Alkaline Modification of Arabica-Coffee and Theobroma-Cocoa Agroindustrial Waste for Effective Removal of Pb(II) from Aqueous Solutions

Arabica-coffee and Theobroma-cocoa agroindustrial wastes were treated with NaOH and characterized to efficiently remove Pb(II) from the aqueous media. The maximum Pb(II) adsorption capacities, qmax, of Arabica-coffee (WCAM) and Theobroma-cocoa (WCTM) biosorbents (qmax = 303.0 and 223.1 mg·g−1, respectively) were almost twice that of the corresponding untreated wastes and were higher than those of other similar agro-industrial biosorbents reported in the literature. Structural, chemical, and morphological characterization were performed by FT-IR, SEM/EDX, and point of zero charge (pHPZC) measurements. Both the WCAM and WCTM biosorbents showed typical uneven and rough cracked surfaces including the OH, C=O, COH, and C-O-C functional adsorbing groups. The optimal Pb(II) adsorption, reaching a high removal efficiency %R (>90%), occurred at a pH between 4 and 5 with a biosorbent dose of 2 g·L−1. The experimental data for Pb(II) adsorption on WACM and WCTM were well fitted with the Langmuir-isotherm and pseudo-second order kinetic models. These indicated that Pb(II) adsorption is a chemisorption process with the presence of a monolayer mechanism. In addition, the deduced thermodynamic parameters showed the endothermic (ΔH0 > 0), feasible, and spontaneous (ΔG0 < 0) nature of the adsorption processes studied.


Introduction
Effluents from industrial activities such as smelting, mining, painting, tanning, etc. are causing severe environmental pollution by depositing heavy metals, particularly in aquatic ecosystems [1]. These metals are highly toxic, are not degradable, and can accumulate in living organisms and affect many of their vital functions. Lead (Pb) is the second most toxic metal and can adversely affect the nervous, digestive, and reproductive systems and can even cause death [2,3]. For these reasons and for preventive purposes, for example, the World Health Organization (WHO) has established the permissible limit of Pb in drinking water at 0.01 mg L −1 [4].
Various methods are used to remove heavy metals, such as Pb, from wastewater: coagulation-flocculation, liquid-liquid extraction, ion exchange, and electrochemical treatment. However, these methods have disadvantages such as the high operating costs, long operating times, and the generation of a large volume of toxic sludge [5,6]. In this context, the removal of contaminants using biological materials (biosorbents), such as algae, cyanobacteria, fungi, or particularly agroindustrial wastes has become an economical, ecological, and promising alternative method compared with the conventional methods mentioned above [7][8][9][10]. These biosorbents can be modified to improve the adsorption capacity, structural stability, and reusability. The modification can be carried out by chemical reagents (chemical modification), physical calcination, or grinding methods [11]. In particular, Calero et al. [12], Moyo et al. [13], Petrović et al. [14], and Ye and Yu [15], among others, reported that chemical modification, with NaOH, of biosorbent-precursors improved the removal capacity of Pb.
Peru is an important producer of coffee and cocoa worldwide, with annual productions of 136 and 218 Kt, respectively [16]. The processing of coffee cherries and cocoa pods generates wastes at approximately 80% (coffee) [17] and 70% (cocoa) [18,19] of the total weight of the product, constituting a serious environmental problem for their producing regions [20]. However, these agro-industrial residues could be used for the effective cleaning of aqueous ecosystems of the surrounding crops, in which we have evidence that there is contamination with heavy metals such as Pb.
In this work, we significantly improved the absorption capacity of Pb(II) by means of alkaline modification (with NaOH) to coffee-and cocoa-untreated wastes [21].

