Oxidized Biomass and Its Usage as Adsorbent for Removal of Heavy Metal Ions from Aqueous Solutions

Nowadays, very coarse wool fibers are considered waste biomass and are discarded at random or burned. Therefore, it is of actual interest to valorize coarse wool fibers as utile products. In this sense, we report herein an environmentally-friendly process for the preparation of a new material based on oxidized wool fibers and designed for efficient adsorption of heavy metals from wastewater. The morphology and the structure of the obtained product were characterized by scanning electron microscopy (SEM) coupled with an X-ray energy-dispersive module (EDX) and by Fourier-transform infrared spectroscopy (FTIR). Likewise, the performances of the oxidized wool fibers for the adsorption of heavy metal cations (Cu2+, Cd2+, Pb2+) from aqueous solutions were tested. The adsorption kinetics data were analyzed by applying the pseudo-first-order (PFO) and pseudo-second-order (PSO) kinetic models. The equilibrium of the adsorption process was investigated by using the Freundlich and Langmuir isotherm models. According to the Langmuir isotherms registered at 300 K, the maximum adsorption capacities of the oxidized wool were found to increase from Cu2+ (9.41 mg/g) and Cd2+ (10.42 mg/g) to Pb2+ (30.71 mg/g). Consequently, the removal efficiency of metal ions was found to vary in the range of 96.8–99.7%. The thermodynamic parameters (e.g., enthalpy, entropy, and Gibbs free energy) were calculated and discussed.


Introduction
Anthropogenic pollution with heavy metals is considered one of the major environmental issues, accelerated by the development of different human activities (e.g., industrial processes, vehicle emissions, agriculture, e-waste disposal, and waste treatments), which generated significant quantities of such persistent pollutants [1,2]. The industrial development has been conducted to the obvious economic growth of modern society, unfortunately it has also caused serious damage to the environment [3]. A cost-benefit analysis of the last decades disclosed a value of at least $200 billion in the United States alone that was spent to clean up hazardous waste using conventional technologies [4,5]. Generally, the term "heavy metals" is attributed to metals that have densities higher than 5 g/cm 3 , atomic weights greater than 50, and atomic numbers bigger than 20 [1,6]. On the other hand, heavy metals are the natural components of the earth's crust. Consequently, the process of rock erosion, natural degradation of plants and animal waste, precipitation, and airborne particles accumulated in the atmosphere (from volcanic eruptions and fire smoke) are all contributing to environmental pollution [1].
Some heavy metals are essential in many structural and biochemical functions of plants (including their growth, e.g., Cu 2+ ) [7,8]. By contrast, other heavy metals (i.e., Pb 2+ and Cd 2+ ) are non-essential and have no known biological function, being toxic

Results and Discussion
The chemical composition of wool fibers is a complex one, with different functional carboxyl, amino, hydroxyl, disulfide and thiol groups that have relevant affinities for heavy metal ions [17,20]. The most prone metal ion binding sites are the carboxyl groups of aspartic and glutamic acid found in the structure of keratine [21,22]. These affinities rely on electrostatic attraction and donor/acceptor interactions between the metal cations and the surface of the material [17,23]. For instance, the carboxyl groups in their anionic form (-COO − ) are able to attract positively charged metal cations by electrostatic forces (ionic bonds). In turn, the pair of non-participating electrons of the nitrogen in the amine groups (-NH 2 ) or of the oxygen in the hydroxyl groups (-OH) may interact with metal cations present in the solution through coordinative (donor/acceptor) bonds [24]. Also, mixed electrostatic and coordinating wool-cation interactions could be envisaged, depending on the solution pH [21,24,25]. Scheme 1 shows a suggested scenario of the mechanism of adsorption of heavy metal ions on oxidized wool fibers.
Wool fibers were already proved as appropriate sorbents for heavy metal cations [20,23,[26][27][28][29][30][31]. Analyzing the dedicated references, one may affirm that the cations' sorption performances of the wool-based materials are strongly affected by their chemical composition and morphology, their specific surface, etc. To increase the superficial content of anionic groups of wool fibers, their oxidation was performed as discussed in the next section. Wool fibers were already proved as appropriate sorbents for heavy metal cations [20,23,[26][27][28][29][30][31]. Analyzing the dedicated references, one may affirm that the cations' sorption performances of the wool-based materials are strongly affected by their chemical composition and morphology, their specific surface, etc. To increase the superficial content of anionic groups of wool fibers, their oxidation was performed as discussed in the next section.

Oxidized Wool Fibers
To increase the surfaces functionality of the wool fibers, they were oxidized in the presence of a mixture of H2O2 and NH4OH, as mentioned in Section 3.2. Newly formed functional groups were expected to emerge as a result of the oxidation process. For example, herein we might expect the appearance of carboxyl and cysteine oxide groups generated by the oxidation of amide bonds from the chain backbone of wool keratin, respectively, of the disulfide cross-linking bridges.
The surface morphology and the structural characteristics of pristine wool as compared with those of the oxidized wool fibers were investigated by using scanning electron microscopy (SEM) coupled with energy-dispersive X-ray (EDX) spectroscopy ( Figure 1 and

