listed in Table 2. The calculated results are represented as lines in Figure 4. Regarding the modeling results, it can be concluded that the model was able to describe reasonably well the experimental data. Figure 5 represents the adsorption of Cu2+ ions at 4 × 10-4 M metal concentration and at two ionic strengths, 0.1 and 0.01 M. Through the Gouy-Capman equation (eq 3), the model allowed for the effect of ionic strength on the metal adsorption to be taken into account. It also allowed the pH and metal concentration to vary. This is the principal advantages of the surface complexation models over other empirical models. A comparison of the calculated intrinsic adsorption int constants shows that the difference between the K≡COOMe + int values for all the metals was not as large as that for K≡COMe +. This effect was previously observed for metal adsorption onto humic acids (22). Because of the protonation of the functional int groups, the difference between K≡COMe + may not be significant at low pH. At higher pH, the phenolic adsorption sites become a dominant factor in determining the metal binding VOL. 36, NO. 10, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2245

was confirmed by the high value of the complexation constant with the phenolic groups. The estimated complexation constants for copper and cadmium were compared with values available in the literature. First, our complexation constants were expressed according to the equilibrium ≡COO- + Me2+ T ≡COOMe+, which led to the calculation int of the stability constant log β≡COOMe+ ) log K≡COOMe + - log int + Ka (24). Then, the two distinct log β≡COOMe values were added together for each of the metal ions to obtain a global stability constant relating to the carboxyl moieties of the polysaccharide. Although different electrostatic models were used, the affinity constants were in the same range concerning humic substances, wheat bran, and sugar beet pulp (Table 3). The complexation constants relating to cadmium element were in very close agreement with the reported values. Whatever the model used, the metal cations appear to have a higher affinity for the phenolic sites than the carboxylic moieties. FIGURE 5. Adsorption of Cu2+ ions as a function of pH at two different ionic strengths, 0.01 and 0.1 M (2.5 g‚L-1 adsorbent and metal concentration around 4 × 10-4 M). The lines represent modeled sorption curves derived from the stability constants given in Table 2. sequence. To compare the metal binding strengths, the three adsorption constants have to be lumped together. Hence, the strong affinity of copper element for the polysaccharide

Adsorption in Binary Metal Systems. Before considering a practical application of the sugar beet pulp in a process, competitive adsorption in metal mixtures must be investigated. Therefore equilibrium experiments were carried out to find out the competitive effects in binary Cu-Zn (Figure 6a), Zn-Cd (Figure 6b), Cd-Ni (Figure 6c), and Ni-Zn (Figure 6d) systems at equimolar concentrations (around 2 × 10-4 M metal concentration) and for 0.01 M ionic strength. The pH effect on the metal adsorption in mixtures was similar

FIGURE 6. Competitive adsorption in (a) Cu/Zn, (b) Zn/Cd, (c) Cd/Ni, (d) and Ni/Zn binary metal system at equimolar metal concentrations (around 2 × 10-4 M) and for 0.01 M ionic strength (2.5 g‚L-1 adsorbent). The dashed lines represent the model predictions. 2246

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 10, 2002

TABLE 3. Values of log β≡COOMe+ and log β≡COMe+ Obtained for Copper and Cadmium Elements; Comparison with Literature Values humic substances (22)

references

agricultural byproducts

(39)

(40)

(41)

(24) (wheat bran)

this study (beet pulp)

metal

Cu

Cd

Cu

Cd

Cu

Cd

Cu

Cu

Cu

Cd

log β≡COOMe+ log β≡COMe+

3.60 7.60

3.12 4.60

3.2-5.9 8.4-9.9

3.0-5.1 4.5-4.7

3.4 8.2

2.9 4.9

0.40 6.42

3.58 7.56

0.18 7.97

3.24 4.77

FIGURE 7. Competitive adsorption of Cu2+ and Zn2+ ions at non equimolar metal concentrations and for 0.01 M ionic strength (2.5 g‚L-1 adsorbent). The dashed lines represent the model predictions.

