New biosorbents based on the seeds, leaves and husks powder of Moringa oleifera for the effective removal of various toxic pollutants

ABSTRACT A new hydrogel beads-sponge structure (these new beads combine the properties of the beads with the sponge character) with a super rate of swelling hydrogel in water and wastewater and a large porosity was created during this research. New biosorbents were obtained by modification of three parts of Moringa oleifera (seeds, leaves and husks). Prepared adsorbent beads were characterised by some analytical methods, viz. FT-IR, SEM and SEM-EDX. The effects of operational parameters, such as pH, dose of sorbents, temperature, and others were investigated and compared with those of initial powder. EDX results of M. oleifera (MO) and M. oleifera-gelatin (MO-GEL) had the highest percentage of amino groups. The FT-IR spectra showed the existence of – NH, – OH, – COOH functional groups in the biomass and the modified beads. The maximal adsorption potential was observed at the range of pH value of (5–6) for all pollutants tested and the pHpzc showed the charge carried by MO and MO-GEL in different pH conditions. Conclusively, the objective of this study was the valorisation of the powder in the form of beads to facilitate its use in sorption processes, and without disturbing or diminishing efficiency. This study showed that the sorbents based on Moringa oleifera in their bead forms were good alternative adsorbent that could be used for recovery of dyes (the sorption capacity reaches up to 175.28 mg CV g−1 (≈44%) and 493.76 mg BG g−1 (≈65.8%) for MOS-GEL beads and reaches up to 475.27 mg CV g−1 (≈63%) and 575.82 mg BG g−1 (≈83%) for MOL-GEL) and heavy metal from water (47.08 mg Pb g−1 ( 57% for MOS-GEL), 25.99 mg Pb g−1 ( 38.41% for MOL-GEL) and 32.25 mg Pb g−1 ( 24,6% for MOH-GEL)).


Introduction
Biosorption is a specific type of adsorption/sorption, where pollutants are removed using different types of biomass that can be used as effective bio-adsorbents compared to conventional adsorbents.Interest in these biopolymers has developed due to their vegetable oil, to enhance them in the form of beads to facilitate their use for water treatment.However, no previous work has been developed about the modification of MO 'seed, leaves and husks' for elimination of the selected pollutants from aqueous solution or even detailed studies on the mechanisms in CV, BG and Pb(II) biosorption by MO.In this context, these new sorbents were characterised by FT-IR, SEM and SEM-EDX to understand their behaviour and to estimate their sorption mechanism for the select pollutants.Third, we tested their efficiency in the treatment of pollutants and sought the optimal operating conditions to improve the efficiency of the sorption process 'pH, temperature, dose, etc.Also, we tried to make a comparison of our material synthesised with other materials known in the literature.

Metal solutions and measurement
The initial standard solutions of 1 g•L −1 of CV, BG dyes and Pb(II) were prepared by dissolving the precise quantity of each of them in distilled water.Working solutions were obtained by diluting these solutions to the desired concentrations.Before the biosorption process, the pH of each test solution was adjusted to the desired value by using 0.5 M HCl or NaOH solutions.The final pH was determined in order to evaluate the potential precipitation phenomena associated with pH variation, by using a pH-metre model: HANNA INSTRUMENTS, pH 209, pH METRE (measurement error of ± 0.01).
Initial 'C 0 , mg L −1 or mmol L −1 ' and equilibrium 'C eq , mg L −1 or mmol L −1 ' metal concentrations were subsequently determined using Atomic absorption spectrophotometer (type: 'PERKIN ELMER Atomic Absorption Spectrometer PinAAcle 900 H') and for the dyes, UV-VISIBLE model spectrophotometer ("ANALYTIK JENA, SPECORD 210/PLUS and a wavelength accuracy of ± 0.1 with the Deuterium line at 656 nm) was used.
For the structure and peak identification available in MO/MO-GEL, we used IRTF-ATR (Attenuated Total Reflection) Spectrophotometer: PERKIN ELMER, Spectrum FT-IR with a resolution of 0.5 cm −1 .And, the morphological structure of biosorbents was visualised by scanning electron microscopy (SEM) and SEM-EDX.This apparatus was used for the surveillance and examination of surface and pore structure of MO and MO-GEL.Additionally, SEM-EDX analysis was used to detect the composition of the structure and the main elements present at the surface of MO and MO-GEL.

Process for manufacturing the various powder based beads
Samples of the seeds and leaves of Moringa oleifera were collected from the second region of Mali in Dioila (Koulikoro, Mali), harvested during the period between 'August and September 2018'.The powder manufacturing procedure is shown in AMS, Section I.The MOS-GEL, MOH-GEL and MOL-GEL beads were prepared using a food gelatin, these new composite material were prepared using a new method developed for the first time in this work.Gelatin-Moringa oleifera beads, i.e. seed, leaves and husks were prepared by a general emulsification method.The gelatin powder was dissolved in distilled water with a percentage of 32% w/v and a volume of 300 mL, heated to 45-50°C and stirred overnight.After homogeneously dissolving the gel in warm water, the solution was mixed using an electric mixer.This quantity is divided into three identical quantity and the quantities of Moringa oleifera powder, i.e. husks, seeds and leaves were added to the gel solution and stirred homogeneously to disperse the powder in the network of gel for 3 h.Reciprocally, for the three types of powders, we added 100 g of gel in 3.5 g of powder.Second, the homogeneous solutions prepared (i.e.MOS-GEL, MOL-GEL and MOH-GEL) were added dropwise to 200 mL of the olive oil (which was left 60 min beforehand in a freezer) using a 3 mL syringe (approximately opening of 2 mm in diameter) with continuous stirring or else a separatory funnel.Then, to complete the gelation, the beakers containing the olive oil and the prepared beads were kept in an ice-water mixture for 2 h, and then left in the refrigerator overnight.Finally, the beads obtained were filtered and washed several times, using distilled water cooled and dried at room temperature overnight.The photos of the wet and dry fabricated beads are represented in Figures 1 and 2. The size of the beads  were characterised by determining the diameters of the dried beads by statistical analysis on samples of some beads (scaled photographs [18]).The diameter and the shape of the different beads are represented in Figure 3.