Oxidized Wool Fibers
To increase the surfaces functionality of the wool fibers, they were oxidized in the presence of a mixture of H 2 O 2 and NH 4 OH, as mentioned in Section 3.2. Newly formed functional groups were expected to emerge as a result of the oxidation process. For example, herein we might expect the appearance of carboxyl and cysteine oxide groups generated by the oxidation of amide bonds from the chain backbone of wool keratin, respectively, of the disulfide cross-linking bridges.
The surface morphology and the structural characteristics of pristine wool as compared with those of the oxidized wool fibers were investigated by using scanning electron microscopy (SEM) coupled with energy-dispersive X-ray (EDX) spectroscopy (Figures 1 and 2, respectively) and FTIR analysis ( Figure 3). Wool fibers were already proved as appropriate sorbents for heavy metal catio [20,23,[26][27][28][29][30][31]. Analyzing the dedicated references, one may affirm that the cations' so tion performances of the wool-based materials are strongly affected by their chem composition and morphology, their specific surface, etc. To increase the superficial c tent of anionic groups of wool fibers, their oxidation was performed as discussed in next section.

Oxidized Wool Fibers
To increase the surfaces functionality of the wool fibers, they were oxidized in presence of a mixture of H2O2 and NH4OH, as mentioned in Section 3.2. Newly form functional groups were expected to emerge as a result of the oxidation process. For exa ple, herein we might expect the appearance of carboxyl and cysteine oxide groups gen ated by the oxidation of amide bonds from the chain backbone of wool keratin, resp tively, of the disulfide cross-linking bridges.
The surface morphology and the structural characteristics of pristine wool as co pared with those of the oxidized wool fibers were investigated by using scanning electr microscopy (SEM) coupled with energy-dispersive X-ray (EDX) spectroscopy (Figur and Figure 2, respectively) and FTIR analysis ( According to SEM images, pristine wool was found to be of very coarse grade [32] as the average fiber diameter is 66.0 ± 6.7 µm (Figure 1a). The fibers show a roughness emerging as a scale-like pattern (cuticle cells), which is characteristic of wool fibers [28]. This specific surface morphology does not change following oxidation (Figure 1b). The EDX spectra of both pristine ( Figure 2a) and oxidized (Figure 2b) samples unveiled the presence of all expected chemical elements, i.e., C, O, N, S (Pt is due to the fiber metallization). As well known, EDX elemental analysis provides local element concentrations and depends on the depth from the surface the elements are counted. To have a better certitude on the fiber elemental composition, five different zones on the fibers were analyzed. The weight and atomic proportions reported in the inset tables ( Figure 2) represent the average values.
Hence, the EDX technique provides qualitative elemental analysis, which is surfacespecific (i.e., the estimation has a local character). This could be the reason the obtained composition (C = 60.4%; N = 14.7%; O = 19.1%; S = 4.0%) of the pristine wool is somewhat different as compared to the averages values found for wool in the literature (C = 50%; N = 16-17%; O = 22-25%; S = 3-4%) [33]. Looking to the composition of pristine ( Figure 2a) and oxidized wool fibers (Figure 2b), one may observe that both oxygen and nitrogen contents of the oxidized sample are higher as compared to the pristine one.    To further clarify the structural aspects, Figure 3 compares the specific IR absorptions of pristine wool and of the oxidized sample. The interpretation of the FTIR spectra of both materials was performed based on the information available in the literature [33][34][35][36][37][38][39][40]. As one may see from Figure 3a, the pristine wool shows the following characteristic absorptions: a large band centered at 3630 cm −1 (O-H groups and adsorbed water possibly superposed on amide A absorptions), bands located between 2955 and 2853 cm −1 (C-H vibrations), 1707 and 1688 cm −1 (COO-from aspartic and glutamic units), 1661 cm −1 (C=O, amide I vibration specific to secondary structure of wool keratin), 1576 and 1541 cm −1 (C-NH, amide II), 1229 cm −1 (amide III vibrations), 1161, 1066, and 1040 cm −1 (oxidized cystine groups, already present in the pristine sample). After oxidation (Figure 3b), the amide A (3250 cm −1 ), amide B (3057 cm −1 ) and amide III absorptions (1228 cm −1 ) become more visible, while the amide II vibrations move to the 1549-1500 cm −1 region. Moreover, the increase of the intensities of O-H (3587 cm −1 ), carboxylate overlapped on amide I (1707 cm −1 ) and of cysteine oxides bands (1167, 1069, 1036 cm −1 ) as compared to those of C-H groups (2957-2872 cm −1 ) is easily visible, confirming the higher content of oxygen in the oxidized sample evidenced by EDX analysis. Similar results were obtained when keratin was extracted by using H 2 O 2 in alkaline media [36].
Thus, the increase of the oxygen and nitrogen contents in the oxidized sample could be mainly explained by the splitting of disulfide cystine linkage, respectively, by the neutralization of cysteic acid groups with ammonium hydroxide, according to the reactions represented in Scheme 2. According to SEM images, pristine wool was found to be of very coarse grade [32] as the average fiber diameter is 66.0 ± 6.7 μm (Figure 1a). The fibers show a roughness emerging as a scale-like pattern (cuticle cells), which is characteristic of wool fibers [28]. This specific surface morphology does not change following oxidation (Figure 1b). The EDX spectra of both pristine ( Figure 2a) and oxidized (Figure 2b) samples unveiled the presence of all expected chemical elements, i.e., C, O, N, S (Pt is due to the fiber metallization). As well known, EDX elemental analysis provides local element concentrations and depends on the depth from the surface the elements are counted. To have a better certitude on the fiber elemental composition, five different zones on the fibers were analyzed. The weight and atomic proportions reported in the inset tables ( Figure 2) represent the average values.
Hence, the EDX technique provides qualitative elemental analysis, which is surfacespecific (i.e., the estimation has a local character). This could be the reason the obtained composition (C = 60.4%; N = 14.7%; O = 19.1%; S = 4.0%) of the pristine wool is somewhat different as compared to the averages values found for wool in the literature (C = 50%; N = 16-17%; O = 22-25%; S = 3-4%) [33]. Looking to the composition of pristine ( Figure 2a) and oxidized wool fibers (Figure 2b), one may observe that both oxygen and nitrogen contents of the oxidized sample are higher as compared to the pristine one.
To further clarify the structural aspects, Figure 3 compares the specific IR absorptions of pristine wool and of the oxidized sample. The interpretation of the FTIR spectra of both materials was performed based on the information available in the literature [33][34][35][36][37][38][39][40]. As one may see from Figure 3a, the pristine wool shows the following characteristic absorptions: a large band centered at 3630 cm −1 (O-H groups and adsorbed water possibly superposed on amide A absorptions), bands located between 2955 and 2853 cm −1 (C-H vibrations), 1707 and 1688 cm −1 (COO-from aspartic and glutamic units), 1661 cm −1 (C=O, amide I vibration specific to secondary structure of wool keratin), 1576 and 1541 cm −1 (C-NH, amide II), 1229 cm −1 (amide III vibrations), 1161, 1066, and 1040 cm −1 (oxidized cystine groups, already present in the pristine sample). After oxidation (Figure 3b), the amide A (3250 cm −1 ), amide B (3057 cm −1 ) and amide III absorptions (1228 cm −1 ) become more visible, while the amide II vibrations move to the 1549-1500 cm −1 region. Moreover, the increase of the intensities of O-H (3587 cm −1 ), carboxylate overlapped on amide I (1707 cm −1 ) and of cysteine oxides bands (1167, 1069, 1036 cm −1 ) as compared to those of C-H groups (2957-2872 cm −1 ) is easily visible, confirming the higher content of oxygen in the oxidized sample evidenced by EDX analysis. Similar results were obtained when keratin was extracted by using H2O2 in alkaline media [36].
Thus, the increase of the oxygen and nitrogen contents in the oxidized sample could be mainly explained by the splitting of disulfide cystine linkage, respectively, by the neutralization of cysteic acid groups with ammonium hydroxide, according to the reactions represented in Scheme 2.

