Optimization of synergistic biosorption of oxytetracycline and cadmium from binary mixtures on reed-based beads: modeling study using Brouers-Sotolongo models

1 Optimization of synergistic biosorption of Oxytetracycline and Cadmium from binary mixtures on reed-based-beads: modeling study using BrouersSotolongo models Abstract The first aim of this study was to synthesize and characterize reed-based-beads (BBR), an enhanced adsorbent from Tunisian-Reed. The second purpose was to evaluate and optimize the BBR efficiency for the simultaneous removal of oxytetracycline (OTC) and Cadmium (Cd(II)), using Central Composite Design under Response Surface methodology. The third goal was to elucidate the biosorption mechanisms taking place. It was shown that under optimum conditions (4.19 g L-1 of BBR, 165.54 µmol L-1 of OTC, 362.16 µmol L-1 of Cd(II), pH of 6, and 25.14 h contact time) the highest adsorption-percentages (63.66 % for OTC and 99.99 % for Cd(II)) were obtained. It was revealed that OTC adsorption mechanism was better described by Brouers-Sotolongo fractal equation, with regression coefficient (R2) of 0.99876, and a Person’s Chisquare (X2) of 0.01132. The Weibull kinetic equation better explained Cd(II) biosorption (R2=0.99959 and X2=0.00194). FTIR and isotherm studies confirmed that the BBR surface was heterogeneous, and that adsorption mechanisms were better described by the Freundlich/Jovanovich equation (R2=0.99276 and X2=0.04864) for OTC adsorption, and by the Brouers-Sotolongo model (R2=0.98510 and X2=0.77547) for Cd(II) biosorption. Overall results indicate that, at last, the BBR lignocellulosicbiocomposite-beads could be considered as cost-effective and efficient adsorbent, which could be of socioeconomic and environmental relevance. Introduction Nowadays, large amounts of medicines (including antibiotics), after being discharged by pharmaceutical industries can be found in the aquatic environment [1]. Antibiotics are also excreted and discharged into water courses as unmetabolized compounds directly within feces and urine [2], thus menacing human beings and ecosystems [3]. In particular, Oxytetracycline (OTC) is an antibiotic very relevant in the treatment of various bacterial infections and animal diseases [4]. Because of its broad antimicrobial activity spectrum, OTC is also widely used as an anti-microbial additive [5,6]. In addition to antibiotics, “heavy metals” are directly and/or indirectly discharged into the environment, by mining operations, metal plating facilities, fertilizer and battery industries, 2 causing serious environmental pollution [7]. Cadmium (Cd(II)) is considered one of the most hazardous and toxic within these substances. It can cause kidney damage, high blood pressure and destruction of red blood-cells and testicular tissue [8]. Thus, removing this contaminant from water is an urgent task. Antibiotics and heavy metals are mutagenic and carcinogenic, and not easily biodegradable. Therefore, they present a major danger to the environment and for human beings [3,8]. In addition, OTC and Cd(II) often coexist in soils and wastewater. Due to the existence of this environmental and public health issue, various biological, physical and chemical techniques have been used to remove these pollutants. Examples of these process are membrane filtration [9–11], coagulation-flocculation [12,13], photocatalytic degradation [14,15], chemical oxidation [16,17], ion exchange [18], electrochemical degradation [12,19], sonocatalytic degradation [20,21], and adsorption [22,23]. Among all of them, adsorption was considered as the most appropriate method to remove pollutants from wastewater, because of its rapidity, simplicity, cost effectiveness, high efficiency, and eco-friendliness [24]. Additionally, there was a wealth of study on the single adsorptive removal of several kinds of contaminants, such as dyes [25], antibiotics [26], and heavy metals [27]. In the case of binary systems, But, most researches focused on the removal of pollutants of the same type However, in spite of the common existence of antibiotics and heavy metals in soils and wastewaters, there is a dearth of works studying their simultaneous removal from binary solution. Specifically, there is a lack of works studying OTC and Cd(II) binary adsorption, when compared with other pollutants. Furthermore, various adsorbents have been investigated to remove pollutants from wastewater, mainly activated carbon [28], cinnamon soil [29], biomaterials [25,30,31], and many other adsorbents, some of them considered as not highly effective. However, it is not easy to find an efficient adsorbent with potential to retain two different kinds of contaminants at the same time. Therefore, great efforts have been made to produce new adsorbents with high efficiency for binary contaminants removal. This study is the first comprehensive work that assessed the efficacy of Reed-based biocomposite for the simultaneous adsorption of the antibiotic OTC, and the heavy metal Cd(II), from water. 3 Therefore, this paper reported a two-fold aim. (a) The first objective was to produce a novel, cost-effective and environment-friendly biocomposite beads based on Tunisian Reed (Phragmites australis) to be used as an adsorbent. This reed is an invasive plant that flourishes widely in marshes in the Gabes region (Tunisia). (b) The second objective was to evaluate the efficiency and capability of this biosorbent to remove heavy metals (specifically Cd) and antibiotics (specifically OTC) from wastewater simultaneously. (c) The third objective of this study was to evaluate the effect of various biosorption parameters using the CCD (central composite design) methodology. In this regard, compared to traditional single factor testing approaches, the designs approaches are clearly advantageous, because they could be used to model and optimize the effects of various variables with a