FIGURE 8. Comparison of the predicted and observed metal adsorption capacities (qe in mmol‚g-1 of dried sorbent) in binary metal system.

to that in a single-metal system. The competitive effects of the metals were evaluated by comparing the single adsorption percentage at the plateau (4.5-5 pH range) with that in mixtures and, of course, under close experimental conditions. Figure 7 represents the binary Cu-Zn mixture at nonequimolar concentrations (3.8 × 10-4 M Cu2+ and 7.5 × 10-4 M Zn2+) and for 0.01 M ionic strength. Taking into consideration that the Zn2+ concentration was 2 times higher than that of Cu2+, it can be observed that Cu2+ removal was slightly affected by the presence of Zn2+ ions. When compared with Cu2+ removal in single-metal system and under the same experimental conditions (Figure 5), the efficiency decreased by 10%. Zn2+ removal efficiency was also affected by Cu2+ ions (about 10% decrease) because the latter ions are the more strongly adsorbed to the polysaccharide. The negligible effect of Ni2+ ions on Cd2+ adsorption should be noted (Figure 6c), whereas Zn2+ led to a decrease of about 15% of Cd2+ removal in the 4.5-5 pH range (Figure 6b). Last, Ni2+ adsorption was slightly affected by the competing Zn2+ ions (Figure 6d), whereas it was decreased by 10% with the presence of Cd2+ cations (Figure 6c). In all cases, metal adsorption onto the polysaccharide was found to be favorable in multicomponent systems. Model Predictions. One of the purposes of modeling is the prediction of free metal ion concentrations for various experimental conditions because the direct measurements can be time-consuming. The multispecies metal adsorption equilibrium provide a further test to validate the affinity constants estimated from this study. Consequently, we assessed the ability of the present modeling approach to predict the adsorbed metal percentage for binary metal systems. The calculations were carried out under the same conditions as those used in the experiments. The dashed lines in Figures 6 and 7 have been calculated with the combinations of individual adsorption reactions and their stability constants calculated in single-metal system (Table

2). As shown in Figures 6 and 7, the model described well the competition effects between the various metal ions. Despite the simplicity of the model used, the description of the pH dependency of all competitive metal ions and the percentage of bound metal were reasonably well simulated. The tendency of the model to overestimate the Zn2+ removal for pH above 4.5 in Cu-Zn, Ni-Zn, and Cd-Zn mixtures should be noticed. Figure 8 shows the predictive ability of the model for the adsorption in binary metal systems. Student t statistics prove that the slope of the regression is not significantly different from the perfect agreement for a 95% confidence interval. It can be concluded that a multimetallic adsorption process could be predicted on the basis of a combination of simple single-metal sorption reactions. This study showed that sugar beet pulp might be an interesting material in the treatment of metal-contaminated water. The data sets for metal sorption onto the polysaccharide were interpreted in the framework of a unifying modeling approach, which resulted in the calculation of the metal complexation constants. Moreover, the heterogeneous properties of the polysaccharide surface were taken into consideration because the phenolic and carboxylic groups were assumed to be the adsorption sites. The simple model used allowed the experimental parameters such as metal concentration, ionic strength and solution pH to vary. Further studies integrating other metals and more competitive metal binding effects could allow the proposed model to be tested under wider experimental conditions. Last, all of these data will be useful for the development and the optimization of a new process involving this low-cost adsorbent.

Acknowledgments The authors gratefully acknowledge Dr. Catherine FaurBrasquet from the Mining School of Nantes (France) for valuable discussions. VOL. 36, NO. 10, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2247

Supporting Information Available The speciation diagrams of the soluble metallic species were calculated under the experimental conditions used in this work. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Ricou, P.; Lecuyer, I.; Le Cloirec, P. Water Res. 2001, 35, 965976. (2) He´quet, V.; Ricou, P.; Lecuyer, I.; Le Cloirec, P. Fuel 2001, 80, 851-856. (3) Brown, P. A.; Gill, S. A.; Allen, S. J. Water Res. 2000, 34, 39073916. (4) Ho, Y. S.; McKay, G. Water Res. 2000, 34, 735-742. (5) Texier, A.-C.; Andre`s, Y.; Le Cloirec, P. Environ. Sci. Technol. 1999, 33, 489-495. (6) Laszlo, J. A.; Dintzis, F. R. J. Appl. Polym. Sci. 1994, 52, 531-538. (7) Wartelle, L. H.; Marshall, W. E. Adv. Environ. Res. 2000, 4, 1-7. (8) Dronnet, V. M.; Renard, C. M. G. C.; Axelos, M. A. V.; Thibault, J.-F. Carbohydr. Polym. 1997, 34, 73-82. (9) Ge´rente, C.; Couespel du Mesnil, P.; Andre`s, Y.; Thibault, J.-F.; Le Cloirec, P. React. Funct. Polym. 2000, 46, 135-144. (10) Farley, K. J.; Dzombak, D. A.; Morel, F. M. M. J. Colloid Interface Sci. 1985, 106, 226-242. (11) Hayes, K. F.; Leckie, J. O. J. Colloid Interface Sci. 1987, 115, 564-572. (12) Dzombak, D. A.; Morel, F. M. M. Surface Complexation Modeling; Hydrous ferric oxide; Wiley: New York, 1990. (13) Hayes, K. F.; Redden, G.; Ela, W.; Leckie, J. O. J. Colloid Interface Sci. 1991, 142, 448-469. (14) Stumm, W.; Morgan, J. J. Aquatic chemistry, 3rd ed.; John Wiley & Sons: New York, 1996. (15) Yiacoumi, S.; Tien, C. Kinetics of Metal Ion Adsorption from Aqueous Solutions. Models, Algorithms, and Applications; Kluwer Academic Publishers: London, U.K., 1995; Chapter 3. (16) Tiffreau, C.; Lu ¨ tzenkirchen, J.; Behra, P. J. Colloid Interface Sci. 1995, 172, 82-93. (17) Bonnissel-Gissinger, P.; Alnot, M.; Lickes, J.-P.; Ehrhardt, J.-J.; Behra, P. J. Colloid Interface Sci. 1999, 215, 313-322. (18) Corapcioglu, M. O.; Huang, C. P. Water Res. 1987, 21, 10311044. (19) Gabaldon, C.; Marzal, P.; Ferrer, J.; Seco, A. Water Res. 1996, 30, 3050-3060. (20) Seco, A.; Marzal, P.; Gabaldon, C.; Ferrer, J. Sep. Sci. Technol. 1999, 34, 1577-1593.