Scheme 2.
Reaction scheme for splitting of disulfide linkage (cystine) and neutralization of cysteic acid groups with ammonium hydroxide.

Adsorption/Desorption of Metal Cations on/from Wool Fibers
Once the superficial composition of the oxidized wool fibers was identified, one may propose a possible scenario of the mechanism of adsorption of the heavy metal cations on the oxidized wool fibers (Scheme 1). Many years ago, Friedman [21] demonstrated that the free carboxyl groups of aspartic and glutamic acid (available in the amorphous polypeptide sequences of wool), almost completely dissociated at pH 7 as well as cystine Scheme 2. Reaction scheme for splitting of disulfide linkage (cystine) and neutralization of cysteic acid groups with ammonium hydroxide.

Adsorption/Desorption of Metal Cations on/from Wool Fibers
Once the superficial composition of the oxidized wool fibers was identified, one may propose a possible scenario of the mechanism of adsorption of the heavy metal cations on the oxidized wool fibers (Scheme 1). Many years ago, Friedman [21] demonstrated that the free carboxyl groups of aspartic and glutamic acid (available in the amorphous polypeptide sequences of wool), almost completely dissociated at pH 7 as well as cystine groups, seem to be the most prone to bind metal cations. Instead, the coordination binding of metal cations to amine and amide groups takes place in alkaline pH conditions [21,24].
The SEM images in Figure 1c-e show that the morphology of wool fibers was kept after the cation sorption, while the EDX elemental analysis identified the presence of the adsorbed Cu 2+ (Figure 2c), Cd 2+ ( Figure 2d) and Pb 2+ (Figure 2e) cations.