Effect of Alkaline Treatment
After treatment with NaOH, both the arabica-coffee and theobroma-cocoa biosorbents lost weight, 40.2% and 38.4%, respectively (Table 1). This loss may be due to the fact that, during NaOH treatment, the hydrolysis reactions that take place would cause a high dissolution of organic compounds from the biomass and, therefore, its considerable disintegration [12,22]. The basic titrable sites on both WACM-and WTCM-treated biosorbents were almost 1.4 times higher than that in the respective untreated precursors WAC and WTC (see Table 1). According to Santos et al. [22], and Bulgariu and Bulgariu [23], among others, the NaOH treatment provides, due to the hydrolysis reactions, the formation of more carboxylic (-COO − ) and hydroxyl (-OH) groups (basic titrable sites), both in undissociated as dissociated forms, that improve the Pb-binding properties of WACM and WTCM biosorbents. It is interesting to mention the absence of acid-titrable sites on the surface of each treated biomass (See Table 1).
The point of zero charge, pH PZC , values for WACM and WTCM were higher than those for WAC and WTC, respectively (Table 1). This result indicates an increase in the surface basicity of the treated biosorbents, which is consistent with the increase in the concentration of its basic titrable sites, described above. A similar feature was reported by Blázquez et al. [24] for olive stone biomass modified with NaOH.
An interesting consequence of the alkaline treatment of the studied biomasses is the considerable increase in the Pb(II) removal capacity. Thus, taking into account optimal conditions, described below (Section 1.3), the Pb(II) adsorption capacity q e of WACM and WTCM were almost three times higher than of corresponding non-treated WAC and WTC biomasses (see Figure 1). Mangwandi et al. [6], Ren et al. [11], and Gupta et al. [25], among others, reported that chemical modification of a biomass considerably improves its adsorption capacity of heavy metals.
Molecules 2023, 28, x FOR PEER REVIEW 3 of 16 conditions, described below (Section 2.3), the Pb(II) adsorption capacity qe of WACM and WTCM were almost three times higher than of corresponding non-treated WAC and WTC biomasses (see Figure 1). Mangwandi et al. [6], Ren et al. [11], and Gupta et al. [25], among others, reported that chemical modification of a biomass considerably improves its adsorption capacity of heavy metals.

SEM/EDX and FTIR Analysis
SEM micrographs of the WACM and WTCM biosorbents before and after Pb(II) was loaded are shown in Figure 2. Typical uneven and rough surface morphologies are observed. Before adsorption, the images show more porous and less compact structures, than that with Pb(II) loaded. The results of the EDX spectra ( Figure 3) confirmed that Pb(II) is adsorbed on the surface of both Pb-loaded biosorbents. Furthermore, the disappearance of the Na peak on these samples (Figure 3 right) can be attributed to ionic exchange with Pb(II). Similar morphologies were reported by Jaihan et al. [5] in papaya peels loaded with Pb(II).

SEM/EDX and FTIR Analysis
SEM micrographs of the WACM and WTCM biosorbents before and after Pb(II) was loaded are shown in Figure 2. Typical uneven and rough surface morphologies are observed. Before adsorption, the images show more porous and less compact structures, than that with Pb(II) loaded. The results of the EDX spectra ( Figure 3) confirmed that Pb(II) is adsorbed on the surface of both Pb-loaded biosorbents. Furthermore, the disappearance of the Na peak on these samples (Figure 3 right) can be attributed to ionic exchange with Pb(II). Similar morphologies were reported by Jaihan et al. [5] in papaya peels loaded with Pb(II).      Figure 4 shows the FTIR spectra of WACM and WTCM before and after Pb (II) sorption. The FTIR spectra of the unloaded-Pb samples show the positions of the peaks and absorption bands (in parentheses for WTCM) at the following: 1) 3339.3 (3335.7) cm −1 , assignable to typical OH bond stretching vibrations in samples such as cellulose, lignin, or water [26][27][28][29]; 2) 2919.9 (2910.9) cm −1 , assignable to the symmetric stretching of the CH bonds of aliphatic acids [30]; 3) 1636.2 (1621.6) cm −1 assignable to the asymmetric stretching of the double bond of C=O carbonyl groups [29]; 4) 1420.4 (1420.4) cm −1 , assignable to the stretching COH and C=O groups of carboxylates [29,31]; 5) 1019.1 (1028.8) cm −1 , characteristic of COC stretching in polysaccharides [32]; The FTIR spectra after Pb(II) was loaded show changes in the intensity and position of some peaks and bands with respect to those of the clean samples. Thus, for both biosorbents, the positions of the peaks or bands 1, 3, and 4 are displaced with respect to the clean sample values at 1 = −13.3 (−9.7), 3 = −4.9 (−23.2), 4 = −8.5 (−3.6) cm −1 . These results indicate that the OH, C=O, and CO groups would be involved in the biosorption of Pb (II). A similar behavior was reported, among others, by Barka et al. [33] and Mahyoob et al. [29] in the removal of Pb(II) by biomasses such as cladodes of prickly pear or olive tree leaves.