2248

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 10, 2002

(21) Chen, J. P.; Lin, M. Water Res. 2001, 35, 2385-2394. (22) Liu, A.; Gonzalez, R. D. Langmuir 2000, 16, 3902-3909. (23) Fein, J. B.; Daughney C. J.; Yee, N.; Davis, T. A. Geochim. Cosmochim. Acta 1997, 61, 3319-3328. (24) Ravat, C.; Dumonceau, J.; Monteil-Rivera, F. Water Res. 2000, 34, 1327-1339. (25) Lee, S. M.; Davis, A. P. Water Res. 2001, 35, 534-540. (26) Michel, F.; Thibault, J.-F.; Barry, J.-L.; De Baynast, R. J. Sci. Food Agric. 1988, 42, 77-85. (27) Dronnet, V. Ph.D. Dissertation, University of Nantes, 1996. (28) Reddad, Z.; Ge´rente, C.; Andre`s, Y.; Ralet, M.-C.; Thibault, J.-F.; Le Cloirec, P. Carbohydr. Polym. 2002, 49, 23-31. (29) Hohl, H.; Stumm, W. J. Colloid Interface Sci. 1976, 2, 281-288. (30) Boehm, H.-P. Adv. Catal. 1966, 16, 179-225. (31) Bertin, C.; Rouau, X.; Thibault, J.-F. J. Sci. Food Agric. 1988, 44, 15-29. (32) Ge´rente, C.; Reddad, Z.; Andre`s, Y.; Le Cloirec, P. Water Sci. Technol. 2002. (33) Smith, R. M.; Martell, A. E. Critical Stability Constants, 4, Inorganic complexes; Plenum press: New York, 1976. (34) Herbelin, A. L.; Westall, J. C. FITEQL. A Computer Program for Determination of Chemical Equilibrium Constants from Experimental Data. Version 4.0; Department of Chemistry, Oregon State University: OR, 1999. (35) Lifson, S.; Katchalsky, A. J. Polym. Sci. 1954, 13, 43-55. (36) De Wit, J. C. M.; Van Riemsdijk, W. H.; Koopal, L. K. Environ. Sci. Technol. 1993, 27, 2015. (37) Milne, C. J.; Kinniburgh, D. G.; De Wit, J. C. M.; Van Riemsdijk, W. H.; Koopal, L. K. Geochim. Cosmochim. Acta 1995, 59, 11011112. (38) Martell, A. E.; Hancock, R. D. Metal Complexes in Aqueous Solutions; Plenum: New York, 1996. (39) Van Riemsdijk, W. H.; De Wit, J. C. M.; Mous, S. L. J.; Koopal, L. K.; Kinniburgh, D. G. J. Colloid Interface Sci. 1996, 183, 35. (40) Benedetti, M. F.; Milne, C. J.; Kinniburgh, D. G.; Van Riemsdijk, W. H.; Koopal, L. K. Environ. Sci. Technol. 1995, 29, 446. (41) Kinniburgh, D. G.; Milne, C. J.; Benedetti, M. F.; Pinheiro, J. P.; Filius, J.; Koopal, L. K.; Van Riemsdijk, W. H. Environ. Sci. Technol. 1996, 30, 1687.

Received for review September 12, 2001. Revised manuscript received February 6, 2002. Accepted February 20, 2002. ES010237A