Adsorption Kinetics
The adsorption kinetics of three representative heavy metal ions (Cu 2+ , Cd 2+ , Pb 2+ ) onto the surface of the oxidized wool fibers was investigated for single-component systems (i.e., retention of each metal ion separately) and for a multi-component system (i.e., competitive retention of three metal ions simultaneously). Figure 4 illustrates the relationship between the adsorption capacity q t (mg/g) and the contact time t (min) for the single-solute systems ( Figure 4a) and for the competitive system ( Figure 4b). The equilibrium adsorption capacities were found to be 8.73 mg/g (for retention of Pb 2+ ), 8.30 mg/g (for retention of Cu 2+ ), and 7.29 mg/g (for retention of Cd 2+ ). For the multi-component system ( Figure 4b) the results revealed the best selectivity for Cd 2+ ions and the lowest selectivity for Pb 2+ ions. By comparison with the single component systems, the adsorption capacities of oxidized wool strongly decreased when all the studied pollutant cations were mixed together. At equilibrium, the sum of the three maximum adsorption capacities was equal to 24.3 mg/g when summing up the adsorption capacities for the individual cation solutions. In turn, the same totaling value was found to be 14.6 mg/g, when summing up the adsorption capacities for the multi-component system. This discrepancy might be attributed to the fact that these two situations are not thermodynamically equivalent, that is, the equilibria are established in different ways. In this sense, ionic strength plays an important role. For instance, the salt source for Cu 2+ ions was copper sulfate (CuSO 4 × 5H 2 O). Instead, for Pb 2+ and Cd 2+ ions, their nitrate salts were employed, i.e., (Pb(NO 3 ) 2 and Cd(NO 3 ) 2 × 4H 2 O). Hence, for the individual systems, the following ionic strengths were calculated: (1) I1 = 0.80 mM for CuSO 4 , (2) I2 = 0.49 mM for Cd(NO 3 ) 2 , and (3) I3 = 0.45 mM for Pb(NO 3 ) 2 . Whereas, for the multi-component system (Cu(II)+Cd(II)+Pb(II)), the ascertained ionic strength was equal to I = 0.70 mM. Notice that this value is lower than the value of the ionic strength for the Cu 2+ individual system. Thus, these evident discrepancies between ionic strengths of different adsorption systems may contribute to the discrepancy between the sum of the adsorption capacities for the multi-component system and individual systems. Therefore, the maximum capacity values were reduced in the multi-component system compared to individual systems as follows: by 56.01% for lead cations, 39.06% for copper cations and by 22.50% for cadmium cations, respectively. These findings are in close agreement with previous literature reports [41,42] and are usually attributed to the significant competition between the studied cations on the adsorption sites of the oxidized wool [41]. According to the literature, there are numerous factors that can be considered for ordering the cations adsorption capacities in multicomponent systems, such as: covalent index, ionic charge, ionic radii or hydrated ionic radii, electronegativity, among others [41][42][43]. In our study, adsorption capacities follow the order: Cd 2+ > Cu 2+ > Pb 2+ for the multi-component system. This result indicates that the order is reversely correlated with the hydrated ionic radii of the cations (4.01 Å for lead, 4.19 Å for copper and 4.26 Å for cadmium [43]). In fact, one may suppose that, once the bulkiest hydrated cation Cd 2+ were adsorbed on the oxidized wool surface they are hindering the access of the other cations to the active adsorption sites. This fact resulted in the stronger decrease of the adsorption capacity values for Pb 2+ and Cu 2+ individually and in the premature exhaustion of the adsorbent. In addition, the slower diffusion of Pb 2+ might be also associated with a greater atomic radius (1.80 Å) of this cation, compared to Cd (1.55 Å) and Cu (1.35 Å) atoms [44]. Once the adsorption equilibrium in the multi-component system has been established, it might be assumed that the subsequent exchange between these three cations is minimized. This might be deduced from experimental observations regarding desorption in distilled water, which showed insignificant desorption of heavy metal cations in water eluent (as further discussed in Section 2.2.4). Generally, the adsorption capacity gets higher with the increment of the contact time, until the stationary regime (equilibrium) is established. According to kinetic data reported in Figure 4, the adsorption equilibrium is generally established for contact time of more than 60 min. insignificant desorption of heavy metal cations in water eluent (as further discussed in Subsection 2.2.4.). Generally, the adsorption capacity gets higher with the increment of the contact time, until the stationary regime (equilibrium) is established. According to kinetic data reported in Figure 4, the adsorption equilibrium is generally established for contact time of more than 60 min.
(a) (b) The experimental data were interpolated by means of the pseudo-first-order (PFO) and pseudo-second-order (PSO) model equations given by Equation (1) and Equation (2), respectively [45]: where qt and qe represent the adsorption capacities at time t and at equilibrium, respectively; k1 and k2 are constants for the equilibrium rate of the PFO and PSO adsorption. The parameters of kinetic models were calculated by the non-linear regression tool implemented in OriginPro software (version 9.9.0.225/SR1 2022). The concordance between observations and models was estimated by using the reduced χ-square statistic test (χ ) as the error function, which can be calculated according to Equation (3): where q (obs) denotes the observed adsorption capacity (mg/g); q (calc) is the calculated adsorption capacity (mg/g); i represents an integer index of summation; σi 2 is the variance (if no data replication, then σi 2 = 1); n and p represent the numbers of observation and fitted parameters, respectively, and the difference n − p gives the degree of freedom (v = n − p). The experimental data were interpolated by means of the pseudo-first-order (PFO) and pseudo-second-order (PSO) model equations given by Equation (1) and Equation (2), respectively [45]: where q t and q e represent the adsorption capacities at time t and at equilibrium, respectively; k 1 and k 2 are constants for the equilibrium rate of the PFO and PSO adsorption. The parameters of kinetic models were calculated by the non-linear regression tool implemented in OriginPro software (version 9.9.0.225/SR1 2022). The concordance between observations and models was estimated by using the reduced χ-square statistic test (χ 2 v ) as the error function, which can be calculated according to Equation (3): where q (obs) denotes the observed adsorption capacity (mg/g); q (calc) is the calculated adsorption capacity (mg/g); i represents an integer index of summation; σ i 2 is the variance (if no data replication, then σ i 2 = 1); n and p represent the numbers of observation and fitted parameters, respectively, and the difference n -p gives the degree of freedom (v = n -p). Typically, the lower the error function is, the better the model prediction. As one can see from Figure 4 and by inspecting χ 2 v values (Tables 1 and 2), both PFO and PSO kinetic models interpolate appropriately the experimental data. However, one model may be somewhat better than another one for a particular situation. For example, for the adsorption of Cu 2+ ions in the single-component system, the PSO kinetic model gives a better prediction. Instead, Cu 2+ adsorption in the competitive system is appropriately predicted by both PFO and PSO models.