Influence of pH Solution
The pH plays an important role in the adsorption process and provides necessary information on the adsorption-desorption mechanisms. The effect of pH was studied in the range of 2 to 5 ( Figure 5) since Pb precipitates, at pH > 5, into Pb(OH)2 [27]. The Pb(II) adsorption capacity qe, was very low for an acidic medium close to pH 2. It is due to a competing effect between H + and Pb(II) ions for fill surface active sites [30,34]. For pH > 2, qe increased greatly, reaching values of 227.1 and 214.3 mg g −1 at pH 4 for WACM and WTCM, respectively. With a further increase at pH 5, the qe values showed improvements by almost 17% for WACM and by only 2% for WTCM. Accordingly, when pH increases, the repulsive interactions between H + and Pb(II) ions decrease, facilitating access of Pb(II) to the surface adsorption sites. (1) 3339.3 (3335.7) cm −1 , assignable to typical -OH bond stretching vibrations in samples such as cellulose, lignin, or water [26][27][28][29]; (2) 2919.9 (2910.9) cm −1 , assignable to the symmetric stretching of the C-H bonds of aliphatic acids [30]; (3) 1636.2 (1621.6) cm −1 assignable to the asymmetric stretching of the double bond of C=O carbonyl groups [29]; (4) 1420.4 (1420.4) cm −1 , assignable to the stretching C-OH and C=O groups of carboxylates [29,31]; (5) 1019.1 (1028.8) cm −1 , characteristic of C-O-C stretching in polysaccharides [32]; Molecules 2023, 28, 683

of 15
The FTIR spectra after Pb(II) was loaded show changes in the intensity and position of some peaks and bands with respect to those of the clean samples. Thus, for both biosorbents, the positions of the peaks or bands 1, 3, and 4 are displaced with respect to the clean sample values at ∆ 1 = −13.3 (−9.7), ∆ 3 = −4.9 (−23.2), ∆ 4 = −8.5 (−3.6) cm −1 . These results indicate that the OH, C=O, and C-O groups would be involved in the biosorption of Pb(II). A similar behavior was reported, among others, by Barka et al. [33] and Mahyoob et al. [29] in the removal of Pb(II) by biomasses such as cladodes of prickly pear or olive tree leaves.

Influence of pH Solution
The pH plays an important role in the adsorption process and provides necessary information on the adsorption-desorption mechanisms. The effect of pH was studied in the range of 2 to 5 ( Figure 5) since Pb precipitates, at pH > 5, into Pb(OH) 2 [27]. The Pb(II) adsorption capacity q e , was very low for an acidic medium close to pH 2. It is due to a competing effect between H + and Pb(II) ions for fill surface active sites [30,34]. For pH > 2, q e increased greatly, reaching values of 227.1 and 214.3 mg g −1 at pH 4 for WACM and WTCM, respectively. With a further increase at pH 5, the q e values showed improvements by almost 17% for WACM and by only 2% for WTCM. Accordingly, when pH increases, the repulsive interactions between H + and Pb(II) ions decrease, facilitating access of Pb(II) to the surface adsorption sites.
olecules 2023, 28, x FOR PEER REVIEW Figure 5. Influence of pH on the Pb(II) adsorption capacity, qe, for T = 293 K, adsorptio min, biosorbent dose = 2 g L 1 , C0 = 130.8 mg L 1 . Figure 6 depicts the Pb(II) removal efficiency, %R, as a function of the biom A significant increase in %R is observed for both the WACM and WTCM b reaching almost 55% and 60%, respectively, for a biomass dosage of 2 g L − increase in the dosage produces a slight increase in %R for WTCM but a rap for WACM. The latter trend is due to the agglomeration of the WACM biomas during the experimentation, which would reduce the effective surface area a the interaction between Pb(II) and the biosorbent [35].

