Adsorption of Pb (II) ions on variable charge oxidic calcined substrates with chemically modified surface

Oxidic lithological material

The raw oxidic lithological material (OLM) was collected from a natural deposit, a mine, 5 kg were extracted manually with the help of pickaxes and shovels in the coodinates 8^o 28´47´´ N and 71^o 23´47´´ W. This zone is an arid type, where the temperature varies between 17 and 30 ^oC along the year, and maximum rainfall of 200 mm per year. The OLM has already been chemically characterized,^19–21 consisting mainly of Fe, Al as well as Ti and Mn which can form refratory amphoteric oxides. Due to its thermal resistence this lithologic material is used by potters for preparing bricks by thermal treatment.

Preparation of the adsorbent substrate and activation of negative surface charges

The lithologic material was crushing using a rubber hammer so as not to destroy the mineral structures, and sieved for 5 minutes, using Octagon 200CL Digital Sieve Shaker (Endecotts Ltd, England) to obtain particle-size fractions of 800 μm, then mixed up with distillated water in order to obtain a homogeneous saturated paste with an easily moldable consistency. With the use of a 60 mL syringe, 3 mm diameter “spaghettis” are extruded and cut out into 5 mm long pieces, are dried in ambient air for 24 hours, then in a furnace at 150 ^oC another 24 hours (FELISA FE-293, Jalisco, Mexico). Finally, the dried solid adsorber is calcined up to 750 ^oC at the rate of 6.25 ^oC min^-1, across four hours (Thermolyne FB141OM muffle, Thermo Scientific, Waltham, USA) and then cooled down for 12 hours, until 20 ^oC. The calcination process favors the oxides formation and cementation of the pellets to avoid dispersion in the aqueous solution. Also, during the calcination process, organic matter fraction, which is very small, is complete burned out, so only the oxidic phase participates in the adsorption reaction.

The calcined substrate is chemically treated in alkaline media with 0.1 N NaOH solution for 12 hours in a flask at room temperature, in order to enlarge the negative charge density on the surface because of the oxide deprotonation reaction. The product is then called activated substrate. (The non-treated substrate is labeled with the name of non-activated substrate). After this time the substrate is washed out with distilled water until neutral pH is achieved, then dried in the furnace at 120 ^oC for 12 hours.

Adsorption studies

The adsorption study was performed in triplicate in isothermal conditions (20 ± 2 ^oC) using batch equilibration procedure, treating 2 g of activated and non-activated calcined substrate, with increasing volume of 5, 10, 15, 20, 25, 30 and 40 mL of 0.001 M Pb⁺² solution, prepared from CH₃COOPb analytical grade Merck reagent, during 24 hours. in 100 mL glass beakers. Suspensions were periodically shaken every hour. Differences between Ci and C_eq were assumed to be due to adsorption. Adsorption isotherms were obtained by plotting the amount of lead adsorbed (q_e), against the Pb⁺² equilibrium concentration (C_eq) and fitted to the linear forms of the Freundlich and Langmuir equations.^21,30–33

Freundlich Isotherm is an empirical model which assumes an adsorption process characterized by multy layer adsorption on heterogeneous surfaces. The model is described by the equation 1 and the linear form by equation 2.^21,30,31 A graph representation of log (q_e) vs log (C_eq) is a straight line, with slope equal to n and intercept equal to log (K_F):

In equations (1) and (2), q_e is the amount adsorbed per adsorbent weight unit, C_eq is the equilibrium concentration of adsorbate in solution after adsorption, K_F is the Freundlich constant related to adsorption capacity, and n is a constant related to adsorption intensity or energetic homogeneity of active sites of adsorption. n may take values near unity or greater. The lower the value of n is, the lower the energy heterogeneity in the active adsorption sites.

The Langmuir model describes a reversible process with the formation of adsorbate monolayers on the adsorbent surface. The nonlinear and linear forms of the Langmuir isotherm are described in Equations (3) and (4)^21,22,33:

In equations (3) and (4), q_m is the adsorption capacity, which represents the maximum amount of adsorbate in a monolayer, and K_L is the Langmuir constant related to the affinity between adsorbate and adsorbent. The parameters of the isotherm are obtained by plotting 1/q_e versus 1/C_eq, resulting in a straight line with slope equal to 1/K_L*q_m and the intersection equal to 1/q_m.

The best fit between the isotherm function and the experimental data was verified by carrying out linear regressions of the isotherm linear equation. In addition, the adjustment of each isotherm was verified by comparing the experimental value obtained from the amount adsorbed (q_e exp) with the value obtained by calculating it by means of the isotherm equation (q_e calc) and determined the linear correlation between the two values given by the linear regression coefficient (r).