Adsorption Isotherms
Isotherms describe the equilibrium between adsorption and desorption phenomena at a given temperature, representing the basis for the design of the adsorption system. The importance of adsorption isotherms relies on detailed information on the distribution of the adsorbate species (metal cations in this case) between the liquid phase and solid surface of the adsorbent at equilibrium and at different temperatures. In this study, the adsorption isotherms were explored at two different temperatures (300 K and 323 K), and the adsorption capacity was recorded at the contact time t = 240 min to ensure the attaining of the equilibrium. Figure 5 presents the adsorption isotherms describing the retention of heavy metal cations (Pb 2+ , Cd 2+ , Cu 2+ ) onto the oxidized wool fibers at 300 K ( Figure 5a) and 323 K (Figure 5b). As shown in Figure 5, as the equilibrium concentration C e (mg/L) increases, the adsorption capacity q e (mg/g) gets higher as well, attaining a stationary plateau for a given level of temperature. The equilibrium data revealed that the adsorption of Pb 2+ cations is favored by the increase in temperature, while the adsorption of Cu 2+ and Cd 2+ cations is practically not so sensitive to temperature variation. This observation is analyzed in more detail in the next section dealing with the thermodynamics of the adsorption process. The experimental data were interpolated by using Freundlich and Langmuir isotherm equations (Equations (4) and (5), respectively), which are the most frequently used models in this sense [46][47][48][49][50]: where q e and q m represent the adsorption capacity at equilibrium and the maximum adsorption capacity, respectively, C e is the metal concentration at equilibrium, K F and K L are the Freundlich and Langmuir isotherms constants, and 1/n F is a variable heterogeneity factor (between 0 and 1).
where qe and qm represent the adsorption capacity at equilibrium and the maximum adsorption capacity, respectively, Ce is the metal concentration at equilibrium, KF and KL are the Freundlich and Langmuir isotherms constants, and 1/nF is a variable heterogeneity factor (between 0 and 1).
(a) (b) According to the Langmuir model, every adsorption site is energetically equivalent, and it can hold just one adsorbate molecule. Thus, the Langmuir approach only considers the formation of a monolayer of adsorbate onto the outer surface of the adsorbent. In turn, the Freundlich isotherm considers both the initial adsorption onto the heterogeneous surface as well as the condensation effect occurring due to adsorbate-adsorbate interaction [50]. Previous research has shown that the non-linear regression methods were more precise in estimating the model parameters compared to the linear regression techniques employed for the same purpose [45]. Hence, in this work, the non-linear regression method was applied with the aid of the OriginPro program in order to calculate the isotherm equations parameters. Also, the error function (χ ) was calculated to estimate the goodness-offit. Table 3 lists the calculated parameters of the isotherm models along with computed χ values. According to Figure 5 and Table 3, the Langmuir isotherm model provided better predictions than the Freundlich model. Hence, the Langmuir approach (i.e., monolayer adsorption) seems to represent better the equilibrium distribution of metal ions between the oxidized wool fibers and the liquid phase. However, both models for the adsorption isotherms for Pb 2+ show important deviation from the experimental values at low equilibrium concentration ( Figure 5) and the corresponding computed χ values are quite high (Table 3). The maximum adsorption capacities at 300 K were found to be 9.41, 10.42, and 30.71 mg/g for retention of Cu 2+ , Cd 2+ , and Pb 2+ cations, respectively. These values were found to be substantially greater than the values registered for the adsorption capacities of pristine wool fibers (2.49, 3.87, and 8.66 mg/g for retention of Cu 2+ , Cd 2+ , and Pb 2+ ions, respectively). At 323 K, the maximum adsorption capacities were somewhat higher (Table 3). According to the Langmuir model, every adsorption site is energetically equivalent, and it can hold just one adsorbate molecule. Thus, the Langmuir approach only considers the formation of a monolayer of adsorbate onto the outer surface of the adsorbent. In turn, the Freundlich isotherm considers both the initial adsorption onto the heterogeneous surface as well as the condensation effect occurring due to adsorbate-adsorbate interaction [50]. Previous research has shown that the non-linear regression methods were more precise in estimating the model parameters compared to the linear regression techniques employed for the same purpose [45]. Hence, in this work, the non-linear regression method was applied with the aid of the OriginPro program in order to calculate the isotherm equations parameters. Also, the error function (χ 2 v ) was calculated to estimate the goodness-of-fit. Table 3 lists the calculated parameters of the isotherm models along with computed χ 2 v values. According to Figure 5 and Table 3, the Langmuir isotherm model provided better predictions than the Freundlich model. Hence, the Langmuir approach (i.e., monolayer adsorption) seems to represent better the equilibrium distribution of metal ions between the oxidized wool fibers and the liquid phase. However, both models for the adsorption isotherms for Pb 2+ show important deviation from the experimental values at low equilibrium concentration ( Figure 5) and the corresponding computed χ 2 v values are quite high ( Table 3). The maximum adsorption capacities at 300 K were found to be 9.41, 10.42, and 30.71 mg/g for retention of Cu 2+ , Cd 2+ , and Pb 2+ cations, respectively. These values were found to be substantially greater than the values registered for the adsorption capacities of pristine wool fibers (2.49, 3.87, and 8.66 mg/g for retention of Cu 2+ , Cd 2+ , and Pb 2+ ions, respectively). At 323 K, the maximum adsorption capacities were somewhat higher (Table 3). Table 3. Isotherm models and parameters for adsorption of Cu 2+ , Cd 2+ and Pb 2+ cations onto the oxidized wool fibers; experimental conditions: adsorbent dose = 0.5-9.0 g/L; C 0 = 50 mg/L; t = 240 min, pH 5.0 ± 0.2.