reduced number of experiments. CCD-RSM (central composite design under response surface methodology) is considered the most efficient methodology to precisely analyze the correlation between dependent and independent variables [32]. It can perform experiments under extreme variables value conditions, guaranteeing accurate results. (d) The fourth goal in this work was to model biosorption kinetics and isotherms data by nonlinear fitting using Brouers-Sotolongo models. The Brouers-Sotolongo kinetic and isotherm models take into consideration the complexity of the biosorption process, and can be used to elucidate the biosorption mechanism. The Brouers-Sotolongo equations have been used in various previous studies [30,33,34], but, to our knowledge, this is the first comprehensive work that has assessed the synergetic CCD optimization of simultaneous antibiotics and heavy metals biosorption and data modelling using Brouers-Sotolongo family equations. In this way, batch biosorption experiments were performed for OTC and Cd(II), to examine the adsorption thermodynamics, kinetics and isotherms data. Finally, possible biosorption mechanisms were also suggested. Overall, the results of these studies could be of relevance as regards treatments to reduce environmental and public health issues related to the simultaneous presence of OTC and Cd(II) in environmental compartments. Materials and methods Reagents Oxytetracycline (OTC, 95 % oxytetracycline hydrochloride (C22H24N2O9HCl)), and Cadmium (Cd(II), Cadmium nitrate tetrahydrate (CdN2O6.4H2O)) were supplied by Sigma-Aldrich (country). In addition, 1-Butyl-3-methylimidazolium chloride (BMIMCl) was purchased from abcr GmbH(country). All mentioned reagents were used without supplementary purification. 4 Solutions of OTC and Cd(II), at concentrations of 400 µM and 1600 µM, respectively, were prepared and stocked. Afterward, the working solutions were prepared daily by appropriate dilution. Apparatus The quantification of OTC in the suspension was carried out on a HPLC liquid chromatograph (Dionex Corporation, Sunnyvale, USA). The injection volume and the flow rate were 50 µL and 1.5 mL min-1, respectively. The mobile phase consisted of acetonitrile (phase A) and 0.02 mol L-1 oxalic acid/0.01 mol L-1 triethylamine (phase B). The wavelength used for detection of OTC was 360 nm. The quantification of Cd(II) was carried out using Inductively Coupled Plasma equipped with Optical Emission Spectrometer (ICP-OES) (PerKin Elmer Optima 4300 DV, USA). To designate the functional groups involved in the sorption of OTC and Cd(II) by the BBR beads, infrared spectra were obtained for the biosorbent, before and after contact with OTC and Cd(II) pollutants, using a Fourier Transform Infrared Spectrophotometer equipped with Attenuated Total Reflection Method (FTIR-ATR, PerkinElmer, USA). The baseline correction related to the atmospheric air was repeated for each sample. The spectral resolution was 2 cm-1 and four scans were realized for each sample, with 450-4000 wave-numbers range. To observe the surface morphology of biosorbent, a scanning electron microscope (SEM) equipped with energy dispersive X-ray (EDX) (ESEM EVO LS15, Zeiss, Germany) was used. SEM-EDX allows quantifying the surface chemical composition before and after sorption. Thermo-gravimetric measurement of BBR was performed using thermogravimetric analyzer (TGA, 4000 PerkinElmer, USA), under N2 atmosphere, with heating rate of 10 °C min-1, to determine the BBR thermal stability. The sample was heated from 30 to 900 °C. The amount of the BBR sample was about 4.813 mg. The specific surface area of BBR was determined using the N2-BET (Nitrogen-BrunauerEmmett-Teller) adsorption test (Micrometrics Instrument Corp., USA). Bio-sorbent preparation The biocomposite beads of Reed (referred to as BBR) were synthesized using the Karoui et al.'s protocol [30], as follows: The Tunisien reed (R) was cleaned, dried and sieved to obtain the reed powder. In the present work the used biosorbent is composed of regenerated Tunisien reed (RTR) and charcoal of Tunisien reed (ChTR) with equal quantities. On the one hand, 5 after treated the reed powder with the orthophosphoric acid (H3PO4) solution (1M), the obtained mixture was neutralized until pH = 6.95 ± 0.5, and the powder was carbonized at 250 °C to obtain the ChTR. On the other hand, the reed was dissolved in BMIMCl, and then the ChTR was added [30]. Afterwards, the viscous and homogeny mixture was added drop-wise into a distilled water bath using pipette with a diameter of 3 mm. , In a further step, the obtained beads were thoroughly rinsed with distilled water until disappearance of BMIMCl (control realized by electric conductivity and pH measurements). Finally, the BBR were frozen at T = -17 °C, to prepare it for lyophilization. The formed beads were stocked in desiccators before other use. Adsorption Experiments Batch binary biosorption experiments were carried out in 50 mL glass bottles. All solutions were stirred at 150 rpm by the mean of horizontal shaker at room temperature. The central composite design (CCD) methodology was used to obtain the optimum sorption conditions and to study the effect of various factors on pollutants removal efficiency. Then, the temperatures (20–50±1°C) were chosen to perform and discuss the thermodynamic study for OTC and Cd(II) binary sorption onto BBR beads. Percentage removal (%R) and the adsorbed amount qe at equilibrium were determined using Eq(1) and Eq(2), respectively [35]: qe =(C0 – Ce)V/m (1) % R= 100(C0-Ce)/C0 (2) Where, C0 and Ce are the initial and the equilibrium concentrations (mg L-1), respectively; V is the solution volume (L); m is the mass of biosorbent (g). Error analysis To designate the best kinetic or isotherm equation applicable in the current study, non-linear correlation coefficients (R2), chi-square test (χ2), and residual sum of squares (RSS) were