Batch and Fixed-Bed Biosorption of Pb (II) Using Free and Alginate-Immobilized Spirulina

: The valorization of Spirulina as a potential biosorption material to treat contaminated wastewater was evaluated. Batch experiments were conducted to study the inﬂuence of pH value and ionic strength on the biosorption capacity of Spirulina . Higher removal capacity was observed at pH 5.2, while higher ionic strength was found to result in lower adsorption capacity, which suggests that ion exchange is a relevant mechanism for Pb (II) adsorption on Spirulina . The immobilization of Spirulina on alginate beads was found not only to increase the adsorption capacity, but also to overcome limitations such as unacceptable pressure drops on column systems. The Langmuir model was the most appropriate model to describe the biosorption equilibrium of lead by free and immobilized Spirulina . The experimental breakthrough curves were evaluated using the Thomas, Bohart-Adams, and dose-response models. The experimental results were most properly described by the dose-response model, which is consistent with previous results. The adsorption capacity of Spirulina was found to increase linearly with the inﬂuent lead concentration (in the range 4–20 mg L − 1 ) at 1.6 mL min − 1 ﬂow rate. Batch and column experiments were compared to better understand the biosorption process. The promising results obtained indicate the potential use of Spirulina immobilized on alginate beads to treat industrial wastewater polluted with toxic metals.


Introduction
The increasing pollution of the environment by heavy metals due to industrial activity poses serious risks for human health and living organisms. Lead, which is nonbiodegradable and toxic even at low concentrations, is one of the most serious environmental pollutants [1]. The presence of lead in various types of wastewater should be controlled according to the standard for the permitted amount of metal determined by the Environmental Protection Agency [2].
Several conventional technologies have been proposed to remove lead from aquatic environments. Recently, biosorption has been proposed as an emerging and low-cost alternative based on the sorption of dissolved pollutants on a biomaterial. This technology overcomes the most relevant drawbacks of conventional methods, mainly the indirect disposal of toxic metal sludge and the limited adsorption efficiency at low metal concentrations [3]. Biomaterials, such as bacteria, algae, fungi, and agricultural wastes, have been proposed as low-cost biosorbents for the removal of heavy metals from wastewater [4]. Among these biomaterials, algae have been widely suggested as ideal biosorbents for the removal of toxic metals from water effluents [5]. The presence of negatively charged groups on its surface, such as amino, hydroxyl, carboxyl, sulfhydryl, and sulfonate, allows the binding of heavy metals on algal biomass [6].
Brouers and Al-Musawi proposed the use of a fresh mixture of green and blue-green algae as a biosorbent for lead removal contained in aqueous solutions. They concluded that algal biomass is a promising biosorbent with a maximum removal efficiency of 98% MO, USA) dissolved in deionized water (25 mL). The resulting solution was heated to 85 • C for 15 min. With the aim of obtaining beads with diameters ranging between 1.5 and 2.0 mm, the mixture was dropped through a syringe with an internal diameter of 2 mm into a solution of 0.5 M CaCl 2 . After 24 h, the solution was rinsed with deionized water and immersed in a solution of 0.5 M HCl for more than 24 h. Finally, the beads were washed again with deionized water.