Cation T (K)
Freundlich Isotherm Langmuir Isotherm K F (mg/g)(L/g) 1/n n F χ 2 v q m (mg/g) Further analysis of the Langmuir equation may allow the calculation of the dimensionless parameter (R L ) for equilibrium, also known as the separation factor, which may be expressed by Equation (6) [46,[50][51][52]: where C 0 denotes the initial solute concentration, and K L -Langmuir constant. The calculation of the separation factor (R L ) may offer information on whether the studied adsorption process is favorable or not. Generally, if the separation factor ranges between 0 and 1, then the studied adsorption process is favorable [46,[50][51][52]. In our experiments, the separation factor R L was found to be in the range of 0.017-0.024 (Table 3). Thus, the results pointed out that the adsorption of all three studied metal cations (Cu 2+ , Cd 2+ , Pb 2+ ) onto the oxidized wool fibers was favorable under the experimental conditions considered in this work. In addition, the removal efficiency (Y, %) of metal cations from aqueous solutions by adsorption onto the oxidized wool fibers is shown in Figure 6. Hence, the influence of the adsorbent dose on the removal efficiency was recorded for both temperature levels, i.e., 300 K ( Figure 6a) and 323 K ( Figure 6b). As one can see from Figure 6, the increment of the adsorbent dosage resulted in an increase in the removal efficiency. However, this increase tends to approach a horizontal asymptote; this effect is more evident for the removal of Pb 2+ cations. The highest removal efficiency of metal ions was found to be 96.80 ± 1.69% (for Cu 2+ ); 97.13 ± 1.09% (for Cd 2+ ), and 99.70 ± 0.20% (for Pb 2+ ).
Likewise, a comparative analysis was performed in order to compare the performances of our adsorbent with those registered for other keratin-based materials reported in the literature ( Table 4). As one may see from Table 4, the adsorption performance of metal cations onto the oxidized wool fibers (obtained in this work) is intermediate if compared to other existing wool-based materials. However, the obtaining method of the oxidized wool reported in this work is simple and economically viable, enabling the production of a low-cost and environmental-friendly adsorbent.
adsorption onto the oxidized wool fibers is shown in Figure 6. Hence, the influence of the adsorbent dose on the removal efficiency was recorded for both temperature levels, i.e., 300 K (Figure 6a) and 323 K (Figure 6b). As one can see from Figure 6, the increment of the adsorbent dosage resulted in an increase in the removal efficiency. However, this increase tends to approach a horizontal asymptote; this effect is more evident for the removal of Pb 2+ cations. The highest removal efficiency of metal ions was found to be 96.80 ± 1.69% (for Cu 2+ ); 97.13 ± 1.09% (for Cd 2+ ), and 99.70 ± 0.20% (for Pb 2+ ).

Thermodynamic Adsorption Parameters
To comprehend in more detail the nature of adsorption, one should calculate the basic thermodynamic parameters such as the Gibbs free energy, enthalpy, and entropy [50]. In order to estimate these thermodynamic parameters for the adsorption process, one should consider the equilibrium constants at least for two distinct temperature levels. The thermodynamic parameters of adsorption can be estimated by using the Equations (7)-(9) [46,50,51]: where ∆G (kJ/mol), ∆H (kJ/mol), and ∆S (J/(K·mol)) indicate the changes in Gibbs free energy, enthalpy and entropy of the adsorption process, respectively; T-the absolute temperature (K); R g -the universal gas constant (R g = 8.314 J/(K·mol)), and K d is the equilibrium constant of adsorption. Note that, in this calculation, the equilibrium constant of adsorption was approximated with the Langmuir parameter (K d~KL ) after converting the latter one (K L ) into (L/mol) units. On the basis of the above equations (Equations (7)- (9)), the values of the thermodynamic parameters were calculated and reported in Table 5. As given in Table 5, the positive values of the enthalpy (∆H) for Cd 2+ and Pb 2+ removal indicate endothermic processes, the adsorption of these cations being favored by the temperature increase. For Cu 2+ removal, the negative value of the enthalpy suggested an exothermic process, meaning that a certain amount of heat was released during the adsorption of copper cations onto the surface of the oxidized wool. These different values of the enthalpies of adsorption (∆H) for the cations under consideration are in agreement with the experimental data at equilibrium reported in Figure 5. The positive values of the entropy (∆S) summarized in Table 5 indicate an increase in the degree of freedom or disorder of the adsorbed species (metal cations) at the solid/solution interface. However, the entropy variation was smaller for Cu 2+ removal, suggesting that this system was more stable from a thermodynamic standpoint and reached the equilibrium more rapidly as compared to the processes of Pb 2+ and Cd 2+ removal. The negative values of Gibbs free energy (∆G) indicate the spontaneity of the investigated adsorption processes aiming to remove Cd 2+ , Cu 2+ and Pb 2+ cations from aqueous solutions. In an approximation, the ranges of Gibbs free energy ∆G of (0 to −20 kJ/mol), (−20 to −80 kJ/mol), and (−80 to −400 kJ/mol) might be associated with physical adsorption, ion exchange, and chemisorption, respectively [59]. In this study, the values of ∆G are found between −33.74 and −28.59 kJ/mol suggesting the predominance of the electrostatic interactions (ion exchange) when retaining metal ions onto the oxidized wool fibers. In addition, we calculated the free energy of sorption (Es, kJ/mol) according to Dubinin-Radushkevich (D-R) isotherm. Hence, the free energy of sorption (Es, kJ/mol) was found to lay into the interval 14.78−28.98 kJ/mol. These data suggested that the adsorption mechanism was based on both ion-exchange and chemisorption. Thus, in the proposed Scheme 1c, the most probable route of the mechanism for cation retention is the path based on the electrostatic and coordinating interactions.