evaluated as alternatives for informing on the sorption process. Table 1 shows the used equations to calculate the R2, χ2 and RSS values. Table 1. Experimental design and statistical treatment 6 STATISTICA12.0 software was used to analyze the experimental data, obtained by central composite design under response surface methodology (CCD-RSM). Both OTC and Cd(II) removals were determined by using batch-type binary adsorption experiments. The CCDRSM methodology was used to optimize the experimental conditions and minimize the number of experiments [30]. The effect of all tested parameters and the interactions among the controlled factors on the biosorption percentage were evaluated using CCD. [36] In this paper, five factors including the adsorbent dosage (F1: 2 –4 g L-1), the OTC concentration (F2: 175-275 µmol L-1), the Cd(II) concentration (F3: 400-800 µmol L-1), the pH (F4: 4-8) and the reaction time (F5: 12-24 hours), were evaluated to determine the optimum sorption conditions. The used low values of adsorbent dosage were chosen to save biomaterials. In addition, taking into account that the concentrations of antibiotic and heavy metal detected in wastewater were low, narrow ranges of OTC and Cd(II) concentrations were chosen. The experimental design was performed in a large range of pH (acid and basic media). Some preliminary tests showed that the OTC and the Cd(II) biosorption process onto BBR was slow. Therefore, the time range estimated in this work was 12-24 h. A number of 47 experiments were carried out under non-factorial central composite rotatability design. This number is the sum of 5 center points (replicates), 10 axial (star) points, and 32 factorial (cube) points. Each factor was coded at 5 levels, which are the lowest (-α), low (-1), the center point (0), high (+1) and the highest (+α) levels. Table 2 represents the value corresponding to each level for both OTC and Cd(II) contaminants. Table 2. For a central rotatable composite design, the axial distance α (distance between the star point and the design center) was expressed by Eq. (6): 𝛂𝛂= [nc]1 4 (6) where 𝑛𝑛𝑐𝑐 stands for the number of cube points in the design. The developed empirical model to correlate the response (adsorption percentage) to the five studied independent variables was given by the following mathematical relation (Eq. (7)), which is based on second-order quadratic model. 7 Y = b0+∑biFi k i=1 + ∑biiFiFi k i=1 + ∑ ∑ bijFiFj k j=i+1 k i=1 (7) Where, Y is the predicted dependent variable (predicted adsorption percentage); b0 is the model constant; bi is the linear coefficient; bij is the squared (quadratic) coefficient; bij is the cross-product coefficient; and Fi, Fj are the independent variables. The experimental design points and the findings for binary antibiotic and heavy metal adsorption were illustrated in Table S1. The five central points show the experimental data reproducibility. Table S1 (Supplementary Material) Kinetic models OriginPro 8.5 software for Windows was used to carry out processing of the non-linear data and the derivatives spectra. The kinetics Brouers-Sotolongo family model, based on the BurrXII statistical distribution [37], was the most developed fractal theory applied to study the kinetics sorption of pollutants onto porous adsorbents [38].The statistical macroscopic kinetic model BSf (n,α) widely integrates irreversibility, fractal adsorption and diffusion, and the sorption-desorption mechanism nature of the process [30]. The BSf (n,α) was expressed by Eq (8) and Eq (9) (Table. 3). Five approximate models can be obtained from the general BSf model (n,α) by giving well-defined values to n and α. The five equations are indicated in Table. 3. Table 3. In this work, the adsorption kinetic results of the OTC and Cd(II) uptake onto BBR were depicted using Brouers-Sotolongo-fractal BSf equation (n, α), as was described in various previous papers [30,33,34]. The pseudo˗first˗order model (PFO; n = 1 and α = 1), the pseudo˗second˗order model (PSO; n = 2 and α = 1), the Weibull kinetic (n =1 and α ≠ 1), the Hill kinetic (n = 2 and α ≠ 1), and the Brouers˗Gaspard kinetic (BG; n = 1.5) were employed to study the binary biosorption of organic and inorganic substances onto the BBR in liquid phase and to analyze the experimental data obtained. Adsorption isotherms The equilibrium adsorption isotherm is based on the relation between the equilibrium concentration in the liquid˗phase (Ce) and the amount adsorbed at equilibrium per unit of mass (qe) [25]. 8 The General˗Brouers˗Sotolongo (GBS) equation is expressed by Eq (15) and Eq (16) (Table. 4), and 𝐶𝐶𝑒𝑒1 2(50 % of Ce), that depends on the “a”, “b” and “c” constants, was calculated using Eq (17) (Table. 4). Six approximate models can be obtained from the GBS equation by giving well-defined values to “a” and “c” parameters. The six equations are indicated in Table. 4. Table 4. In this work, the GBS, the normal Brouers˗Sotolongo (BS; c = 0), the Freundlich (c = 0 and 𝐶𝐶𝑒𝑒≪𝑏𝑏), the Jovanovich (c = 0 and a = 1), the Hill˗Sips (HS; c = 1), the Langmuir (c = 1 and a = 1) and the Brouers˗Gaspard (BG; c = 0.5) isotherms were used to analyze adsorption data and describe the interaction between the BBR adsorbent and the OTC and Cd(II) adsorbates. The experimental conditions used in the kinetic and isotherm studies are the optimal adsorption conditions obtained by the CCD˗RSM design. The consistency between the experimental and the predicted values was assessed using the coefficients of correlation (R2). Moreover, the kinetic and isotherm equations appropriateness was verified by the reduced Chi-square (ꭕ2) and the residual sum of square (RSS). Results and discussion Biosorbent Characterization The determination of pHpzc was performed to examine the relationship between the adsorbent surface charge and the solution pH. NaOH and HCl solutions were used to adjust the pH of 0.01M NaCl solutions, in the range 1-12. Then 50 mL of NaCl at each pH and 0.15 g of BBR biosorbent were mixed for 24 h and stirred at 350 rpm. Then, the pHfinal of each solution was measured using an Orion Star A211 pHmeter (Thermo Fisher Scientific, USA) [39]. The pHpzc value corresponds to the intersection of the bisector line and the graphical representation of pHf versus pHi. The pHpzc value of BBR adsorbent was found to be 4.7, as indicated in Fig. 1. At pH < pHpzc, the biosorbent surface is positively charged, and the biosorption of anions is favored. Whereas, at pH > pHpzc, the biosorbent surface is negatively charged, and biosorption of cations is favored [40,41]. Fig. 1. 