Batch Biosorption Studies
Batch experiments were proposed to evaluate the adsorption of Pb (II) on free and immobilized Spirulina. The experiments with free biosorbent were conducted by adding a known mass of Spirulina to 1 L of aqueous Pb(NO 3 ) 2 solutions of different concentrations. Following the guidelines published by the Organization for Economic Cooperation and Development (OECD) [30], the solid to liquid ratio (S/L) was 50 mg L −1 . The samples were stirred in an orbital shaker at 120 rpm at 25 • C. After completion, the mixture was filtered using Whatman GF/C filters (pore size 0.6 µm, Whatman, Maidstone, UK) through vacuum filtration. Each experiment was performed in triplicate using the same operational conditions. The effect of the initial pH of the solutions, i.e., 2, 3, 4, 4.5, and 5.23 (the pH value of the original Pb solution, not adjusted), was evaluated using Pb (II) solutions with initial concentrations in the range 0.6 to 5.6 mg L −1 . For isotherm studies, the initial Pb (II) concentration was varied from 0.6 to 5.6 mg L −1 at pH 5.23 for 72 h. The effect of ionic strength on the biosorption of Pb (II) by Spirulina was evaluated by adding 0.01 and 0.1 M NaNO 3 . The concentration of Pb (II) was varied from 0.6 to 5.6 mg L −1 to obtain isotherms with different ionic strengths.
The behavior of Spirulina immobilized on alginate gel was also evaluated by performing batch experiments. The essays were conducted by immersing 35 mg of alginate or Spirulina immobilized on alginate gel in 100 mL of aqueous Pb(NO 3 ) 2 solutions of different concentrations (from 20 to 150 mg L −1 ) until the equilibrium was reached. The final concentration of Pb (II) in the aqueous solution was analyzed by atomic absorption spectroscopy (AAS) (Varian SpectrAA-110, Varian, Palo Alto, CA, USA).
With the aim of determining the total content of Pb (II) in solid samples, microwaveassisted acid digestion was conducted. The concentration of Pb (II) was also determined by AAS (Varian SpectrAA-110). Each experiment was performed in triplicate under the same experimental conditions. The biosorbent capacity and the removal percentage of Pb (II) were obtained as follows: where q (mg g −1 ) is the amount of adsorbed lead ions, C 0 (mg L −1 ) is the initial concentration of lead, C f (mg L −1 ) is the final lead concentration, V (L) is the volume of the metal solution, and W (g) is the weight of the biosorbent.

Column Biosorption Experiments
Experiments in continuous systems, shown in Figure 1, were carried out using a glass column with a length of 3.7 cm and an internal diameter of 3.1 cm. The alginateimmobilized Spirulina mass contained in the column was of 10 g. The synthetic solutions containing Pb was circulated with a flow rate of 1.6 mL min −1 through the column using a peristaltic pump. Samples were collected at selected times until bed saturation.
lized Spirulina mass contained in the column was of 10 g. The synthetic solutions contai ing Pb was circulated with a flow rate of 1.6 mL min −1 through the column using a per staltic pump. Samples were collected at selected times until bed saturation. The adsorption capacity was calculated from the breakthrough curve as: is the amount of metal ions in the treated solution, Q (mL min −1 ) is th volumetric flow rate, W (g) is the dry weight of the biosorbent, and te (min) is the operatio time of the column when C/C0 = 1. The initial concentration of Pb was varied from 4 to 20 mg L −1 to evaluate the influ ence of the metal concentration on column performance. The initial and final concentr tions of Pb in aqueous solutions were determined by AAS (Varian SpectrAA-110).
where qmax_L (mg g −1 ) indicates the maximum amount of metal ions per unit mass of bi mass at equilibrium, and kL (L mg −1 ) is the equilibrium adsorption constant.
where qmax_DB (mg g −1 ) is the maximum adsorption capacity, β (mol 2 J −2 ) is a constant ass ciated with the adsorption energy, and ε (J mol −1 ) is Polanyi potential obtained as: The value of the mean adsorption energy, E (J mol −1 ), is obtained as: The adsorption capacity was calculated from the breakthrough curve as: where C (mg L −1 ) is the amount of metal ions in the treated solution, Q (mL min −1 ) is the volumetric flow rate, W (g) is the dry weight of the biosorbent, and t e (min) is the operation time of the column when C/C 0 = 1. The initial concentration of Pb was varied from 4 to 20 mg L −1 to evaluate the influence of the metal concentration on column performance. The initial and final concentrations of Pb in aqueous solutions were determined by AAS (Varian SpectrAA-110).
q e = k f C 1/n e (4) where q e (mg g −1 ) is the amount of lead adsorbed per unit weight of biosorbent at equilibrium with a given solution concentration C e (mg L −1 ), k f (mg (n−1)/n L 1/n g −1 ) is the adsorption capacity, and n (-) is a parameter related to the adsorption intensity.
where q max_L (mg g −1 ) indicates the maximum amount of metal ions per unit mass of biomass at equilibrium, and k L (L mg −1 ) is the equilibrium adsorption constant.
where q max_DB (mg g −1 ) is the maximum adsorption capacity, β (mol 2 J −2 ) is a constant associated with the adsorption energy, and ε (J mol −1 ) is Polanyi potential obtained as: where R is the universal gas constant (8.314 J mol −1 K −1 ). The value of the mean adsorption energy, E (J mol −1 ), is obtained as: The value of E is widely used to make a prediction about the nature of the adsorption processes, either physical or chemical [34].