Batch sorption experiments
To investigate the adsorption characteristics of the powder MOS, MOL, MOH, and beads of the MOS-GEL, MOL-GEL, MOH-GEL, batch sorption tests were performed.Also, to understand the sorption process and the estimated parameters of the adsorption kinetics and isotherm experiments of these adsorbents, all the experiments were performed at shaking rate of 150 rpm and room temperature for three contaminants: Green Brilliant (BG), Crystal violet (CV) and Pb(II).In all series of experiments, a fixed volume of 50 mL metal ion solution was added to a quantity of powder/beads predetermined of 0.5 g in 75 mL flasks.The mixtures were then stirred at 150 rpm for 24 h for the powder to ensure equilibrium, and 2-3 h for the beads.
Kinetic adsorption experiments were performed using 500 mL of Brilliant Green, Crystal violet and Pb(II) solutions with 5 g of the powder MO, and MO-GEL beads.At the desired time intervals (1 min, . . . 1, 2, 3,4, 8, and 24 h).After kinetics study, the stirring time was fixed.
For the isotherm adsorption experiments, a fixed volume of pollutant solution (50 mL of all pollutants) with the initial concentration of 2.5-150 mg/L solutions was stirred with 0.5 g of the adsorbents (Equilibrium data were collected from 2. 5,5,7,10,15,20,25,50,75, 100 and 150 mg/L of initial concentration).The mixtures were shaken for 3-4 h, and filtered through a 0.22 μm inorganic filter.The metal solution of Pb(II) was collected and analysed by a spectrophotometric method and Crystal violet and Green Brilliant dyes were analysed by UV-visible spectrophotometer.
In all cases, the amounts of CV, BG and Pb(II) were calculated based on the variation of the concentration in aqueous solutions before and after sorption (C 0 -C eq ) according to the Equations ( 1) and ( 2).The units used in these equations are: sorbent dosage (SD: m/V, g L −1 ), metal concentration (C 0 and C eq ; mg L −1 or mmol L −1 ); the biosorption capacity q (mg of pollutants/g of biosorbents based on MO and/or mmol of pollutants/g of biosorbent based on MO).
Whereas, the% pollutants removed were determined by the following equation: The biosorption capacity was calculated by the following equation: The non-linear forms of Langmuir [36], Freundlich [37], and Sips [38] models as well as all the necessary information about the modelling by these models are presented in the Supplementary material section, Section II.To identify the bio-sorption mechanism and control the speed of adsorption, two kinetic models are used for the analysis of the experimental results (See Supplementary material section, Section II, for a reminder on equations Eq.AM4 (PSORE) and Eq.AM5(PFORE)).