9 Fig. 2. exhibits the FTIR˗ATR spectra for the BBR particles, before and after adsorption of OTC and Cd(II). The spectra of BBR revealed the presence of the characteristic functional groups of lingo-cellulosic material. In line with Karoui et al. [30], the O-H and C-H stretching medium band between 3650 and 3000 cm−1 can be related to the presence of cellulose and hydrogen bonds association. Peaks at 2918 and 2850 cm-1, attributed to C–H bonds, were observed on the BBR and BBR-OTC-Cd(II) spectra. As was reported by Alshaheri et al. and Guedidi et al. [42,43], the peak at 1713 cm-1 on the BBR and BBR-OTC-Cd(II) spectra is related to the CO2 and C=O elongation vibration obtained from the transformation of phenols, ketones, esters, ether and/or aromatic carboxylic groups. As was previously explained by Guedidiet et al., Song et al., and Karoui et al. [30,43,44], the vibrations at 1595 and 1600 cm−1 could be ascribed to C=C of aromatic vibrations, asymmetric COO–, and/or the C=O conjugated with aromatic bonds. These signals would indicate the presence of lignin [45]. In line with Taylor et al. [46], the peak observed at 1234 cm−1 could be related to the C-H, C–O, and C-O-H bonds in carboxylic acids. The peak, present in BBR spectra, near 1039 cm−1 originating from carbonyl peak can be ascribed to the cellulose structure. The BBR spectra before and after bisorption showed a peak at 897 cm-1, assigned to C-C band. As observed, OTC and Cd(II) adsorption showed H-C-H peak appearance at 1317 cm-1 [47].This finding indicates the OTC antibiotic adsorption on the BBR surface. After binary biosorption onto the BBR surface, the band at 1039 cm-1, attributed to the C=O bond, moved to 1034 cm-1. This shift confirmed that direct complexation between OTC and the BBR functional groups could be occurred. In the same way, the band at 1595 cm-1 related to the C=C of aromatic vibrations shifted to 1588 cm-1. These shifts should be due to the biosorption phenomena. Furthermore, the presence of Cu(II), simultaneously to OTC, lead to slight shifts of wave-numbers from 1600 to 1609 cm-1 and from 1595 to 1588 cm-1 in the C=C , COO–, and C=O peaks [48], and enhancement in the intensity of carboxylic acids bonds at 1234 cm-1 and C-C deformation at 1370 cm-1. Finally, after OTC and Cd(II) adsorption, a decrease in the intensity of the CO2 and C=O bands at 1713 cm-1 and a phenyl vibration band at 740 cm-1 were observed [49]. These changes suggested that the carbonyl and phenolic groups of BBR reacted with Cd(II) to form inner-sphere surface complexes, thus generating BBR-O≡Cd-OTC ternary complexes [44]. In summary, the appearance of new characteristic groups for OTC and Cd(II), the increase/decrease of the intensity of some bands and the shifts of other peaks would confirm the adsorption of OTC and Cd(II) pollutants onto the BBR particles. 16 and the adsorbed molecular species interact among them to form a contaminants poly-layer. Secondly, these results confirm the high heterogeneous systems. Finally, the adsorption process of OTC and Cd(II) imply sorbing materials with various structural and chemical characteristics. Fig. 8. Table. S4 (Supplementary Material) Biosorption thermodynamics The spontaneity, endothermic or exothermic nature of the adsorption phenomenon of binary OTC and Cd(II) onto BBR were evaluated using the enthalpy (ΔH0), entropy (ΔS0), and Gibbs free energy (ΔG0). The Equations (27)-(30) and the plotting ln(Kd) versus 1/T were used to determine the mentioned basic thermodynamic parameters [63–66]: ∆𝐺𝐺0=∆H0−T∆S0 (27) Kd=qe Ce (29) ln Kd=∆S0 R−∆H0 RT (30) where, T represents the absolute temperature (K); R is the universal gas constant (8.314 J.mol/K); and Kd is the distribution coefficient. The linear plot of lnKd against 1/T is given in Fig. S4. Fig. S4. (Supplementary Material) All thermodynamic data are illustrated in Table 6. The negative values of ∆𝐺𝐺0 obtained at different temperatures demonstrated the thermodynamic spontaneity of the adsorption reaction [30]. It seems that the increase of temperature resulted in a decrease in the ∆𝐺𝐺0 values and an increase in the Kd values. This result implied that OTC and Cd(II) adsorption is more favorable at high temperature, through the improvement of the adsorption capacity of BBR surface. The ∆𝐻𝐻0 values were 22.76 kJ mol-1 and 35.72 kJ mol-1, for OTC and Cd(II) adsorption, respectively. The mentioned positive values of ∆𝐻𝐻0 indicated that OTC and Cd(II) 17 adsorption was endothermic. Furthermore, the ∆𝐻𝐻0 values were lesser than 80 kJ/mol, indicating the physical nature of OTC and Cd(II) adsorption onto BBR. The ∆𝑆𝑆0 values were 70.28 J/mol.k and 204.94 J/mol.k, for OTC and Cd(II) biosorption, respectively. These positive values suggest the increased randomness at the solid/solution interface during the adsorption of OTC and Cd(II) by BBR. Similar results were found in many thermodynamic studies about adsorption of antibiotic and heavy metal onto various biomaterials [67,68]. Finally, by comparing thermodynamic parameters of OTC and Cd(II), it seemed that the adsorption of Cd(II) is faster than the OTC adsorption, due to the affinity and the stronger interaction between the BBR surface and Cd(II) cations. This result confirms the