Sorption Dynamics in Fixed-Bed Columns
The understanding of the sorption phenomena in a fixed column is required for design and operation of a full-scale adsorption process. It should be noted that the description of the sorption in a continuous process is not straightforward due to the importance of axial dispersion, sorption kinetics, intraparticle diffusion resistance, and mass transfer. Therefore, several mathematical models have been reported to predict breakthrough behaviors and to calculate the design parameters for fixed-bed column adsorption.
The Thomas and Bohart-Adams models are frequently used to predict the dynamic behavior in fixed-column processes. Thomas' equation is obtained from a Langmuir adsorption equilibrium system with a pseudo second-order reaction kinetic law and without axial dispersion [35]: where C and C 0 (mg L −1 ) are the effluent and influent lead concentration, k TH (mL mg −1 min −1 ) is the Thomas rate constant, q TH (mg g −1 ) is the adsorption capacity, W (g) is the mass of biosorbent, Q (mL min −1 ) is the flow rate, and V eff (L) the volume of the effluent. The Bohart-Adams model was founded on the surface rate theory and supposes that equilibrium is not achieved instantaneously. This model is usually applied to describe the initial part of the breakthrough curve [36]. The equation can be expressed as: where k AB (L mg −1 min −1 ) is the rate constant, N 0 (mg L −1 ) is the saturation concentration, Z (cm) is the bed depth, and ν (cm min −1 ) is the linear velocity obtained by dividing the flow rate by the column section area. The modified dose-response model [37] was also implemented to evaluate experimental values. This empirical model focuses on the description of the kinetics of metal removal in a biosorption column and minimizes the error generated by previous models, especially for lower and higher times of the breakthrough curve.
where a (-) and b (L) are the model parameters.

Mathematical Model Parameter Determination
With the aim of determining all model parameters, a non-linear square method was applied using Solver-add MS-Excel. The differences between values predicted by the model and the experimental values were estimated by the root mean square errors (RMSEs) as: where m is the number of data points and X exp and X model are the experimental and the predicted values, respectively, of the adsorption capacity, q (mg g −1 ), or the fractional concentration, C/C 0 , for isotherm or fixed-bed adsorption models, respectively. The minimization of RMSE values with simultaneous variation of model parameters allowed the fit of the experimental data to the non-linear forms of the equations. With the purpose of comparing the results obtained from different models, the coefficient of determination (R 2 ) of parity plots (i.e., experimental versus model values) was determined.

Effect of Initial pH of the Solution
The effect of the pH value of the aqueous solution on the biosorption of Pb (II) ions on Spirulina was evaluated. The changes of Pb (II) adsorption capacity at different initial pH values (from 2 to 5.2) are depicted in Figure 2. As can be seen, the Pb (II) biosorption capacity of Spirulina is strongly dependent on pH value. When the solution pH was below 3, the biosorption capacity was significantly lower. These results are in agreement with the competition between protons and Pb (II) for adsorption sites, and with the increase in electrostatic repulsion between positively charged species. At these pH values, Pb is mainly present as Pb 2+ [38]. With increasing pH, the biosorption capacity of Pb (II) on Spirulina increased. These results have been previously related to the occurrence of more ligands carrying negative charges, such as carboxyl groups, which entails higher attraction of metal ions and biosorption onto biosorbent surfaces [39]. According to the experimental results, no significant differences in Pb (II) removal were observed for pH values from 4.5 to 5.2. As 5.2 was the pH of the original aqueous Pb (II) solution, this value was selected for further experiments. Previous studies dealing with the biosorption of Pb (II) have found similar optimal pH values [40][41][42].
predicted values, respectively, of the adsorption capacity, q (mg g −1 ), or the fr centration, C/C0, for isotherm or fixed-bed adsorption models, respectively.
The minimization of RMSE values with simultaneous variation of mode allowed the fit of the experimental data to the non-linear forms of the equati purpose of comparing the results obtained from different models, the coeffic mination (R 2 ) of parity plots (i.e., experimental versus model values) was de