Characterisation of adsorbents
The physical characteristics such as humidity rate and swelling rate for the precision of the functional groups in the surface of our materials were studied using the experimental protocols used by several researchers.The results are summarised in Table 1, knowing that for the calculation of swelling rate, we used the following formula; While, for the% of water, we used the formula: Where: W W = mass or weight of wet beads (g), W d = mass or weight of the dried beads (g) at 3, 24 or 72 hours.The swelling behaviour of the MO-GEL (see Figure 4) hydrogel was measured by immersion of dry beads in distilled water for 24 h at room temperature and neutral pH.We were surprised that the beads resist 7-8 h and not 24 h and we have a swelling effect and after the beads fill it explodes.In addition, the resistance of these beads changes depending on the environment and this swelling behaviour increases three times if these beads are added to colour solutions but the explosion time changes from 3 to 4 h (depending on the type of powder).These phenomena was followed by camera, the degree of swelling was found to be in the range of 3-10 times compared to the initial volume and mass, and the swelling/deswelling was observed to be significant in both cases of water/ pollutant.So we need to mention the irreversible hydrogel network effect known by these beads.These hydrogels underwent a significant change in volume which was attributed to its porous morphology and the presence of hydrophilic functional groups on the mixture, similar behaviour was observed by S. Lone et al. [35].These two properties allow water/ pollutants to pass through the hydrogel mixture via capillary forces.
According to several researches and the resultants of IR obtained in this study, we can say that MO is rich in functional groups such as: amide C-N, amine N-H, carboxylic acid C-O, hydroxyl O-H and other functional groups [39].So, these functional groups play an important role in sorption phenomenon by developing some electrostatic forces or complexion [40].Generally, the groups -OH, N-H and C-H are due to the presence of proteins, fatty acids and carbohydrates [41].Figure 5(a-c) shows the FTIR spectra of GEL-MOS Beads, GEL-MOH Beads and GEL-MOL Beads, respectively.
The comparisons of results of FTIR spectrum for MO and MO-GEL showed a displacement of some bands which may be due to a chemical reaction between the functional groups in Moringa and Gelatin.The same observations have been reported by several researchers [42][43][44].
For Figure 6(a) we will observe on the seed powder the existence of carboxylic acid and its derivatives and a very wide OH band at a frequency of 3286 cm −1 , some other researchers prove that this wide bands of -OH is mixed with the band of -NH 2 , and confirms that moringa powders contain -OH and -NH 2 .At 3000 cm −1 , we observe an elongation vibration of C-H for the aromatic group 2924 and 2922 cm −1 for the powder and beads, respectively, due to the asymmetric and symmetric stretch of aliphatic chains (-CH) [45,46] of CH 3 , CH 2 and CH.The same observation is also present at 2853 and 2854 cm −1 .After the reaction between GEL and MOS, the disappearance of O-H band at a frequency of 3286 cm −1 was observed.
In MOS-GEL and MOS, the bands at 1745, 1748 cm −1 showed the existence of non-ionic carboxylic groups and its derivative with a strong vibration of elongation for the group C = O, respectively.We observed the bands at 1647, 1636 and 1544 cm −1 which were due to the amine group, NH 2 and the amide group, N-H (2i-amide), respectively.Also, the band at 1643-1646 was associated with ionic carboxylic groups (-COO) carbon dioxide elongation vibration [47].The band at 1636 cm −1 in GEL and 1653 cm −1 in MOS is replaced by 1577 cm −1 in MOS-GEL due to the reaction between -COO of GEL and NH 2 of MOS to obtain an amide -NH group.In the MOS spectrum, a very wide O-H band at a frequency of 3286 cm −1 was disappeared due to the -OH reaction with -COOH for esterification reaction.For the MOS, the spectra showed two strong absorption bands at 1653 cm −1 and a 1549 cm −1 characteristics of amide I and II, respectively, which confirmed the structure of the protein present in Moringa seeds.
In addition, the presence of intensive absorbance band around 1053, 1057 and 1054 cm −1 , respectively, for MOS, MOL and MOH was due to the carbohydrates, aromatics, ethers and polysaccharide [48].We have deformations of alkane CH 3 at 1456 cm −1 , 1232, 1161 and 1057 cm −1 which represented the amines, C-N, with an elongation vibration.The bands at 796, 722 and 645 cm −1 represented alkenes with deformation vibration, alkanes with a deformation vibration and alcohol and phenol with a deformation vibration O-H, respectively.
Figure 5(b) shows that the Leaf's compositions of Moringa oleifera and those of the Seeds were identical.There was a small difference in some bands of the modified leaf powder which may be due to a chemical reaction when the beads form with gelatin.This difference was noted for example: Presence of new peaks in the GEL-MOL structure at 1741 cm −1 , this bands showed the existence of non-ionic carboxylic groups and its derivative with a strong vibration of elongation for the group C = O, and the band at frequency of 3265 cm −1 replaced by a wide band in MOL-GEL at 3302 cm −1 confirmed the existence of carboxylic O-H groups and free COOHs after MOL modification.Also, by the bands obtained with the leaf balls from 1575 and 1539 cm −1 which corresponds to an amine function with a vibration of deformation in NH 2 .At 1486 cm −1 , we observed in the neutral powder a possibility of the existence of chemisorption of H 2 O or of O-H group.In Figure 5(c) (for the husks), we observed an O-H band (with an H bond) wide at 3316 cm −1 of the husk powder for the alcohol and phenol functions of elongation vibration.The variation in transmittance observed in the husks modified by the addition of gelatin may be influencing the pH on biosorption process were Pb(II) species and surface functional groups on the adsorbents [46].The same functional groups in the powder and the beads were observed.Indeed, the rest of the spectrum generally has the same shape as for seeds and leaves and we observed the existence of protein, fatty acid and carbohydrate functional groups in the different parts of Moringa oleifera.The band ranging from 1800 to 1600 cm −1 indicated the presence of stretches of C = O, which were associated with fatty acids and protein structures [46], the presence of carboxylic acids in the husks' powders was confirmed by the rowing peaks 1700 cm −1 .
After biosorption, there was a reduction in intensity and disappearance of some peaks and the shift occurred slightly in some others peaks which manifests the prominent role of these functional groups in the adsorption of cationic dyes 'CV and BG' and Pb(II) on the surface of Moringa oleifera (See Supplementary material section IV, Figure AM5).

SEM and SEM-EDX
The efficiency of the biosorption process depends on the morphology of surface area and functional groups available in the biosorbent.In this study, SEM (Figure 6) and SEM-EDX (Figure AM6, Supplementary section) are used to check the change in the morphological features of MOS, MOL and MOH after cold gelling MOS-GEL, MOL-GEL and MOH-GEL.The surface morphology of MO was different from that of MO-GEL.The scanning electron microscopy (SEM) technique is necessary to observe the physical morphology of the surface of the samples.This analysis shows a heterogeneous structure rich in active function.EDX spectra are presented with element maps to estimate the composition of materials (see Supplementary material section, Section V, Figure AM6).According to the results seen in figure AM6, the composition of MO, MOL-GEL is rich in -N (more than 12%), the proposed modification protected the amine functions existing on the surface of the materials and increased the chance of using MO in the industrial field.In addition, given the richness of structure in Na + , Ca +2 and other ion exchange it is one of the most used mechanisms because this method is usually used when porous sorbents have the function of cation exchange [49].