isotherm studies. Table 6. Suggested biosorption mechanism BBR is a reed-based composite. The reed is mainly composed of celluloses, hemicelluloses and lignin. The FTIR, MEB analysis and modeling study confirmed the chemical and physical heterogeneity of the BBR surface. BBR biocomposite was then rich in carbonyls, hydroxyls, ethers, aldehydes, phenols, aldehydes, and aromatic compounds. The optimum pH for OTC and Cd(II) biosorption was found to be 6, using CCD analysis. At this pH, the BBR surface (pHpzc = 4.7) was negatively charged. Consequently, the adsorption of Cd(II) cations takes place through electrostatic attraction between them and the BBR surface. Knowing that OTC pKa1, pKa2 and pKa3 values were 3.3, 7.7 and 9.9 respectively, at pH 6, OTC would have a zwitterionic form (OTC+-) and electrostatic interaction between the OTC cationic groups (RN+) and the BBR surface would occur. The biosorption of OTC could also take place via π-π stacking, as well as hydrophobic interactions, as shown in Fig. 9. According to the CCD application and FTIR analysis, the formation of BBR-O≡Cd-OTC ternary complex could take place. This complex would be caused by an electrostatic interaction between the cationic Cd(II) and the anionic group (COO-) of OTC. Moreover, the explanation of OTC and Cd(II) adsorption through the BBR pores would be equally possible. Fig. 9. Regeneration study 18 The regeneration study of binary OTC and Cd(II) biosorption was performed at the optimal conditions, obtained by CCD analysis. Desorption experiments were used to regenerate the biosorbent and recover the antibiotic and heavy metal. In this study, ultra-pure water, 0.1 M HCl, and 0.1 M CH3COOH were examined as eluents. The obtained results are illustrated in Fig. 10 a1 and a2. The maximum percentage recovery of OTC was 92.44% with 0.1M CH3COOH, whereas that of Cd(II) was 99.96% with ultra-pure water in the first cycle. It was found that 0.1M CH3COOH and ultra-pure water were good eluents for desorption of OTC and Cd(II), respectively, compared to other assessed eluents. The strong desorption efficiency of CH3COOH may simply be explained by the electrostatic repulsion occurring between the BBR adsorbent and OTC antibiotic, due to the fact that both acquired positive charges in an acidic medium. Then, 1M CH3COOH and ultra-pure water were used as an eluent agent for OTC and Cd(II), respectively. The regenerated biosorbent was reused for three adsorption-desorption cycles, as shown in Fig. 10. b1 and b2. A gradual decrease in OTC and Cd(II) adsorption efficiency with the increase in the cycle’s number was observed. After a sequence of three cycles, the adsorption capacities of the BBR adsorbent were reduced from 63.65 % to 56.78 % for OTC, and from 99.99 % to 98.68 % for Cd(II). The loss in the biosorption capacity of the biosorbent for antibiotic and heavy metal ions were 6.87 % for OTC and 1.31 % for Cd(II). This might be due to the alteration of the BBR adsorbent and/or the ignorable weight loss of BBR biosorbent during the adsorption-desorption process. Therefore, BBR could be used for three cycles in OTC and Cd(II) biosorption with a small loss in its biosorption capacity. Conclusion In this study, a reed-based-beads biocomposite was developed from an invasive reed plant, to be used as adsorbent, which would be easy to produce, efficient and cost effective. A detailed characterization of the BBR beads showed that this biocomposite was made of celluloses, hemicelluloses and lignin, which indicate its potential as enhanced adsorbent. The BBR beads were also characterized after OTC and Cd(II) adsorption, finding that the biosorption process was successfully implemented. Moreover, the optimization of the biosorption was achieved through the combination of five factors; namely, the adsorbent amount, the OTC concentration, the Cd(II) concentration, the pH, and the reaction time. The optimal conditions were found to be 4.19 g L-1 of biosorbent, 165.54 µmol L-1 of OTC, 362.16 µmol L-1 of Cd(II), a pH of 6, and 25.14 h contact time. This study proved that at the mentioned optimal 19 conditions, the percentages of adsorption would reach 63.66 % and 99.99 % for OTC and Cd(II), respectively. Furthermore, at equilibrium, biosorption efficiency values were 16.05 mg g-1 and 23.54 mg g-1 for OTC and Cd(II), respectively. Therefore, the current study showed that the developed biosorbent is economic and effective for the simultaneous uptake of OTC and Cd(II) from aqueous solution. Regarding the elucidation of mechanisms involved in the simultaneous biosorption of OTC and Cd(II) onto BBR, the kinetic study showed that the adsorption of Cd(II) was faster than that of OTC. Indeed, the states of equilibrium were reached at 20 h and at 5 h for OTC and Cd(II), respectively. OTC and Cd(II) adsorption mechanisms were described by the BSf equation and the Weibull model, respectively. The isotherm study confirmed that the biosorbent surface was heterogeneous and the Freundlich/Jovanovich and BS equations led to the best non-linear fits for the biosorption of OTC and Cd(II), respectively. Furthermore, the BBR efficiency was assessed thermodynamically, showing that higher temperature favored better biosorption percentages, reaching up to 80.58 % and 100 % for OTC and Cd(II), respectively, at 50 °C. Furthermore, it was shown that the biosorption mechanisms for OTC and Cd(II) are physisorption, spontaneous and endothermic. The adsorption mechanisms took place through the electrostatic interactions between the Cd(II) cations and the anionic groups (COO-, OH) of the BBR adsorbent, and/or via π-π stacking interactions between lignin aromatic groups of BBR and OTC pollutant. In addition, electrostatic interaction between the OTC cationic groups (R-N+) and the BBR surface, would occur. The formation of BBR-O≡Cd-OTC ternary complex could take place. This complex would be caused by electrostatic interactions between the cationic Cd(II) and the anionic (C=O) groups of OTC. Moreover, the explanation of OTC and Cd(II) adsorption through the BBR pores and through formation of BBR-O≡CdOTC ternary complex would be also possible. Overall, the main objective of this work being to promote the valorizing of the reed plant and ridding the aquatic environment of large amounts of polluted wastewater at little expenses, was successfully reached, which would be of aid to facilitate an improvement of environmental health. 