Effect of Initial pH of the Solution
The effect of the pH value of the aqueous solution on the biosorption o on Spirulina was evaluated. The changes of Pb (II) adsorption capacity at di pH values (from 2 to 5.2) are depicted in Figure 2. As can be seen, the Pb (II capacity of Spirulina is strongly dependent on pH value. When the solution p 3, the biosorption capacity was significantly lower. These results are in agr the competition between protons and Pb (II) for adsorption sites, and with th electrostatic repulsion between positively charged species. At these pH mainly present as Pb 2+ [38]. With increasing pH, the biosorption capacity Spirulina increased. These results have been previously related to the occurr ligands carrying negative charges, such as carboxyl groups, which entails h tion of metal ions and biosorption onto biosorbent surfaces [39]. According mental results, no significant differences in Pb (II) removal were observed f from 4.5 to 5.2. As 5.2 was the pH of the original aqueous Pb (II) solution, th selected for further experiments. Previous studies dealing with the biosorpt have found similar optimal pH values [40][41][42].

Effect of Ionic Strength on Pb (II) Biosorption
The

Effect of Ionic Strength on Pb (II) Biosorption
The influence of ionic strength on Pb (II) biosorption was evaluated by adding solutions of 0.01 and 0.1 M NaNO 3 at pH 5.2. The experimental results, presented in Figure 3  Changes in the ionic strength not only affect the interfacial po quently, the activity of electrolyte ions, but also the competition betw ions. According to the experimental results, it could be concluded th Pb ions. Previous studies have explained the effect of ionic strength competition of cations added as salts with heavy metals. In other w strength, adsorption sites are covered by counter-ions, which entail a consequently, this weakens the binding forces due to electrostatic inte results suggest that ion exchange is the most relevant mechanism for on Spirulina, as discussed in previous work carried out with other alg actions, i.e., complexation/coordination, electrostatic interactions, ch sisorption, microprecipitation, and reduction, can also coexist with ion

Adsorption Isotherms
The Freundlich, Langmuir, and Dubinin-Radushkevich isotherm plied to study the equilibrium data. The experimental results were s linear forms of these model equations, as presented in Figure 4. As can these three models properly describe the biosorption process with hig cients and low RSME values. The Langmuir isotherm shows a sligh correlation coefficient (R 2 ) of 0.992. Pb (II) biosorption can be regarded monolayer adsorption, as concluded in previous studies dealing with als onto algae [44][45][46]. The maximum adsorption capacity accordin model was 114.47 mg g −1 , which is comparable to the results from oth The Dubinin-Radushkevich isotherm allows the description of the na tion process through the estimation of the free energy of sorption. From value of the free energy obtained was higher than 8 kJ mol −1 , which h Changes in the ionic strength not only affect the interfacial potential and, subsequently, the activity of electrolyte ions, but also the competition between the electrolyte ions. According to the experimental results, it could be concluded that Na ions displace Pb ions. Previous studies have explained the effect of ionic strength as an electrostatic competition of cations added as salts with heavy metals. In other words, for high ionic strength, adsorption sites are covered by counter-ions, which entail a loss of charge, and, consequently, this weakens the binding forces due to electrostatic interactions [39]. These results suggest that ion exchange is the most relevant mechanism for Pb (II) biosorption on Spirulina, as discussed in previous work carried out with other algae [43]. Other interactions, i.e., complexation/coordination, electrostatic interactions, chemisorption, physisorption, microprecipitation, and reduction, can also coexist with ion exchange [4].