The pH study
Figures 7 and 8 represent the evolution of the percentage of adsorption as a function of the initial pH of the pollutant solution, respectively, for the dyes, namely, Crystal Violet, Brilliant Green (Figure 7(a,b)) and for the Lead (Figure 8).We carried out adsorption tests by changing the initial pH of a solution in the pH range varied between 2 and 6 by the addition of 0.5 M HCl or the NaOH.After stirring for 3 h at a homogeneous speed of 150 rpm, the equilibrium dye concentrations were determined using UV-visible and Pb(II) using the atomic absorption.
The analysis of these results generally showed that when the pH of the solution increased, the quantity adsorbed by these biosorbants was increased, then tends to stabilise or decrease in some cases from pH = 6.However, the percentage adsorption of Pb(II) varies from zero by 27% (Figure 8) in the pH range studied (2.9 to 6).Whereas, for powders the percentage of sorption varies from 20% to 80% depending on the type of powder MOS, MOL and MOH (MOS; 31.16%,MOL; 76.16%, and MOH; 20.82%).Thus, we chose the value of 6 as an optimal pH to study the effectiveness of our materials towards the adsorption of these pollutants.
For the sorption of different dyes by Moringa, we noted that the efficiency was increased with increasing the pH up to the value of 7 and after that, we noticed the steady state or it was decreased [50,51], it is usually due to precipitation.For example, for CV the best elimination percentage is achieved at pH = 6.0 with MOS (89.12%) and MOS-GEL (79.93%) and for BG it is achieved with MOS (64.66%) and MOS-GEL (73.07%).
In the case of the sorption of Pb(II) by moringa oleifera, Tavares et al. [52] found an optimum pH value of 6. Imran et al. studied the sorption of Pb(II) by Moringa oleifera leaves and found similar trends in terms of pH [53].It is generally known that at low pH value, concentration of H + ions far exceeds that of the metal ions and hence H + ions compete with Pb(II) ions for the surface of the adsorbent which would hinder the Pb(II) ions from reaching the binding sites of the adsorbent resulting in low adsorption amount of Pb(II).As the pH increases, there are fewer protons in the solution and consequently there is lesser competition with Pb(II) for binding to the adsorbent.Above pH 6, the potential of adsorption was reduced which could be due to the formation of Pb(OH) Pb(OH) + , similar investigations were reported in previous studies [54].Gomez-Serrano et al. mentioned that in dilute aqueous solutions of pH <6, lead ions exist as Pb(II) or Pb (OH) + or both, whereas the formation of Pb(II) hydrolysis occurred at pH >6.0 [55,56].
In order to verify the presence of precipitation, we must follow the variation in the initial pH value as a function of the pH after adsorption.The figures AM2-4 (see Supplementary Material, Section III) show the variation of the initial pH of the solutions as a function of the pH at equilibrium during sorption.We must mention that the variation in pH, under the experimental conditions: concentration, dosage, stirring time and temperature, is quite reduced to low values and therefore makes it possible to avoid the phenomenon of precipitation.As shown in Figure AM2-4, the pH PZC value of MOS, MOL, MOH and MOS-GEL, MOL-GEL and MOH-GEL were 5.1, 6.21, 6.09, 5.3, 6.7, 5.7, respectively.Thus, the surface charge of MO is protonated (i.e.positive charge) when the solution pH is lower than their pH PZC , otherwise it is negative.This is due to the fact that the replacement of protons of hydroxyl and carboxylic groups presented on the MO surface; thus, the adsorption of these ions is partially via ion exchange mechanism.This explains the formation of new peaks, disappearance of others and shifts observed in some other bands in the FITR.The same observation was mentioned by Kebede et al., when using water-soluble protein extracted from Moringa Stenopetala seeds, they detected band shifts on the water-soluble protein powder after elimination as well as formation of new peaks and the disappearance of others [57].

Effect of sorbent dosage and metal concentration
In the following work, the mass of the adsorbent was fixed at a dosage equivalent to 1 g/L low or the optimum dose can identify resistance to intraparticle diffusion.Obviously, when the SD is increased, the recovery of pollutants increases, thus shortening the equilibrium time, but does not make it possible to identify the contribution of resistance to intraparticle diffusion in the control of kinetic profiles.In fact, an excess of adsorbent (relative to the content of metal ions/dyes) limits the sorption to external accessible sorption sites [5, 6,58,].
From Figure AM7 (See Supplementary material section), it can be shown that the increases in the quantity adsorbed (mg/g) can be evidenced by the decrease in the mass of the adsorbent(g), which shows that mass of adsorbent is inversely proportional to the adsorption capacity, while the sorption efficiency increases up to a certain equilibrium due to the balanced distribution of the same quantity of pollutant at the active site which are saturated at a time despite the increase in mass.At the start of the adsorption process, the increase in the adsorption percentage is due to the high availability of active sites, i.e. the more the adsorbent dose provides more surface area, which ultimately gives more binding sites for the adsorption of the dyes/metal, the decrease in the number of these active sites increases the adsorption of the dyes/metal.A further increase in the dose of adsorbent had no significant effect on the percentage of elimination of the dye or Pb(II), since the adsorbent was already occupied by all possible binding sites.
When the mass of the solid in the solution becomes large, the number of adsorption sites becomes too, this additional increase in the adsorbent dose does not have impact on the metal/dyes removal.Consequently, the probability of encounter (molecule-site) also increases, leading to better retention.It has been reported by different authors that as the adsorbent dose increases, the movement of dye ions to the energetic adsorption sites will be restricted as well, hence reducing the adsorption efficiency [59].
In the study of the effect of initial concentration of pollutants, in our case, Pb(II), CV and BG, the initial concentrations for pollutant were varied from 2.5 to 150 mg pollutant.L −1 .The effect of the initial concentration of the adsorbate is reversed compared to the effect of sorbent dosage, because if the initial concentration of the adsorbate increases, the quantity of biosorbent pollutant is increased per unit of weight of biosorbent, but decreases its elimination efficiency.However, if the dosage of biosorbents is increased, the amount of biosorbent pollutant is reduced per unit weight of biosorbent, but increases its elimination efficiency.
The fast initial biosorption percent are attributed to the high number of accessible free functional groups, e.g.-OH, -COOH and -NH 2 on the surface of the natural MO and modified MO-GEL at the start of the biosorption process and less steric obstruction for the approaching of pollutant.The slow sorption process is a result of the reduction in the available sorption positions and the build-up of contaminant on the surface of the MO/ MO-GEL adsorbent; this hinders the diffusion of more pollutant into the MO/MO-GEL pores.Slow diffusion rates are related to pores which are of a similar size to the diffusing adsorbates [60].It is for this reason Persson [61] discusses the influence of the hydration of metal ions on their structure and the dimensions of these molecules and their influence on sorption.Therefore, once the molecules of metals or dyes are hydrated, they have different ion radius configurations and other behaviour.