Influence of Biomass Dosage
For both the WACM and WTCM biosorbents, the dose of 2 g L −1 was sel optimal value, which would provide a suitable surface area for the efficient Pb(II).  Figure 6 depicts the Pb(II) removal efficiency, %R, as a function of the biomass dosage. A significant increase in %R is observed for both the WACM and WTCM biosorbents, reaching almost 55% and 60%, respectively, for a biomass dosage of 2 g L −1 . A further increase in the dosage produces a slight increase in %R for WTCM but a rapid decrease for WACM. The latter trend is due to the agglomeration of the WACM biomass, observed during the experimentation, which would reduce the effective surface area available for the interaction between Pb(II) and the biosorbent [35].

Influence of Biomass Dosage
For both the WACM and WTCM biosorbents, the dose of 2 g L −1 was selected as the optimal value, which would provide a suitable surface area for the efficient removal of Pb(II). during the experimentation, which would reduce the effective surface area availa the interaction between Pb(II) and the biosorbent [35].
For both the WACM and WTCM biosorbents, the dose of 2 g L −1 was selected optimal value, which would provide a suitable surface area for the efficient rem Pb(II). The effect of the initial Pb(II) ion concentration C0 on the adsorption capacity removal efficiency %R was studied in the range of 5.6 to 130.8 mg L −1 (See Figure both the WACM and WTCM biosorbents, the highest %R values (> 92%) were ob for low C0 concentrations (in the range of 5.6 to 40 mg L −1 ), where the ratio of surfac sites to the free Pb(II) ions is high, resulting in rapid adsorption [36]. For high C0 c The effect of the initial Pb(II) ion concentration C 0 on the adsorption capacity q e and removal efficiency %R was studied in the range of 5.6 to 130.8 mg L −1 (See Figure 7). For both the WACM and WTCM biosorbents, the highest %R values (>92%) were obtained for low C 0 concentrations (in the range of 5.6 to 40 mg L −1 ), where the ratio of surface active sites to the free Pb(II) ions is high, resulting in rapid adsorption [36]. For high C 0 concentrations, %R decreased until almost 60% was reached at C 0 = 130.8 mg L −1 . The diminishing %R with increasing C 0 is attributed to the saturation of the available adsorption sites [32,36].
Molecules 2023, 28, x FOR PEER REVIEW trations, %R decreased until almost 60% was reached at C0 = 130.8 mg L −1 . The dim ing %R with increasing C0 is attributed to the saturation of the available adsorptio [32,36].
On the other hand, the Pb(II) adsorption capacity qe of both the WACM and biosorbents showed an opposite trend to %R, since it increases with increasing C result can be explained considering that for a given amount of adsorbent, an incr the amount of Pb(II) ions, in solution, produces a concentration gradient that d greater interaction with the active binding sites of the biosorbent [37,38].

Adsorption Isotherms
The adsorption isotherms were studied in a range of initial Pb(II) concentrat between 5.6 and 130.8 mg L −1 , at pH 4 and pH 5, and T = 293 K and t = 120 min. The are depicted in Figure 8, where Pb(II) adsorption capacity qe vs. Ce concentration o in equilibrium is presented. These data were fitted to two very common isotherm [39,40]: i) Langmuir model, which assumes solute sorption in monolayers with a geneous sorption energy; ii) Freundlich model, which assumes multilayer sorptio heterogeneous sorption energies. The adjustment parameters with both models w On the other hand, the Pb(II) adsorption capacity q e of both the WACM and WTCM biosorbents showed an opposite trend to %R, since it increases with increasing C 0 . This result can be explained considering that for a given amount of adsorbent, an increase in the amount of Pb(II) ions, in solution, produces a concentration gradient that drives a greater interaction with the active binding sites of the biosorbent [37,38].