Adsorption Isotherms
The Freundlich, Langmuir, and Dubinin-Radushkevich isotherm models were applied to study the equilibrium data. The experimental results were simulated with non-linear forms of these model equations, as presented in Figure 4. As can be seen in Table 1, these three models properly describe the biosorption process with high correlation coefficients and low RSME values. The Langmuir isotherm shows a slightly better fit with a correlation coefficient (R 2 ) of 0.992. Pb (II) biosorption can be regarded as more likely with monolayer adsorption, as concluded in previous studies dealing with biosorption of metals onto algae [44][45][46]. The maximum adsorption capacity according to the Langmuir model was 114.47 mg g −1 , which is comparable to the results from other studies ( Table 2). The Dubinin-Radushkevich isotherm allows the description of the nature of the biosorption process through the estimation of the free energy of sorption. From the value of β, the value of the free energy obtained was higher than 8 kJ mol −1 , which has been associated with chemical adsorption of Pb on Spirulina's surface [34]. These results are consistent with previous studies dealing with Spirulina as a biosorbent [18].    Batch experiments were carried out with alginate gel and alginate-immobilized Spirulina to evaluate the isotherm models, i.e., Freundlich, Langmuir, and Dubinin-Radushkevich. The equilibrium isotherm parameters for the biosorption of Pb (II) under these experimental conditions are presented in Table 3. As can be seen, the three selected models fit properly to the experimental data. According to the Langmuir model, the maximum adsorption capacity was 303.94 and 282.17 mg g −1 for alginate and immobilized Spirulina, respectively. These values are consistent with other studies dealing with alginate as a matrix of a natural polymer to immobilize algal biomass [25]. The higher values obtained for biosorption capacity corroborate the potential of alginate as biosorbent in Pb (II) removal. Regarding the value of the free energy obtained from the Dubinin-Radushkevich parameters, it could be concluded that the adsorption of Pb is of a chemical nature in agreement with the results obtained for free Spirulina.

Fixed-Bed Column Studies
The breakthrough curve obtained for a flow rate of 1.6 mL min −1 and for an inlet metal ion concentration of 20 mg L −1 is presented in Figure 5. Initially, most of the Pb (II) bound close to the inlet of the column, while the downstream sections were not exposed to the metal. The shape of this curve depends on several experimental factors, such as inlet flow rates, concentrations, and other properties of the column, i.e., diameter and bed height. The section of the bed in which the concentration of Pb (II) changes importantly is known as the mass transfer zone (MTZ). The shape of this section depends on the equilibrium equation, the axial dispersion in the reactor, and the kinetics of adsorption. A square-wave curve is the ideal shape of a breakthrough, since the column would allow the largest volume of contaminated water to be treated, and the maximal amount of metal would be removed during operation. Therefore, the shape of the breakthrough curve experimentally obtained could indicate that the equilibrium between the adsorbent and the surrounding solution was not reached instantaneously [50].  The mathematical description of this curve is required not only for designing a fixedbed adsorber, but also to predict the behavior of the dynamic system. The estimated parameters obtained by non-linear regression analysis are presented in Table 4. Neither the Thomas nor the Bohart-Adams model adequately reproduced the experimental data, giving R 2 values of 0.907 and 0.849, respectively. As can be concluded from the RSME and R 2 values (Table 4), the dose-response model shows the best fit to the experimental data. These results are in accordance with previous studies that found that this model minimizes the most relevant errors obtained from other models, particularly at low and high values of the operation time [51].
The dependence of the adsorption capacity of immobilized Spirulina on alginate on the initial metal concentration of Pb (II) is shown in Figure 6. According to the experimental results, the biosorption capacity increased linearly with an increase in the initial Pb (II) concentration. As the system is characterized by a Langmuir isotherm, the maximum value of adsorption capacity could be determined from model parameters presented in Table 3. The maximum adsorption capacities (q*) for initial concentrations of 4 and 20 mg L −1 were 243.91 and 273.51 mg g −1 , respectively, while the experimental adsorption capacity (q) was 47.98 and 202.85 mg g −1 , respectively. These results indicate that sorption was controlled by the mass transfer for the flow rate studied. In other words, the metal The mathematical description of this curve is required not only for designing a fixedbed adsorber, but also to predict the behavior of the dynamic system. The estimated parameters obtained by non-linear regression analysis are presented in Table 4. Neither the Thomas nor the Bohart-Adams model adequately reproduced the experimental data, giving R 2 values of 0.907 and 0.849, respectively. As can be concluded from the RSME and R 2 values (Table 4), the dose-response model shows the best fit to the experimental data. These results are in accordance with previous studies that found that this model minimizes the most relevant errors obtained from other models, particularly at low and high values of the operation time [51].
The dependence of the adsorption capacity of immobilized Spirulina on alginate on the initial metal concentration of Pb (II) is shown in Figure 6. According to the experimental results, the biosorption capacity increased linearly with an increase in the initial Pb (II) concentration. As the system is characterized by a Langmuir isotherm, the maximum value of adsorption capacity could be determined from model parameters presented in Table 3. The maximum adsorption capacities (q*) for initial concentrations of 4 and 20 mg L −1 were 243.91 and 273.51 mg g −1 , respectively, while the experimental adsorption capacity (q) was 47.98 and 202.85 mg g −1 , respectively. These results indicate that sorption was controlled by the mass transfer for the flow rate studied. In other words, the metal was not able to reach all positions of the biosorbent surface in a time shorter than residence time, in agreement with the conclusions from the breakthrough curve shape ( Figure 5).
Processes 2021, 9,466 was not able to reach all positions of the biosorbent surface in a time shorter than res time, in agreement with the conclusions from the breakthrough curve shape (Figur