Desorption Assay and Re-Use Tests
Additionally, desorption experiments were carried out in batch mode by mixing the spent adsorbent samples (loaded with heavy metal cations) with different types of eluents (water or solutions of HCl, NaOH and NaCl). The results of desorption tests are shown in Figure 7. As seen from this diagram, the used liquid phases to desorb the cations from spent adsorbents can be ordered in a decreasing series of efficiencies as follows: HCl > NaOH ≈ NaCl > H 2 O. Maximum releases of heavy metal cations between 67 and 88% were obtained in 0.1 M HCl electrolyte solution (i.e., the best eluent). Thus, the adsorbed metal ions can be released in greater amounts at acidic pH, except for the cations linked in more stable mercaptide salts [24]. The reversibility of the wool-metal ion interactions is attributed to the competition between metal ions and hydrogen for the same active sites [21]. stable mercaptide salts [24]. The reversibility of the wool-metal ion interactions is attributed to the competition between metal ions and hydrogen for the same active sites [21]. Considering the Figure 7 data, the sorbent samples recovered from HCl were subjected to two additional cycles of adsorption-desorption, following the same protocol as for the initial tests. The obtained results are presented in Figure 8. One may observe that, for all the studied heavy metal cations, the oxidized wool sorbent preserves the adsorption capacity almost unaffected for at least three consecutive cycles.  Considering the Figure 7 data, the sorbent samples recovered from HCl were subjected to two additional cycles of adsorption-desorption, following the same protocol as for the initial tests. The obtained results are presented in Figure 8. One may observe that, for all the studied heavy metal cations, the oxidized wool sorbent preserves the adsorption capacity almost unaffected for at least three consecutive cycles. stable mercaptide salts [24]. The reversibility of the wool-metal ion interactions is attributed to the competition between metal ions and hydrogen for the same active sites [21]. Considering the Figure 7 data, the sorbent samples recovered from HCl were subjected to two additional cycles of adsorption-desorption, following the same protocol as for the initial tests. The obtained results are presented in Figure 8. One may observe that, for all the studied heavy metal cations, the oxidized wool sorbent preserves the adsorption capacity almost unaffected for at least three consecutive cycles.
The raw wool of very coarse grade (fiber diameter, 66.0 ± 6.7 µm) was supplied by a regional sheep farm (Todiresti, Vaslui, Romania). After removing solid macroscopic impurities, the raw wool was washed several times in tap water tempered at 45 • C to remove the lanoline, followed by rinsing with distilled water. Then, the wool fibers were dried in a laboratory oven at 35 • C for 24 h and stored for further use.

Oxidized Wool-Synthesis Protocol
The chemically modified wool was prepared by means of an oxidation process. To this end, a wool sample (raw fibers) was immersed in an alkaline solution, where oxidation was performed in the presence of hydrogen peroxide. The experimental protocol is detailed as follows: to a Berzelius beaker containing 2 L of distilled water, 256 mL of NH 4 OH (25%), and 400 mL of H 2 O 2 (3%), was added 20 g of wool fibers. The mixture was left for 1 h at room temperature to ensure complete contact between solid and liquid phases. Then, the wool-containing solution was heated to about 95 • C for 1 h, under mechanical stirring. Than the oxidized wool fibers were removed from the reaction medium, washed several times with distilled water and twice with absolute ethanol, dried at room temperature for 24 h, and stored in a desiccator for further studies.

Characterization Techniques
The concentrations of the heavy metal cations in the solutions of the adsorption/ desorption experiments were determined by atomic absorption spectrometry (AAS) using the ContrAA 800D, Analytik Jena spectrometer (Jena, Germany).
Fourier-transform infrared (FTIR) spectra of raw and modified materials were recorded by using a Bruker Vertex 70 model FTIR spectrometer (Ettlingen, Germany) in 4000-400 cm −1 wavenumber interval on KBr pellets.
The fibrous morphology and the qualitative composition of the wool-based materials were examined by scanning electron microscopy (SEM) using an (ESCM) Quanta 200 device equipped with an energy-dispersive X-Ray analysis (EDX) module (Brno, Czech Republic).