Adsorption Isotherms
The adsorption isotherms were studied in a range of initial Pb(II) concentrations C 0 between 5.6 and 130.8 mg L −1 , at pH 4 and pH 5, and T = 293 K and t = 120 min. The results are depicted in Figure 8, where Pb(II) adsorption capacity q e vs. C e concentration of Pb(II) in equilibrium is presented. These data were fitted to two very common isotherm models [39,40]: (i) Langmuir model, which assumes solute sorption in monolayers with a homogeneous sorption energy; (ii) Freundlich model, which assumes multilayer sorption, with heterogeneous sorption energies. The adjustment parameters with both models were obtained by fitting the corresponding experimental data [C e /q e vs. C e ] and [log(q e ) vs. log(C e )]. Figure 7. Effect of the initial Pb(II) concentration C0 on qe (dotted lines) and %R (full l min, T = 293 K, pH 4, biosorbent dose = 2 g L −1 .

Adsorption Isotherms
The adsorption isotherms were studied in a range of initial Pb(II) concen between 5.6 and 130.8 mg L −1 , at pH 4 and pH 5, and T = 293 K and t = 120 min. T are depicted in Figure 8, where Pb(II) adsorption capacity qe vs. Ce concentratio in equilibrium is presented. These data were fitted to two very common isothe [39,40]: i) Langmuir model, which assumes solute sorption in monolayers wit geneous sorption energy; ii) Freundlich model, which assumes multilayer sorp heterogeneous sorption energies. The adjustment parameters with both model tained by fitting the corresponding experimental data [Ce/qe vs. Ce] and [log(qe) We can see ( Table 2) that adsorption isotherms are fitted better with Langmuir (R 2 close to 1) than the Freundlich model (R 2 ≤ 0.87). From the first model and, for both pH 4 and pH 5, lower K L and higher maximum sorption capacity q max values are obtained for WACM than for WTCM. At pH 5, q max = 303.0 mg g −1 for WACM is almost 21% higher than the value at pH 4, while for WTCM, q max = 223.1 mg g −1 is practically the same at both pHs. q max values of agro-industrial wastes, with alkaline treatment, are consigned in Table 3. We can note that q max of WACM and WTCM are among the highest. On the other hand, it is important to mention that that q max value of the treated biosorbent is almost twice that of its corresponding untreated precursor [21].