20 References [1] A. Hartmann, E.M. Golet, S. Gartiser, A.C. Alder, T. Koller, R.M. Widmer, Primary DNA damage but not mutagenicity correlates with ciprofloxacin concentrations in German hospital wastewaters, 119 (1999) 115–119. [2] J.R. V Pils, D.A. Laird, Sorption of tetracycline and chlortetracycline on Kand Casaturated soil clays, humic substances, and clay-humic complexes, Environmental Science and Technology. 41 (2007) 1928–1933. doi:10.1021/es062316y. [3] F. Lian, Z. Song, Z. Liu, L. Zhu, B. Xing, Mechanistic understanding of tetracycline sorption on waste tire powder and its chars as affected by Cu 2 þ and pH, Environmental Pollution. 178 (2013) 264–270. doi:10.1016/j.envpol.2013.03.014. [4] X. Xu, X. Li, Chemosphere Sorption and desorption of antibiotic tetracycline on marine sediments, Chemosphere. 78 (2010) 430–436. doi:10.1016/j.chemosphere.2009.10.045. [5] N. Suga, M. Ogo, S. Suzuki, Risk assessment of oxytetracycline in water phase to major sediment bacterial community : A water-sediment microcosm study, Environmental Toxicology and Pharmacology. 36 (2013) 142–148. doi:10.1016/j.etap.2013.03.013. [6] H. Mihciokur, M. Oguz, Removal of oxytetracycline and determining its biosorption properties on aerobic granular sludge, Environmental Toxicology and Pharmacology. 46 (2016) 174–182. doi:10.1016/j.etap.2016.07.017. [7] F. Fu, Q. Wang, Removal of heavy metal ions from wastewaters : A review, Journal of Environmental Management. 92 (2011) 407–418. doi:10.1016/j.jenvman.2010.11.011. [8] A. Aghababaei, M.C. Ncibi, M. Sillanpää, Optimized removal of Oxytetracycline and Cadmium from contaminated waters using chemically-activated and pyrolyzed biochars from forest and wood-processing residues, Bioresource Technology. (2017). doi:10.1016/j.biortech.2017.04.119. [9] S. Li, X. Li, D. Wang, Membrane ( RO-UF ) filtration for antibiotic wastewater treatment and recovery of antibiotics, 34 (2004) 109–114. doi:10.1016/S13835866(03)00184-9. 21 [10] M. Soylak, U. Divrikli, S. Saracoglu, Membrane filtration – atomic absorption spectrometry combination for copper , cobalt , cadmium , lead and chromium in environmental samples, (2007) 169–176. doi:10.1007/s10661-006-9271-0. [11] W. Li, Q. Wang, Removal of Tetracycline and Oxytetracycline in Water by a Reverse Osmosis Membrane, 2009 3rd International Conference on Bioinformatics and Biomedical Engineering. (2009) 1–4. doi:10.1109/ICBBE.2009.5163315. [12] P. Taylor, S. Bayar, A.E. Yilmaz, R. Boncukcuo, B.A. Fìl, M. Muhtar, B.A. Fil, S. Bayar, A.E. Yilmaz, R. Boncukcuog, Desalination and Water Treatment Effects of operational parameters on cadmium removal from aqueous solutions by electrochemical coagulation aqueous solutions by electrochemical coagulation, (n.d.) 37–41. doi:10.1080/19443994.2012.749201. [13] J. Issn, F.A. Online, Coagulation / flocculation process in the removal of trace metals present in industrial wastewater, (2006) 1–4. [14] Q. Chen, S. Wu, Y. Xin, Synthesis of Au – CuS – TiO 2 nanobelts photocatalyst for efficient photocatalytic degradation of antibiotic oxytetracycline, 302 (2016) 377–387. doi:10.1016/j.cej.2016.05.076. [15] T. Watanabe, T. Taklzawa, K. Honda, Photocatalysis through Excitation ot Adsorbates . 1 . Highly Efficient / V-Deethylation of Rhodamine B Adsorbed to CdS, 81 (1977) 1845–1851. doi:10.1021/j100534a012. [16] I.L.A.K.B. Alciog, Simultaneous Removal of Oxytetracycline and Sulfamethazine Antibacterials from Animal Waste by Chemical Oxidation Processes, (2009) 11284– 11291. doi:10.1021/jf902188j. [17] A. Alwash, Study on the Catalytic Activity of Cerium / Cadmuim Mixed Oxide Catalysts for the Photo catalytic Degradation of Orange G Dye Keywords :, 20 (2017) 71–79. [18] E. Maliou, M. Malamis, P.O. Sakellarides, LEAD AND CADMIUM REMOVAL BY ION EXCHANGE, 25 (2018) 133–138. [19] W. Sun, Y. Sun, K.J. Shah, H. Zheng, B. Ma, Electrochemical degradation of oxytetracycline by Ti-Sn-Sb / γ-Al 2 O 3 threedimensional electrodes, Journal of 22 Environmental Management. 241 (2019) 22–31. doi:10.1016/j.jenvman.2019.03.128. [20] S. Farhadi, F. Siadatnasab, Sonocatalytic degradation of organic pollutants by CdS nanoparticles hydrothermally prepared from cadmium ( II ) diethanoldithiocarbamate, 66 (2017) 5004. doi:10.5004/dwt.2016.0096. [21] J. Qiao, H. Zhang, G. Li, S. Li, Z. Qu, M. Zhang, J. Wang, Fabrication of a novel Zscheme SrTiO 3 / Ag 2 S / CoWO 4 composite and its application in sonocatalytic degradation of tetracyclines, 211 (2019) 843–856. doi:10.1016/j.seppur.2018.10.058. [22] Adsorption of cadmium(II) from aqueous solution onto activated carbon, Water Science and Technology. 35 (1997) 205–211. doi:10.1016/S0273-1223(97)00132-7. [23] I. Sciences, S. Reclamation, E. Science, Adsorption and Desorption of Oxytetracycline and Carbamazepine by Multiwalled Carbon Nanotubes, (2009) 9167–9173. doi:10.1021/es901928q. [24] M. Gaouar-Yadi, K. Tizaoui, N. Gaouar-Benyelles, B. Benguella, Efficient and ecofriendly adsorption using low-cost natural sorbents in waste water treatment, Indian Journal of Chemical Technology. 23 (2016) 204–209. [25] R. Ben Arfi, S. Karoui, K. Mougin, A. Ghorbal, Adsorptive removal of cationic and anionic dyes from aqueous solution by utilizing almond shell as bioadsorbent, EuroMediterranean Journal for Environmental Integration. 