Adsorption Assays for the Retention of Metal Cations
The adsorption of metal cations was carried out using a Biosan ES-20/60 incubator (Riga, Latvia) equipped with an orbital shaker and a temperature control system. The adsorption kinetics was followed on both single-component systems (i.e., retention of each metal cation separately) and a multi-component system (i.e., competitive retention of three metal cations simultaneously).
The kinetic experiments were performed in batch mode, at a temperature T = 300 K (27 • C) and pH = 5.0 ± 0.1. Thus, in an Erlenmeyer flask containing 50 mL of metal cation solution, 0.25 g of wool fibers were introduced under stirring (150 rpm). The initial concentration of heavy metal cations was fixed at 50 mg/L for the single-solute systems and 20 mg/L for each metal ion in the competitive system (i.e., 60 mg/L summing all three metal ions).
The adsorption isotherms were obtained in batch at various ratios between pollutant concentration and adsorbent dosage. In this respect, initial concentrations of metal cations were fixed at 50 mg/L, while the amount of the oxidized wool ranged from 0.5 to 9.0 g/L.
The adsorption isotherms were recorded at two distinct temperatures of 300 K (27 • C) and 323 K (50 • C). The samples were stirred for 4 h at 150 rpm to reach equilibrium.
In all adsorption experiments, aliquots were taken from the liquid phase at different contact times and after reaching the equilibrium, the content of the heavy metal cations was determined by AAS. The adsorption kinetics was followed by using Equation (10): where q t (mg/g) is the amount of heavy metal adsorbed at time t that takes the maximum value (q e ) at adsorption equilibrium (adsorption capacity of the wool); C 0 and C t (mg/L) are the metal concentrations of the initial solution and of the solution after time t of contact with wool fibers, respectively; V (mL) is the volume of the solution; m (g) represents the weight of the dried wool sorbent. In addition, the removal efficiency, Y (%), was calculated by means of Equation (11): where C e is the residual metal ion concentration at equilibrium.

Desorption Assays of Metal Cations from Oxidized Wool Fibers and Re-Use Tests
Desorption studies were performed in distilled water and different eluents (HCl, NaOH and NaCl solutions of 0.1 M concentrations). Around 0.25 g of oxidized wool sorbent samples, previously exposed to the standard solutions of metal cations (according to the methodology described in Section 3.4 for kinetic experiments), were extracted, dried, and introduced into Erlenmeyer flasks containing 15 mL of each eluent. Desorption tests were performed at a temperature of 27 • C, under stirring (150 rpm), by using a Biosan ES-20/60 incubator, for 1 h. The metal ion concentrations in the desorption solutions were determined by AAS on a ContrAA 800D spectrometer. The desorption efficiency (%) was calculated as a ratio between the amount of metal ion desorbed and the amount of metal ion adsorbed (multiplied by 100) [60][61][62]. The oxidized wool samples recovered after desorption studies using HCl were evaluated for their re-use capacities for two supplementary cycles. Note that after recovering from acid solution, the materials were rinsed with NaOH 0.1 M to neutralize the HCl traces, and finally were washed thoroughly with distilled water.

Conclusions
In this study, we proposed a simple and efficient method for oxidizing wool fibers to obtain a modified keratin-based material with good adsorption properties for the removal of heavy metal ions (Cu 2+ , Cd 2+ , and Pb 2+ ) from wastewater.
Kinetic data indicated that the adsorption equilibrium of metal ions onto oxidized wool fibers was attained at a contact time greater than one hour. The experimental observations were fitted to pseudo-first-order (PFO) and pseudo-second-order (PSO) kinetic models, for both single-component and competitive systems. In the case of the multi-component system (competitive) the results disclosed a higher selectivity for Cd 2+ ions and a lower selectivity for Pb 2+ ions. The reason for this might be attributed to the fact that the heavier ions (Pb 2+ ) can diffuse more slowly toward the adsorbent surface than the lighter ones (Cd 2+ and Cu 2+ ).
The Langmuir and Freundlich models of equilibrium were applied to explore the adsorption isotherms at the two temperatures 300 K and 323 K. The Langmuir isotherm model showed better predictions than the Freundlich model for the studied systems (oxidized wool/heavy metal ions). Thus, the Langmuir equation better accounted for the adsorption of metal ions onto the oxidized wool fibers. At 300 K, the maximum adsorption capacities of oxidized wool fibers for the retention of Cu 2+ , Cd 2+ , and Pb 2+ ions were equal to 9.41, 10.42, and 30.71 mg/g, respectively, more than double or triple times higher than the values corresponding to the retention of the same cations on pristine wool fibers. A higher temperature (323 K) resulted in somewhat greater adsorption capacities. Likewise, the removal efficiency of heavy metal ions from aqueous solutions by using the oxidized wool was about 96.80-99.70%, depending on the type of metal ion (Cu 2+ , Cd 2+ , or Pb 2+ ). The experimental data suggested that the adsorption of Pb 2+ and Cd 2+ ions were favored by the increment in temperature, whereas the adsorption of Cu 2+ ions was disfavored.
According to thermodynamic calculations, the adsorption of Pb 2+ and Cd 2+ ions was accompanied by an endothermic effect, while the adsorption of Cu 2+ ions by an exothermic effect. The negative values of the Gibbs free energy (∆G), ranging from −33.74 to −28.59 kJ/mol, suggested spontaneous adsorption, where the electrostatic interactions (ion exchange phenomena) might play a dominant role.
Desorption assays revealed the maximum releases (67-88%) of heavy metal ions from spent adsorbents in 0.1 M HCl electrolyte solution (the best eluent).