Kinetic of Biosorption
The kinetic study is of great importance for the practical and effective use of biosorbents in the industry [41]. In this work, the kinetic studies were carried out by varying the adsorption time from 0 to 180 min, at pH 4; C 0 = 48.42 and 130.8 mg L −1 ; biosorbent dosage = 2 g L −1 ; and T = 293 K.
The experimental kinetic data were modeled using three adsorption kinetic models (Table 4): pseudo first-order model, pseudo second-order model, and ploting q t vs. t 1/2 (Weber and Morris model). The parameters obtained after the non-linear adjustments, including correlation coefficient R 2 , are consigned in Table 4. Figure 9 shows the non-linear fit of the pseudo second-order equation to the kinetic data. For both the WACM and WTCM biosorbents, a better correlation (R 2 ≈ 1) is obtained with pseudo-second-order than the first-order adjustment models. This result indicates that Pb(II) adsorption is a chemisorption process [42]. We can note that the calculated adsorption capacities q e,cal are close to those determined experimentally and, for a given C 0 concentration, greater for WACM than for WTCM. The adsorption rates (k 2 , rate constant adsorption and h, initial adsorption rate) are comparable for both biosorbents.
The q t vs. t 0.5 data, for both WACM and WTCM biosorbents, are depicted in Figure 10 and fitted with the intra-particle diffusion Weber-Morris model. According to the k d intra-particle diffusion rate constants (in mg g −1 min −1/2 , Table 4), we can distinguish three parts: The first part shows rapid growth of q t at time t (k d,I > 10.3), particularly at high initial Pb(II) concentrations (C 0 = 130.8 mg L −1 ), where k d,I can reach values up to seven times higher than for those at low C 0 concentrations (e.g., 48.4 mg L −1 ). These results would indicate the rapid absorption of Pb(II) ions on the surface of the biosorbents. The second part shows slower growth of q t with t (0.16 < k d,II < 12.4), which would be related to a gradual sorption process, where Pb(II) sorbed would fill the biosorbent pores; this part would be related to the diffusion of Pb(II) inside the biosorbent (intraparticle diffusion) [24]. Finally, the 3rd part shows that q t is practically constant with very low k d,III values (k d,III < 0.6). It indicates that the equilibrium between Pb(II) ions in the solution and the sorbent surface is reached.  For both the WACM and WTCM biosorbents, a better correlation (R 2  1) is obtained with pseudo-second-order than the first-order adjustment models. This result indicates that Pb(II) adsorption is a chemisorption process [42]. We can note that the calculated adsorption capacities qe,cal are close to those determined experimentally and, for a given C0 concentration, greater for WACM than for WTCM. The adsorption rates (k2, rate constant adsorption and h, initial adsorption rate) are comparable for both biosorbents.
The qt vs. t 0.5 data, for both WACM and WTCM biosorbents, are depicted in Figure  10 and fitted with the intra-particle diffusion Weber-Morris model. According to the kd intra-particle diffusion rate constants (in mg g −1 min −1/2 , Table 4 gradual sorption process, where Pb (II) sorbed would fill the biosorbent pores; this part would be related to the diffusion of Pb(II) inside the biosorbent (intraparticle diffusion) [24]. Finally, the 3rd part shows that qt is practically constant with very low kd,III values (kd,III < 0.6). It indicates that the equilibrium between Pb(II) ions in the solution and the sorbent surface is reached.

Biosorption Thermodynamics
ΔG 0 was calculated from Equations 4 and 5. The plot lnKc vs. 1/T (eq. 7), depicted in Figure 11, was fitted using the least squares method, aiming to calculate the ΔH 0 and ΔS 0 values. The results are consigned in Table 5 and indicate that the Pb(II) adsorption process on both the WACM and WTCM biosorbents is: i) feasible, spontaneous (positive ΔG 0 values), and more favorable with increasing temperature; ii) endothermic by nature (positive ΔH 0 values); and iii) a process with increasing randomness (positive ΔS 0 values) at the solid-liquid interface [43].
Similar results were reported by Song et al. [27], Morosanu et al. [31], Milojkovic et al. [44], among others, for Pb(II) removal by Auricularia auricular spent substrate, rapeseed biomass, and Myriophyllum spicatum and its compost, respectively. However, a feasible, spontaneous, but exothermic process of Pb(II) adsorption was also reported by Petrović

Biosorption Thermodynamics
∆G 0 was calculated from Equations (4) and (5). The plot lnK c vs. 1/T (Equation (7)), depicted in Figure 11, was fitted using the least squares method, aiming to calculate the ∆H 0 and ∆S 0 values. The results are consigned in Table 5 and indicate that the Pb(II) adsorption process on both the WACM and WTCM biosorbents is: (i) feasible, spontaneous (positive ∆G 0 values), and more favorable with increasing temperature; (ii) endothermic by nature (positive ∆H 0 values); and (iii) a process with increasing randomness (positive ∆S 0 values) at the solid-liquid interface [43].

Preparation of Biosorbents
Arabica-coffee (WAC) and Theobroma-cocoa (WTC) waste were ob Satipo and Chanchamayo provinces, respectively, located at Junín, Perú were previously washed with abundant distilled water, dried in an oven a and finally ground. WAC and WTC were treated with a 0.1 M NaOH sol liquid ratio of 1:10 (g biomass: mL solution) for 24 h with constant stirr Once treated, both products were filtered and washed with abundant d and then were dried in an oven at 60 °C for 48 h, and finally, both treated s and WTCM) were ground again and homogenized with a 70 mesh sieve.
The alkaline treatment of precursor wastes produced a biomass loss. loss was determined as the difference between the initial sample weight (b and the final sample weight (after treatment) All chemical reagents used in this work were of analytical grade.