2 (2017) 20. doi:10.1007/s41207-017-0032-y. [26] Q. Song, Y. Fang, C. Cao, J. Liang, Z. Liu, L. Li, Y. Huang, J. Lin, C. Tang, Selective Adsorption Behavior/Mechanism of Antibiotic Contaminants on Novel Boron Nitride Bundles, Journal of Hazardous Materials. (2018). doi:10.1016/j.jhazmat.2018.10.054. [27] Y. Zhang, Y. Wang, H. Zhang, Y. Li, Z. Zhang, W. Zhang, Resources , Conservation & Recycling Recycling spent lithium-ion battery as adsorbents to remove aqueous heavy metals : Adsorption kinetics , isotherms , and regeneration assessment, Resources, Conservation & Recycling. 156 (2020) 104688. doi:10.1016/j.resconrec.2020.104688. [28] S. Altenor, B. Carene, E. Emmanuel, J. Lambert, J.J. Ehrhardt, S. Gaspard, Adsorption studies of methylene blue and phenol onto vetiver roots activated carbon prepared by 23 chemical activation, Journal of Hazardous Materials. 165 (2009) 1029–1039. doi:10.1016/j.jhazmat.2008.10.133. [29] Y. Wan, Y. Bao, Q. Zhou, Chemosphere Simultaneous adsorption and desorption of cadmium and tetracycline on cinnamon soil, Chemosphere. 80 (2010) 807–812. doi:10.1016/j.chemosphere.2010.04.066. [30] S. Karoui, R. Ben Arfi, K. Mougin, A. Ghorbal, A.A. Assadi, A. Amrane, Synthesis of novel biocomposite powder for simultaneous removal of hazardous ciprofloxacin and methylene blue : Central composite design , kinetic and isotherm studies using BrouerSotolongo family models, Journal of Hazardous Materials. (2019) 121675. [31] R. Ben Arfi, S. Karoui, K. Mougin, A. Ghorbal, Cetyltrimethylammonium bromide‑treated Phragmites australis powder as novel polymeric adsorbent for hazardous Eriochrome Black T removal from aqueous solutions, Polymer Bulletin. 76 (2018) 5077–5102. [32] Z. Parsaee, N. Karachi, S.M. Abrishamifar, M.R.R. Kahkha, R. Razavi, Silver-choline chloride modified graphene oxide: Novel nano-bioelectrochemical sensor for celecoxib detection and CCD-RSM model, Ultrasonics Sonochemistry. 45 (2018) 106–115. doi:10.1016/j.ultsonch.2018.03.009. [33] A.M. Ben Hamissa, F. Brouers, M.C. Ncibi, M. Seffen, Kinetic Modeling Study on Methylene Blue Sorption onto Agave americana fibers: Fractal Kinetics and Regeneration Studies, Separation Science and Technology (Philadelphia). 48 (2013) 2834–2842. doi:10.1080/01496395.2013.809104. [34] S. Gaspard, S. Altenor, N. Passe-Coutrin, A. Ouensanga, F. Brouers, Parameters from a new kinetic equation to evaluate activated carbons efficiency for water treatment, Water Research. 40 (2006) 3467–3477. doi:10.1016/j.watres.2006.07.018. [35] M. Saad, H. Tahir, J. Khan, U. Hameed, A. Saud, Synthesis of polyaniline nanoparticles and their application for the removal of Crystal Violet dye by ultrasonicated adsorption process based on Response Surface Methodology, Ultrasonics Sonochemistry. 34 (2017) 600–608. [36] B. Sadhukhan, N.K. Mondal, S. Chattoraj, Optimisation using central composite design (CCD) and the desirability function for sorption of methylene blue from aqueous 24 solution onto Lemna major, Karbala International Journal of Modern Science. 2 (2016) 145–155. doi:10.1016/j.kijoms.2016.03.005. [37] F. Brouers, O. Sotolongo-Costa, Generalized fractal kinetics in complex systems (application to biophysics and biotechnology), Physica A: Statistical Mechanics and Its Applications. 368 (2006) 165–175. doi:10.1016/j.physa.2005.12.062. [38] F. Brouers, T.J. Al-Musawi, Brouers-Sotolongo fractal kinetics versus fractional derivative kinetics: A new strategy to analyze the pollutants sorption kinetics in porous materials, Journal of Hazardous Materials. 350 (2018) 162–168. doi:10.1016/j.jhazmat.2018.02.015. [39] S. Larous, A.H. Meniai, Adsorption of Diclofenac from aqueous solution using activated carbon prepared from olive stones, International Journal of Hydrogen Energy. 41 (2016) 10380–10390. doi:10.1016/j.ijhydene.2016.01.096. [40] L.R. Bonetto, F. Ferrarini, C. De Marco, J.S. Crespo, R. Guégan, M. Giovanela, Removal of methyl violet 2B dye from aqueous solution using a magnetic composite as an adsorbent, Journal of Water Process Engineering. 6 (2015) 11–20. doi:10.1016/j.jwpe.2015.02.006. [41] Z.L. Yaneva, N. V Georgieva, Insights into Congo Red Adsorption on Agro-Industrial Materials - Spectral , Equilibrium , Kinetic , Thermodynamic , Dynamic and Desorption Studies . A Review, 4 (2012) 127–146. [42] T.A.S. Ali Ahmed Alshaheri, Mohamed Ibrahim Mohamed Tahir, Mohd Basyaruddin Abdul Rahman, Thahira Begum, Synthesis, characterisation and catalytic activity of dithiocarbazate Schiff base complexes in oxidation of cyclohexane, Journal of Molecular Liquids. 240 (2017) 486--496. doi:10.1016/j.molliq.2017.05.081. [43] H. Guedidi, B. Slama, Préparation et modification de carbones activés pour l ’ adsorption de polluants organiques émergents : molécules pharmaceutiques et liquides ioniques, 2016. [44] X. Song, D. Liu, G. Zhang, M. Frigon, X. Meng, K. Li, Bioresource Technology Adsorption mechanisms and the effect of oxytetracycline on activated sludge, Bioresource Technology. 151 (2014) 428–431. doi:10.1016/j.biortech.2013.10.055. 25 [45] W.H. Danial, Z.A. Majid, M. Nazlan, M. Muhid, S. Triwahyono, M.B. Bakar, Z. Ramli, The reuse of wastepaper for the extraction of cellulose nanocrystals, Carbohydrate Polymers. 118 (2014) 165–169. [46] P. Taylor, D.R. Manenti, D.E.G. Trigueros, A.P. De Oliveira, C.E. Borba, A.D. Kroumov, pt us, (2015) 37–41. doi:10.1080/09593330.2015.1051591. [47] T. Oxidation, L. Jiao, H. Bian, Y. Gao, X. Lin, W. Zhu, H. Dai, Highly Dispersible Cellulose Nanofibrils Produced via Mechanical Pretreatment Highly Dispersible Cellulose Nanofibrils Produced via Mechanical Pretreatment and TEMPO-mediated Oxidation, (2018). doi:10.1007/s12221-018-8565-5. [48] A. Alonso-simón, P. García-angulo, H. Mélida, A. Encina, J.M. Álvarez, J.L. Acebes, The use of FTIR spectroscopy to monitor modifications in plant cell wall architecture caused by cellulose biosynthesis inhibitors, (2011). doi:10.4161/psb.6.8.15793. [49] C.S. Transactions, M.K. Mishra, Fourier Transform Infrared Spectrophotometry Studies of Chromium Trioxide-Phthalic Acid Complexes, 5 (2016) 770–774. doi:10.7598/cst2016.1260. [50] R. Ben Arfi, S. Karoui, K. Mougin, A. Ghorbal, Adsorptive removal of cationic and anionic dyes from aqueous solution by utilizing almond shell as bioadsorbent, EuroMediterranean Journal For Environmental Integration. 