Lead analysis

Pb⁺² analyses were performed with a Varian Graphite Furnace Atomic Absorption Spectrophotometer, Spectra AA Zeeman 220, with a pyrolytic coated graphite tube (Palo Alto, California, USA). The spectrophotometer is dotted with a Varian auto sampler model EL-97113008. A Varian Uranium hollow cathode lamp was used.

pH and electric conductivity studies

Solution pH and electric conductivity (EC) variations were studied, in triplicate, using the same isothermal batch equilibration procedure, as in adsorption study. pH was measured with a HI 211 pHmeter (HANNA instruments, Smithfield, Rhode Island, USA), calibrated with commercial buffer solutions of pH 4 and 7. EC was measured with a Trans Instrument HC3010 Conductimeter (Petro Centre, Bukit Merah, Singapure), calibrated with standard reference. All experimental data³⁴ were worked out with the Excel software (Microsoft, Los Angeles, USA)

Isotherm graph

Figure 1a shows the adsorption isotherms of Pb⁺² ions on the activated and non-activated substrates. There are two relevant aspects to consider. First of all, the isotherm profiles show but don´t confirm information which kind of interaction between Pb⁺² ions and substrate surface may take place. In addition, the system under study coincides with a L-type isotherm, that is, at low equilibrium concentrations the amount adsorbed increases rapidly, while at high concentrations it decreases, due to the fewer adsorption sites available on the adsorbent].

Figure 1. a: Isotherms for Pb⁺² ions adsorption and b: fitting to the logarithmic form of the Freundlich model for the activated and non-activated substrates.

Figure 1b shows linear fitting of adsorption data to the Freundlich model for the activated and non-activated substrates, while the Table 1 shows the fitted equations according to the Freundlich model, r, K_F, and n values for activated and non-activated substrates, by triplicated trials. In both cases data present relatively good linearity, r being more favorable for the activated substrate, the negative charge density on the surface is likely to be more homogeny than in non-activate substrate.

Freundlich parameters are different for the activated substrate compared with the adsorption on the non-activated substrate. Average value of K_F, for activated substrate is around 3.7 times greater (1791.73 ± 13.06) with a relative standard deviation (RSD) of 0.72 %, compared to the respective average K_F value for the non-activated substrate (491.54 ± 31.97 with an RSD of 6.5 %). Similarly, the n values suggest greater adsorption intensity on the activated substrate in relation to the non-activated substrate. The smallest negative charge density on the non-activated substrate surface is a limiting factor for the adsorption of Pb⁺² ion.

However, to prove adjustment to the model it is also necessary to show how good experimental data fit to the theoretic model. Figure 2 shows the graphical correlation between calculated and experimental q_e data for the activated and non-activated substrates. Table 2 shows fitted equations for the linear functions. There is good linear correlation showing acceptable adjustment with the Freundlich model. However, calculated values of q_e are biased from the experimental data. The average difference is about 15.16 ± 6.63 % in the case of the activated substrate and 17.41 ± 18.15 % for the non-activated substrate.

Figure 2. Correlations between calculated and experimental values of q_e (Freundlich model) for the activated and non-activated substrates.

Figure 3 shows adsorption data fitting according to the Langmuir model and Table 3 shows fitted equations and values of r, k₁ and k₂ values for the activated and non-activated substrates, by triplicated. The experimental data shows great reproducibility, correlation coefficients are more satisfactory for the activated substrate in relation with the non-activated substrate. The K₁ value for the activated substrate is about 20 times greater than for the non-activated substrate.

Figure 3. Fitting isotherms to the linear form of the Langmuir equation for the activated and non-activated substrates.

Figure 4 show the graphical correlations between calculated and experimental values of q_e for the activated and non-activated substrates and Table 4 shows fitted equations for the linear functions. There is also good linear correlation between calculated and experimental data however, calculated values of q_e are even more biased from the experimental data compared with the Freundlich model. The average difference is 25.29 ± 23.67 % in the case of the activated substrate and 27.22 ± 16.02 % for the non-activated substrate.

Figure 4. Correlations between calculated value of C_ad vs experimental value (Langmuir model) for the activated and non-activated substrates.

pH and electric conductivity study

Figure 5 shows pH variation during adsorption reaction of Pb⁺² ions on raw material, non-activated and activates substrates, by triplicate, from a 1 mM Pb⁺² ions solution. In all cases adsorption reaction take place with production of protons and solution acidification. Experiments are highly reproducible, so reaction takes place through the same mechanism in all the replicates.

Figure 5. pH variation, by triplicate, as a function of mmol of Pb⁺² added to 2 g of substrate from a 1 mM Pb⁺² solution, for raw material, and activated and non-activated substrate.

Figure 6 shows pH variation, by triplicate, as a function of mmol of Pb⁺² added to 2 g of substrate from a 1 mM Pb⁺² solution, for raw material, activated and non-activated substrate. On the calcined substrates, the reaction takes place at lower pH values, so there are more active sites for the adsorption reaction to occur, with a higher production of protons.