The point of zero charge (pHPZC) study was evaluated according to th reported by do-Nascimento et al. [45]. A mixture of 0.05 g of biomas   [31], Milojkovic et al. [44], among others, for Pb(II) removal by Auricularia auricular spent substrate, rapeseed biomass, and Myriophyllum spicatum and its compost, respectively. However, a feasible, spontaneous, but exothermic process of Pb(II) adsorption was also reported by Petrović et al. [14] when removing Pb(II) with a hydrochar of grape pomace. Mahyoob et al. [29] determined that Pb(II) adsorption on co-processed olive tree leaves was an exothermic process with decreasing randomness (negative ∆S 0 ).

Preparation of Biosorbents
Arabica-coffee (WAC) and Theobroma-cocoa (WTC) waste were obtained from the Satipo and Chanchamayo provinces, respectively, located at Junín, Perú. Both samples were previously washed with abundant distilled water, dried in an oven at 60 • C for 48 h, and finally ground. WAC and WTC were treated with a 0.1 M NaOH solution in a solid-liquid ratio of 1:10 (g biomass: mL solution) for 24 h with constant stirring at 300 rpm. Once treated, both products were filtered and washed with abundant deionized water and then were dried in an oven at 60 • C for 48 h, and finally, both treated samples (WACM and WTCM) were ground again and homogenized with a 70 mesh sieve.
The alkaline treatment of precursor wastes produced a biomass loss. The percentage loss was determined as the difference between the initial sample weight (before treatment) and the final sample weight (after treatment) All chemical reagents used in this work were of analytical grade.

Biosorbent Characterization
• The point of zero charge (pH PZC ) study was evaluated according to the methodology reported by do-Nascimento et al. [45]. A mixture of 0.05 g of biomass with 50 mL of an aqueous solution under different initial pHs (pH 0 ) ranging from 1 to 12 was prepared. The acid dilutions were prepared from a 1 M HCl solution, while basic dilutions were from 1 M NaOH. After 24 h of equilibrium, the final pHs (pH f ) were measured.

•
The concentrations of the acid and basic groups (or acid/basic titrable sites) on the surface of WACM and WTCM were determined using the Boehm method reported by Aygun et al. [46]. For acid titrable sites (between brackets for basic sites), mixtures of 0.25 g (0.5 g) of the biosorbent with 50 mL of a standardized 0.05 M NaOH (0.1 M HCl) solution were prepared. All the mixtures were shaken, at room temperature, for 24 h at 100 rpm, and then, for each mixture, 20 mL (10 mL) of the supernatant liquid was pipetted and excess acid (base) was adequately titrated using bromocresol blue or phenolphtalien) as an indicator. • Fourier transform infrared (FTIR, SHIMADZU IR Affinity) spectroscopy, over a spectral range of 4000 to 500 cm −1 was used to characterize the functional groups present on the surface of WACM and WTCM before and after Pb(II) biosorption.

•
Morphological and elemental analysis on the surface of biosorbents were performed by Scanning Electron Microscopy (SEM) coupled with EDX (Energy Dispersive X-rays spectroscopy) (LEO 440 model).

Adsorption Experiments
Batch experiments were carried out using Pb(NO 3 ) 2 solution, with varying Pb(II) concentrations between 5.63 to 130.8 mg L −1 . The dose of the modified biomass was varied in the range of 0.5 to 6 g L −1 and adjusted to a pH in the range of 2.0 to 5.0 (using LAQUA PH1200) by adding 0.01 M HNO 3 or 0.01 M NaOH. The solutions, at room temperature, were stirred to 150 rpm for 120 min, and the samples were taken at certain time intervals.
The Pb(II) concentrations, before and after adsorption, were evaluated using an Atomic Absorption Spectrophotometer (SHIMADZU-AAS 6800). All adsorption experiments were replicated three times, and the results were averaged.
where C 0 and C e (in mg·L −1 ) are the initial and equilibrium final Pb(II) concentrations, respectively, and V (in L) is the volume of solution and m (in g) is the biosorbent mass. The adsorption isotherms and kinetic data were obtained by contacting a biomass dose of 2 g L −1 with different Pb(II) concentrations at pH 4 (also pH 5 for isotherms). The experimental data were adjusted to the corresponding adsorption models described in Table 6 (isotherms) and Table 7 (kinetic). Table 6. Adsorption isotherm models.