2 (2017) 20. [51] S.F. Montanher, E.A. Oliveira, M.C. Rollemberg, Removal of metal ions from aqueous solutions by sorption onto rice bran, 117 (2005) 207–211. doi:10.1016/j.jhazmat.2004.09.015. [52] S. Mohan, G. Sreelakshmi, Fixed bed column study for heavy metal removal using phosphate treated rice husk, 153 (2008) 75–82. doi:10.1016/j.jhazmat.2007.08.021. [53] M. Harja, G. Ciobanu, Studies on adsorption of oxytetracycline from aqueous solutions onto hydroxyapatite, Science of the Total Environment. 628–629 (2018) 36–43. doi:10.1016/j.scitotenv.2018.02.027. [54] L. Aristilde, B. Lanson, L. Charlet, Interstratification patterns from the pH-dependent intercalation of a tetracycline antibiotic within montmorillonite layers, Langmuir. 29 (2013) 4492–4501. doi:10.1021/la400598x. 32 Fig. 6. Desirability function profile for simultaneous biosorption of of OTC and Cd(II). Dashed red line indicated optimized values. Fig. 7. kinetic curves for the OTC and Cd(II) adsorption from binary system (F1 = 4.1892 g/L ; F2 = 165.54 µM ; F3 = 362.16 µM ; X4 = 6 ; X5 = 25.135 h ; T = 20 ± 1 °C). 33 Fig. 8. isotherm curves for the OTC and Cd(II) adsorption from binary system (F1 = 4.1892 g/L ; F2 = 165.54 µM ; F3 = 362.16 µM ; X4 = 6 ; X5 = 25.135 h ; T = 20 ± 1 °C). 34 Fig. 9. Suggested biosorption mechanism, Red and orange lines: electrostatic attractions, Green lines: BBR-Cd(II)-OTC complex interactions, blue lines: p-p stacking interactions 35 Fig. 10. Desorption studies of OTC and Cd(II) as function of (a) desorption time and (b) number of adsorption-desorption cycles 36 Table 1. The equations of used error functions. Error Function Definition Eq N° Description References Regression Coefficient R2 = qe exp−qe cal � � � � � � ∑[(qe exp−qe cal � � � � � � )2+ (qe exp−qe cal)2] 3 q e exp and q e cal are the experimental and calculated sorption capacity (mg g-1), respectively. n is the number of data points in the experiment. yP,i is the predicted value, y0,i is the experimental value. [1] Chi-square test χ2 = (qe exp – qe cal)2/qe cal 4 [2] Residual Sum of Squares RSS= ∑(y P,i −y 0,i )2 N i=1 RSS= ∑(qe cal−qe exp)2 N i=1 5 [3] 37 Table 2. Experimental factors and levels in the CCD design. Factor code Factors level Star point (α = 2.3784) Low (-1) Central (0) High (+1) +α -α F 1 Adsorbent dosage (g L-1) 2 3 4 5.38 0.62 F 2 OTC concentration (µmol L-1) 175 225 275 343.92 106.08 F 3 Cd(II) concentration (µmol L-1) 400 600 800 1075.68 124.32 F 4 pH 4 6 8 10.75 1.24 F 5 Contact time (hour) 12 18 24 32.27 3.73 38 Table 3. The names and non-linear-forms of used kinetic models. Kinetic model Non linear form Eq N° Descriptions References BSf(n,α) qn,α(t)= qe[1 −(1 + (n−1)( t τn,α)α) −1 n−1)] 8 q n,α (t) and q e are the adsorbed contaminant amounts (mg g-1 ) at time t (min) and at equilibrium, respectively; n is the fractional reaction order; α is the fractal time index depending on the geometrical and energetic adsorbent heterogeneity; τn,α is the characteristic time of the adsorption kinetics. [4–7] qn,α(t)= qe[1 −Expn(−( t τn,α)α)] 9 PFO=BSf(1,1) q(t)= qe,1[1 −Exp(− t τ1,1)] τ1,1= k1−1 10 k1 is the adsorption rate constant of the pseudo˗first˗order model (min-1). [7] PSO=BSf(2,1) q(t)= qe,2( t τ2,1 1+ t τ2,1) τ2,1= (qe,2k2)−1 11 k2 is the adsorption rate constant of the pseudosecond-order model (g mg-1 min-1). [7] Weibull= BSf(1,α) q(t)= qe,W[1 −Exp(−( t τ1,α )α)] τ1,α= k1,α −1 12 k1,α is the adsorption rate constant of the Weibull model (min-1). [7] Hill= BSf(2,α) q(t)= qe,H( ( t τ2,α) α 1+(t τ2,α)α) τ2,α= (qe,Hk2,α)−1 13 k2,α is the adsorption rate constant of the Hill model (g mg-1 min-1). [7] Brouers˗Gaspard= BSf(1.5,α) q(t)= qe,BG[1 −(1 + 0.5( t τ1.5,α )α)−2] 14 [7] 39 Table 4. The names and non-linear-forms of used isotherm models. Isotherm model Non linear form Eq N° Descriptions References GBS qe,GBS= qe,max �1−Expc�−� C e b� a �� 15 qe,GBS and qe,max are the adsorbed amount at equilibrium and the saturation adsorbed amount at equilibrium, respectively (mg g-1). Ce is the adsorbate concentration in the liquid phase (mg L-1). The coefficients (a) and (c) are related to the form, and (b) represents a scale parameter. [7–9] qe,GBS= qe,max �1−�1+c �Ce b�a� −1 c� 16 BS (c=0) qe,BS= qe,maxBS �1−Exp �−� C e b� a �� 17 qe,maxBS is the maximum adsorption capacity for the BS model. [7] Freundlich (c=0 and Ce≪b) q e,F = k F C e a 18 k F is the Freundlich constant, indicating the adsorption capacity. [7] Jovanovich (c=0 and a=1) qe,J= qe,maxJ �1−Exp �− C e b�� 19 qe,maxJ is the maximum adsorption capacity for the Jovanovich model. [7] HS (c=1) qe,HS= qe,maxHS �1−�1+ �Ce b� a � −1 � 20 qe,maxHS is the maximum adsorption capacity for the HS model. [7] Langmuir (c=1 and a=1) qe,L= qe,maxL � C eb � 1 + Ceb �� 21 qe,maxL is the maximum adsorption capacity for the Langmuir model. [7] BG (c=0.5) qe,BG= qe,maxBG�1−Expc=0.5�−� C e b� a �� qe,BG= qe,maxBG �1−�1+ 1 2�Ce b�a�−2� 22 qe,maxBG is the maximum adsorption capacity for the BG model. [7] 40 Table 5. Model summary OTC Cd(II) R-squared (R2) 0.96056 0.97598 Adjusted R-squared 0.91753 0.94978 MS Residual 18.36288 26.07838 41 Table 6. Thermodynamic parameters for the biosorption of OTC and Cd(II) onto BBR from binary systems at various temperatures (F1 = 4.1892 g/L ; F2 = 165.54 µM ; F3 = 362.16 µM ; F4 = 6 ; F5 = 25.135 h). Contaminant T (°K) K d ∆H0 (kJ mol -1 ) ∆S0 (J/mol.K) ∆G0 (kJ mol -1 ) OTC 293.15 0.42 22.76 70.28 -43.37 303.15 0.55 -44.07 313.15 0.75 -44.77 323.15 0.99 -45.48 Cd(II) 293.15 24289.74 35.72 204.94 -95.80 303.15 32386.40 -97.85 313.15 48579.70 -99.90 323.15 97159.66 -101.94