R. Aravindhan et al., and A.A. Swelam et al.
confirmed that the decrease in biosorption capacity is attributable to the splitting effect of the concentration gradient between sorbate and sorbent [62,63], so, in our case, for example, with increased MOS concentration causing a decrease in amount of Pb(II) adsorbed onto unit weight of MOS.This phenomenon was also found by K. Anupam et al. using a powder of activated carbon, for the removal of Cr(VI) from aqueous solution [64] and some other researcher by some biomass.

Thermodynamic studies
The results of thermodynamics and the values obtained for ΔG 0 , ΔS 0 and ΔH 0 at 313 K are shown in Table AM1.The details of the equations used in this part are summarised in the Supplementary material section (See section VI, Eqs.

(AM6) -(AM9)).
This study performs a significant part in the research of adsorption capability [6566].These tests consist, first of all, of changing the temperature between 25°C and 50°C of the solutions of CV, BG and Pb(II) and following the changes in the adsorption efficiency on the powders of MOS, MOL and MOH.Then, we determined the thermodynamic parameters related to their adsorption using the equations and the adapted models.We have noticed that the beads dissolve at a temperature of 50°C.Therefore, the study of hot adsorption has not been possible in practice.
The study of temperature is carried out at a concentration of 100 mg/L (except for Pb(II)), we noticed an increase in the quantity adsorbed with the increase in temperature from 25°C to 50°C, except for the first case MOS-CV: For the CV; the sorption on MOS goes from 254.38 to 116.63 mg/g (14.10% to 15.9%), on MOL, from 277.84 to 301.87 mg/g (15.4% to 41.14%) and for the MOH from 161.01 to 205.3 mg/g (8.93% to 11.4%).For the BG; adsorption on MOS goes from 173.53 to 251.9 mg/g(13% to 49.5%), on MOL from 149.53 to 239.3 mg/g (12.87% to 39.84%) and for MOH from 128.71 to 135,94 mg/g (11.08% to 11.70%).The swelling effect inside the internal structure allows the activation of pollutant molecules at high temperatures.The increase in thermal agitation under these conditions may these results.However, in the case of Pb(II); the sorption on MOH is stable when the temperature goes from 40°C to 50°C for MOL, the adsorption efficiency goes from 96% to 89.3% on this solid.On the other hand, we detected an improvement in the adsorption efficiency for the MOS powder (passing from 88.73% to 94.6%) in the same temperature range.This behaviour indicated that the interaction between active sites and pollutants improved with increasing temperature.This behaviour which caused quick and efficient sorption of Pb(II), CV and BG over the surface of MO and sorption process was maintained through endothermic process.At higher solution temperature, kinetic energy increased, which resulted in enhanced the adsorption of Pb(II), CV and BG.The increase in sorption capacity may suggest that increasing temperature may increase the driving force of Pb(II), CV and BG over the surface of adsorbent.Such influence was associated to the chemical bonding or reaction that occurred in the adsorption process 68 .The thermodynamic parameters during the adsorption of CV, BG and Pb(II) are indicated, respectively, in the following table AM1.The positive values of ∆H ° (except for the case of Pb (II)-MOL and Pb(II)-MOH) confirm the endothermic nature of the adsorption phenomenon.The values of the free enthalpy ΔG ° are all negative, except for the case of CV-MOH and BG-MOS, this confirms the spontaneity of the reaction.
Increasing the temperature of the desired adsorbed solution generally improves the biosorptive elimination of this pollutant because it causes the increase in surface activity and the kinetic energy of the adsorbate, but, in some cases, this increase in temperature can damage the physical structure of the biosorbent, this is the case with MO-GEL which is sensitive to high temperature.
The increasing values of ∆G ° with rising temperature revealed that the adsorption processes were quite satisfactory at higher temperatures.A positive value of ∆H ° was gained, which approved that the adsorption process over the surface of MO was feasible, spontaneous and endothermic in nature.
It could be suggested that the endothermic evolution might occur by increasing the pollutant solution temperature on the exterior boundary layer and therefore improved the inner pores of applied adsorbent.The positive estimations of ∆S ° confirmed the increased randomness over the liquid-solid interface during the adsorption process [67].Since positive value of ΔS, except for Pb(II)-MOL, indicated that system becomes more random and indicates also an increase in the arbitrariness of the solid-phase and liquid-phase interfaces through the sorption.The only negative value of ΔS 0 , in the case of Pb(II)-MOL, can be indicated that there is a reduction in randomness between the solid-liquid interfaces [68] and also suggests that there are some structural changes in the adsorbate and adsorbent [69].In most cases of sorption, observed in table AM1, |TΔS| is larger than ΔH, which reveals that the biosorption is dominated by entropic changes rather than the enthalpic changes.
Knowing that the values of |∆H °| in the physisorption are between 5 and 40 kJ mol −1 [70] and in the chemisorption the heat of sorption is high between 40 and 800 kJ mol −1 [70], in this study, the values of ∆H ° is between 0.79 and 40 with the exception of BG-MOS, Pb(II)-MOS and Pb(II)-MOL.These values of ∆H ° confirmed that sorption in this study proceeded as physisorption and the forces of attraction are the Van der Waal forces or hydrogen bridge.
According to several researchers, the presence of physisorption is followed by chemisorption, so chemisorption is limited to a single layer of molecules on the surface, although it can be followed by additional layers of physically adsorbed molecules.
Therefore, it can be approximately inferred that the probable mechanisms of removal of pollutants on Moringa oleifera are understood as physical and chemical sorption; this is confirmed by other researchers using Moringa oleifera for the sorption of metals [62] and dyes.