Model Equation Parameters
Langmuir C e q e = 1 q mx k L + C e q mx q e (mg g −1 ): adsorption capacity C e (mg L −1 ): adsorbate concentration in equilibrium q max (mg g −1 ): maximum sorption capacity k L (L mg −1 ): Langmuir constant related to the affinity between sorbent and sorbate Freundlich lnq e = lnk F + 1 n lnC e k F (mg g −1 L (1/n) ·mg −(1/n) ): equilibrium constant n: constant related to the affinity between sorbent and sorbate Table 7. Kinetic adsorption models.

Model Equation Parameters
Pseudo-first order q t = q e 1 − e −K 1 t q e (mg g −1 ): adsorption capacity q t (mg g −1 ): amount of Pb(II) retained per unit biomass at time t. k 1 (min −1 ): first-order kinetic constant k 2 (g (mg min) −1 ): rate constant adsorption h (mg(g min) −1 ): initial adsorption rate Pseudo-second order q t = q e q e K 2 t 1+q e K 2 t h = k 2 q 2 e Weber and Morris q t = k d t 1/2 + B k d (mg g −1 min −1/2 ): intraparticle diffusion rate constant B (mg g −1 ): constant related to the thickness of the adsorbent boundary layer

Thermodynamic Parameters
This study was carried out by varying the temperature from 293 to 313 K (20 to 40 • C) using a biosorbent dose of 2 g L −1 , an initial Pb(II) concentration C 0 = 130.8 mg L −1 , and a contact time 120 min at pH 4. Parameters such as free energy change ∆G • , enthalpy change ∆H • , and entropy change, ∆S • for the adsorption process studied were calculated using Equations (3)-(6) [47,48]: where R is the ideal gas constant (8.314 J mol −1 K −1 ); T is the absolute temperature of the solution; K c is the thermodynamic equilibrium constant; and C es and C e are Pb(II) concentrations at equilibrium, respectively, in the biosorbent and solution.
The point of zero charge, pH PZC values at 6 and 6.8, for WACM and WTCM, respectively, were higher than those for the corresponding WAC and WTC untreated samples. These measurements were consistent with the concentration of basic titrable sites, which was almost 1.4 times higher for treated than untreated biosorbents. The basic sites would be associated with the OH, C=O, COH, and C-O-C functional groups, which were identified by FTIR measurements.
SEM/EDX analyses showed typical uneven and rough surface morphologies, more porous and less compact for clean than for Pb(II)-loaded biosorbents.
Both the WACM and WTCM biosorbents reached high Pb(II) removal efficiency %R values (>90%) for a biomass dosage of 2 g L −1 at pH between 4 and 5, with initial Pb(II) concentrations C 0 in the range of 5 to 40 mg L −1 . For these conditions, the Pb(II) adsorption capacity q e of WACM (and WTCM) was almost three times higher than that of the corresponding untreated biosorbent.
The adsorption isotherm data were well fitted with the Langmuir model (monolayer adsorption mechanism), which provided, at pH 5, maximum Pb(II) adsorption capacities, q max , equal to 303.0 and 223.1 mg g −1 for WACM and WTCM, respectively. These values are i) twice those corresponding to untreated samples and ii) higher than for other similar alkaline-treated biosorbents reported in the literature.
The adsorption kinetic data were well fitted with the pseudo-second-order model, indicating that the Pb(II) adsorption on WACM and WTCM was a chemisorption process.