Conclusions
Immobilized Spirulina on alginate beads was proposed as an effective biosorb lead removal in batch and fixed-bed columns. From batch experiments, the bioso capacity of Spirulina was concluded to be strongly dependent on pH value. The e mental results suggested that ion exchange is probably the most relevant mechani Pb (II) sorption on Spirulina. The Langmuir model properly described the isotherm of lead biosorption by free and immobilized Spirulina, indicating a monolayer so onto the biomass surface. The potential of Pb (II) adsorption from Spirulina using bilized biomass in a continuous fixed-bed column was assessed. The breakthrough was applied to study the efficiency of the column using immobilized Spirulina. The response model described better the experimental data than both the Thomas and t hart-Adams models, which is associated with the minimization of relevant errors ob at low and high values of the operation time. The importance of the influent met centration was also evaluated, concluding that the adsorption capacity of Spirul creased linearly with the initial metal concentration. From the comparison of bat column studies, it was found that equilibrium between Spirulina and the solution w instantaneously reached in the dynamic studies. The promising results obtained f mobilized Spirulina on alginate beads could be the first step of the valorization o biomass for the treatment of industrial wastewater contaminated by toxic metals s Pb.

Conclusions
Immobilized Spirulina on alginate beads was proposed as an effective biosorbent for lead removal in batch and fixed-bed columns. From batch experiments, the biosorption capacity of Spirulina was concluded to be strongly dependent on pH value. The experimental results suggested that ion exchange is probably the most relevant mechanism for Pb (II) sorption on Spirulina. The Langmuir model properly described the isotherm data of lead biosorption by free and immobilized Spirulina, indicating a monolayer sorption onto the biomass surface. The potential of Pb (II) adsorption from Spirulina using immobilized biomass in a continuous fixed-bed column was assessed. The breakthrough curve was applied to study the efficiency of the column using immobilized Spirulina. The dose-response model described better the experimental data than both the Thomas and the Bohart-Adams models, which is associated with the minimization of relevant errors obtained at low and high values of the operation time. The importance of the influent metal concentration was also evaluated, concluding that the adsorption capacity of Spirulina increased linearly with the initial metal concentration. From the comparison of batch and column studies, it was found that equilibrium between Spirulina and the solution was not instantaneously reached in the dynamic studies. The promising results obtained for immobilized Spirulina on alginate beads could be the first step of the valorization of algal biomass for the treatment of industrial wastewater contaminated by toxic metals such as Pb.