Adsorption kinetic models
The study of the physical or/and chemical interactions of the pollutant molecules on the surface of the adsorbent is important because it gives some important indications on the sorption mechanism, the fit of kinetics by PFORE or PSORE is frequently correlated to the predominance of physical or chemical interactions [71], this information is available through the study of kinetics which increases the interest of this study in sorption phenomena.So, in order to determine the stages of the pollutant transport mechanism in batch sorption, it is necessary to go through a study of sorption kinetics.
The results obtained in this study, shown in figure AM9, revealed that the percentage of removal of CV, BG and Pb(II) increases with increasing time and becomes saturated at the contact time between 100 and 150 min for MO and for MO-GEL, except for Pb(II), we noticed fast kinetics.And the elimination rate is significant at the start of the sorption but became slower when it reaches equilibrium.And, any more increasing in the contact time with no other adsorption was observed because the remaining dyes and metal ions become asymptotic with the time axis.Knowing that the kinetics condition is not different and it is improved compared to the isotherm condition, with a fixed dosage of 1 g/L, pH i of 5.00 and a change in equilibrium that does not exceed 0.500 unit and a fixed temperature of almost 30°C.
However, vacant sites have become more difficult to occupy due to the presence of the new forces between solute in the solid phase and bulk in the liquid phase which repulsive each other, the same remark is observed by other researchers [72,73].Then, the final pollutants concentration remained almost similar to each other after a contact time of 100 min to 3-4 h with a difference of � 5% observed.Consequently, they assumed that the process was in a state of equilibrium and that the situation of quasi-equilibrium distinguished at 100 min, but, we have reached a great rate of swelling at 7 h for the beads and the powder we can go until 24-48 h without trace of precipitation.The pseudosecond order results are represented with the kinetics in figure AM9, while pseudo-first order is represented in the additional document(see additional material Figure AM3).
According to Table 2, the q e values of PSORE models were in agreement with the experimental values of q e,exp , but the values of PFORE without far (see Supplementary material section, table AM2).In addition, the correlation coefficients for the pseudosecond order kinetic plots were greater than R 2 ≥ 0.99, if we compare to the pseudo-first order model (PFORE), the PSORE is better suited to the modelling of experimental results because fitted well the experimental results, therefore proved to be a better model than the other PFORE model to interpret the results of sorption which implies according to S. Gogoi et al. [74] and Z. Shirani et al. [75], the results obtained in our study adsorption could be controlled by chemisorption.According to the theory of chemisorption summarised in the paper of J. Zhou et al. [76], and recalled and confirmed by Z. Shirani et al. [75], it is assumed that adsorption is proportional to the content of the specific adsorption sites available on the surface of the sorbents and a chemical bond is formed for the attachment of the adsorbate to the surface of the adsorbent.And, the choice of the sorption site is not random but it searches preferably sought sites that enhance their coordination with the surface [77].
M. Li et al. reported that pseudo-first order kinetics implies that one adsorbate may be adsorbed onto one surface site of sorbents and the applicability of the pseudo secondorder adsorption kinetic rate model indicates that chemisorption may be the rate-limiting step that controls these adsorption processes [78].
The results obtained during the application of the isotherms are confirmed by these kinetic results, as shown in Table 2, and in the table AM2(see Supplementary material section), the pseudo-second order correlation coefficients which have been obtained are very close to 1 (except for some cases) which suggests that the main sorption mechanism could depend on the concentration of contaminants.A similar behaviour was indicated by AN.Módenes et al., (2017) and CDO. Bezerra et al., (2020) [79,49].
In addition, the comparison results of the two kinetic models, PSORE (Figure AM9 and Table 2) vs. PFORE (Figure AM8 and table AM2), indicates that the K 2 constants (between 0.0004×10 −2 -1.09×10 −2 g mg −1 min −1 ) have decreased compared to K 1 (0.0246-0.08 min −1 ), except for Pb(II) we noticed the opposite behaviour.According to MA. Fontecha-Cámara et al., this is due to the increase in the proportion of particles exposed to the constant adsorbent mass and thus increasing the number of collisions between the adsorbent and the adsorbate, producing a rate of higher adsorption in smaller particles [80].
According to the sorption kinetics studies, we concluded that under well-defined conditions of dosage, temperature and pH, the modification carried out can approximate the effectiveness of the powders, without negative effect of the powder during the process of sorption, which means that these new sorbents are good alternatives and are very effective and even we can improve the effectiveness of the powders in some cases because we have seen new characteristics and ease of use of beads compared to the powder.