Figure 6. pH variation, by triplicate, as a function of mmol of Pb⁺² added to 2 g of substrate from a 1 mM Pb⁺² solution, for raw material, activated and non-activated substrate.

Table 5 shows initial and final concentrations of H⁺ ions during the adsorption reaction of Pb⁺² ions and the net amount of mmol of H⁺ ions produced in the whole reaction, when a 1 mM Pb⁺² solution is used. Adsorption reaction on raw material produces 0.29 μmol of H⁺ and 0.26 μmol on non-activated substrate. However, adsorption reaction on activated substrate produces 0.35 μmol of H⁺ ion, which is 1.4 times greater than in the case of non-activated substrate.

Figure 7 shows EC, variation, by triplicate, during adsorption reaction of Pb⁺² ions on raw material and non-activated and activates substrates, when a 1 mM Pb⁺² solution is used for the reaction of adsorption. EC in the solution is basically produced by ions in the solution, i.e., CH₃COO^- and Pb⁺² ions, as well as H⁺ ions produced in the adsorption reaction. As EC decreases, it suggests that Pb⁺² ions are adsorbed on the substrates surfaces, as well as on the raw material. The adsorption reaction produces ion immobilization being unable to make a net contribution to the EC of the solution.

Figure 7. EC variation, by triplicate, as a function of mmol of Pb⁺² added to 2 g of substrate from a solution 0.001 M Pb⁺², for crud material, activated and non-activated substrate.

Figure 8 shows a comparative view of the EC of the solutions for all three cases when a 1 mM Pb⁺² solution is used. Although conductivity decrease due to adsorption of Pb⁺² ions, solution in contact with activated substrate present greater conductivity due to the highest production of H⁺ ions during the adsorption reaction. These H⁺ ions have a net contribution to the EC of the solution producing an increasing of the conductivity in the solution which is in contact with the substrates.

Figure 8. EC variation, by triplicate, as a function of mmol of Pb⁺² added to 2 g of substrate from a solution 1 mM Pb⁺², for crud material, activated and non-activated substrate.

Figure 9 shows pH variation, by triplicate, during adsorption reaction of Pb⁺² ions on raw material, non-activated and activates substrates, from a 10 mM of Pb⁺² ions solution. As in the former cases, the reaction also produces acidification of the solution in contact with the substrates due to the production of H⁺ ions, which also increases as the concentration of Pb⁺² ions increases.

Figure 9. pH variation, by triplicate, as a function of mmol of Pb⁺² added to 2 g of substrate from a 10 mM Pb⁺² solution, for raw material, and activated and non-activated substrate.

Figure 10 shows comparative pH variations during adsorption reaction on the raw material as well an on the activated ad non-activated substrates when a 10 mM of Pb⁺² solution is used. As it can be observed, the acidification process is more accentuated in the case of the adsorption reaction on the activated substrate, as is expected. The similarity of the curves is indicative that the experiments are highly reproducible as in the previous case (Figure 9), so reaction takes place through the same mechanism in all the replicates.

Figure 10. pH variation as a function of mmol of Pb⁺² added to 2 g of substrate from a solution 10 mM Pb⁺², for crud material, activated and non-activated substrate.

Table 6 shows initial and final concentrations of H⁺ ions during the adsorption reaction of Pb⁺² ions and the net amount of mmol of H⁺ ions produced in the whole reaction, when a 10 mM Pb⁺² solution is used. Adsorption reaction on raw material produces 23.3 μmol of H⁺ ion and 15.6 μmol of H⁺ on non-activated substrate. However, adsorption reaction on activated substrate produces 72.2 μmol of H⁺ ion which is 4.6 times greater than in the case of non-activates substrate and 3 times greater than in the case of adsorption reaction on crude material. Moreover, this amount of H⁺ ions is almost 200 times greater than in the former case when a 1 mM Pb⁺² solution is used for adsorption reaction.

Figures 11 and 12 shows EC, variation by triplicate, during adsorption reaction of Pb⁺² ions on raw material and non-activated and activates substrates, when a 10 mM Pb⁺² is used. As it was pointed out in the former case, EC in the solutions decreases, because Pb⁺² ions are adsorbed on the substrate’s surfaces, as well as on the crude material. Nevertheless, the decreasing of the EC in the solutions in contact with the crude material and non-activated substrate is greater due to the minor production of H⁺ ion during the adsorption reaction, while the EC is greater in the solution in contact with the activated substrate.

Figure 11. EC variation as a function of mmol of Pb⁺² added to 2 g of substrate from a solution 10 mM Pb⁺², for crud material, activated and non-activated substrate.

Figure 12. EC variation as a function of mmol of Pb⁺² added to 2 g of substrate from a 10 mM Pb⁺² solution, for raw material, activated and non-activated substrate.

Ethical considerations

Not applicable.