Adsorption isotherms
The sorption isotherms of CV, BG and Pb(II) on MOS, MOL, MOH, MOS-GEL, MOL-GEL and MOH-GEL are shown in Figure 9.These resultants are confirmed by repeating duplicate the experiments for each concentration to guarantee the precision of the results, usually the error is negligible.
Different forms of isotherms are observed in Figure 9, most isotherms in this figure can be classified as L-shaped, except for CV-MOH (C-curve), MOS-GEL-Pb(II) (S-curve), which means that the Langmuir model is not applicable which is confirmed.Different models have been proposed in the literature in fields of biosorption in order to modelling the experimental data, including 'Langmuir, Freundlich, and Sips equations', but, the Langmuir biosorption isotherm has traditionally been used to quantify and contrast the performance of different biosorbents [81].The application of the Langmuir model suggests that the surface of sorbent is a homogeneous, but, Moringa oleifera may have heterogeneous available ad/sorption sites due to its abundant availability of functional groups, considering the origin of the materials used in this study being plant-based materials [82].In addition, ion exchange has been shown to be a dominant mechanism and typically about two protons are released upon binding of a divalent heavy metal ion.The ion exchange model gives a better representation of the biosorption process because it reflects the fact that most of the biomass is either protonated or contains ions of light metals such as in MO 'K + , Na + , Ca 2+ , Mg 2+ , etc' which are released when a heavy metal cation is bonded (for example).However, the model cannot take into account the influence of pH and ionic strength.
The Langmuir parameter b quantitatively reflects the 'affinity' between the sorbent and the sorbate (for the target pollutants) [83].The q m ×b values are included in table AM2(See Supplementary material section, to see the value of this parameter), in most cases, this parameter, for the MO powder, increase (in comparison with MO-GEL beads).
Moringa Oleivera is rich in functional groups which are different; therefore, with a heterogeneous rich structure that can be used depending on the pollutant and its characteristics.The sips model is used in this work to model the experimental results.These results are found in Figure 9 and the modelling results are grouped in Table 2, the assumption of the Sips model suggests that the adsorption occurs first in the monolayer and subsequently in the multilayer, which is very acceptable, since the Moringa oleifera Lam is a heterogeneous biomaterial [84].According to several researchers, the sips model is used to better describe heterogeneous adsorption, the parameters of this equation are sensitive to the operating conditions like, pH, temperature and initial concentration.This model better describes the adjustments of the Langmuir model in heterogeneous sorption systems in which at low concentrations of contaminant; the model is reduced to isotherm of Freundlich.While in high concentrations, it predicts a monolayer sorption capacity characteristic of the isotherm of Langmuir [85].
Sips plots for CV, BG and Pb(II) sorption using MO and MO-GEL adsorbents showed that the model was will fit and followed by all adsorbent considerably with the regression coefficient (R 2 ) greater than 0.9 for almost all sorption cases.It is noteworthy that the value of q m is sorption capacity at saturation for the Sips equation; this value in the present study is systematically slightly lower than the experimental maximum sorption capacity.As indicated in Table 2, the MO powder has slightly grand q m and q max values than MO-GEL beads.
So, as shown by the R 2 values displayed in table below, and the modelling curves shown in Figure 9, the sips isotherm model was found to be the best fit for the experimental data compared by the other models.For Pb(II), the potential of adsorption was in the order of MOH>MOS>MOL>MOS-GEL>MOH-GEL>MOL-GEL (Figure 9(a)).For CV, the adsorption capacities were in the order of BG-MOL>BG-MOL-GEL>BG-MOS-GEL>BG-MOS>BG-MOH>BG-MOH-GEL and for the CV the order is CV-MOL>CV-MOH>CV-MOS>CV-MOL-GEL>CV-MOH-GEL � CV-MOS-GEL (Figure 9(b,c)).And, we obtained values of q max close to experimental value for most of the biosorbents and pollutants, except for CV-MOL-GEL, BG-MOS-GEL and BG-MOL-GEL.
In our case, the model Langmuir does not give an acceptable result for all the adsorbents and pollutants tested, and it has not given acceptable q max compared to the experimental value.
According to the modelling results; For the Pb(II); Langmuir is not applicable in the case of MOS-GEL, MOL-GEL and MOH-GEL, but for the MO better adjustments of the experimental results for both Sips and Langmuir models, but the Freundlich model is not applicable in this case because it gave us inexplicable results.
While, for the CV and BG; the three models are applicable in the case of CV-MOS, BG-MOS while for the CV-MOH we noticed that the adjustment is almost straight and the model of Freundlich and sips are applied only, knowing that we have already said that the curve is in the form of a C-curve.While for BG-MOL, BG-MOH and CV-MOL we observed that except the model of sips gives a good fit for the modification carried out in this work, and considering the richness still of structure due to this modification the model of sips only gave us a better adjustment compared to the others, the modelling results obtained for the two Langmuir and Freundlich models are summarised in table AM2 (See Supplementary material section VII).
These results confirm that the modelling results are according to the active sites responsible for the fixing of pollutant; therefore, it is not necessary to find the same results for the different pollutants and biosorbant.Therefore, the active sites available on the mixed heterogeneous surface of biosorbents react according to the type of pollutants which makes the sorption heterogeneous challenges (participation of all the sites and with a monolayer adsorption and after multilayer) or with specific site precise which makes the monolayer sorption on a homogeneous surface (Langmuir) therefore its results are acceptable due to the difference in the behaviour of CV, BG and Pb(II) in the surface of different materials (selective sorption each pollutant at its active site responsible for its fixation Langmuir/monolayer or heterogeneous mixed sorption with multilayer sorption sips in the case of the application of two Sips/Langmuir), and these results are acceptable given the vague reaction which is known by these types of biomass, this type of biosorbent has so far been the subject of research for several researchers.Therefore, based on the parameters reported in Table 2, it was concluded that multilayer adsorption of CV, BG and Pb(II) occurred on the heterogeneous active sites of MO (Sips, Freundlich hypothesis) after the formation of the monolayer adsorption that corresponded to Langmuir assumption.

Conclusion
The plant Moringa oleifera 'MO' have been used aiming at developing and applying novel materials for biosorption of CV, BG and Pb(II).These three new sorbents in the form of beads prepared by new innovative conditioning of MO biomass, these beads were successfully developed and used for direct biosorption of some of cationic pollutants like dyes; Crystal violet, Brilliant Green, and metal ions Pb(II).
Due to the importance of MO and their efficiency of the adsorption confirmed by the results obtained, on the one hand, and, on the other hand, the difficulty of the use of the powder on the industrial scale and even in the laboratory in the manipulations adsorption (loss in kinetics and disruption of the results obtained), we used a new method of using/ valorising the powder in gelatin matrix (the incorporation of a powder of MO (powder of moringa oleifera seeds, husks and leaves) into gelatin.The sorption properties of these now beads 'MOS-GEL, MOL-GEL and MOH-GEL' have been compared to a reference material in powder forms (i.e.MOS, MOL and MOH).
Generally, the use of the original porous structure of the beads 'wet without drying' facilitates mass transfer while air-drying irreversibly shrinks the porous structure (with limitations in mass transfer properties) [18], this justifies the use of wet force by several researchers [58].The rehydration of the beads is not able to restore the porous structure but in this case of modification; the rehydration of the beads before use is reversible and it protects the porous structure, the sponge-beads structure with a high rate of swelling/deswelling make this modification very interesting in the field of wastewater (reversiblestructure � H 2 OÞ.
The sorption of these materials increases with the increase of pH due to the progressive deprotonation of carboxylic acid and amine groups, if we compare q max with q exp , R 2 also fit curves, the pseudo second order rate equation gave good fitting of kinetic profiles.
In the case of the biosorption of dyes; sorption isotherms are well described by the sips equation due to the heterogeneous character rich in active functions to the structure of these materials, and the sorption capacity reaches up to 175.28 mg CV g −1 and 493.76 mg BG g −1 for MOS-GEL beads and reaches up to 475.27 mg CV and 575.82 mg BG for MOL-GEL (generally these values much higher than the values reached for MOH-GEL, i.e. 180.48 mg CV g −1 and 267.83 mg BG g −1 ).In the case of the sorption of Pb(II): for MOS-GEL, MOL-GEL and MOH-GEL, allows maintaining a sorption capacity close to 47.08 mg Pb g −1 , 25.99 mg Pb g −1 and 32.25 mg Pb g −1 .
The modelling by the model of sips and seeing the character of MO rich in active function, we can confirm that it has a heterogeneous structure, but the application of these sites in the sorption process for a homogeneous or heterogeneous adsorption is according to the nature of pollutants and sorbent.Consequently, MO and its derivatives have heterogeneous surfaces but the application of the homogeneous and/or heterogeneous sorption sites depends on the nature/size of pollutants and the behaviour/ properties of sorbents.
Several methods are available for the desorption into biomass (and sorbent recycling), but in our study, this is not necessary: the relatively cheap cost of the sorbent does not necessarily require the re-use of the sorbent and metal/dyes could be recovered by simple way.Due to this reversible structure, we have a clear vision that desorption will be easy in an acid medium.However, the preliminary tests showed that metal desorption was easy: using NaOH/HCl eluent for obtaining appreciable levels of desorption.This phenomena due thanks to the presence of amine groups, hydroxyl and carboxylic, due to the ionexchange mechanism, facilitates ion exchange in sorption/desorption

Figure 4 .
Figure 4.The scale of the GLE-MOH, GLE-MOS and GLE-MOL beads before and after adsorption (Swelling rate).

Figure 7 .
Figure 7. Effect of initial pH on the capacity of sorption and efficacity of CV and BG (a) powder (b) beads (black symbols; CV and grey symbols BG) at a room temperature; agitation; 150 tr/min; contact time, t = 3 h; sorbent dosage, SD = 1 g •L −1 , C 0 ≈ 80 mg •L −1 ).

Table 1 .
Characteristics and physico-chemical properties of the synthesised and modified beads.For the pH PZC the protocol used thus the results are detailed in additional material section ( *

Table 2 .
Sorption isotherms -Parameters of the sips models and modelling of resultants of uptake kinetics are used.