REMOVAL OF METHYLENE BLUE BY ADSORPTION ONTO RETAMA RAETAM PLANT : KINETICS AND EQUILIBRIUM STUDY

The feasibility of using medicinal plants species Retama raetam as a low cost and an eco-friendly adsorbent for the adsorption of cationic dye methylene blue from simulated aqueous solution has been investigated. Adsorption kinetics of methylene blue onto Retama raetam plants was studied in a batch system. The effects of pH and contact time were examined. The methylene blue maximum adsorption occurred at pH 8 and the lowest adsorption occurred at pH 2. The apparent equilibrium was reached after 120 min. Optimal experimental conditions were determined. Adsorption modelling parameters for Freundlich and Langmuir isotherms were determined and, based on R, various error distribution functions were evaluated as well. Adsorption isotherm was best described by linear Freundlich isotherm model. Thermodynamic studies show that adsorption was spontaneous and exothermic. For determining the best-fit-kinetic adsorption model, the experimental data were analyzed by using pseudo-firstorder, pseudo-second-order, pseudo-third-order, Esquivel, and Elovich models. Linear regressive and non-linear regressive method was used to obtain the relative parameters. The statistical functions were estimated to find the suitable method that fit better the experimental data. Both methods were appropriate for obtaining the parameters. The linear pseudo-second-order (type 9 and type 10) models were the best to fit the equilibrium data. The present work showed that plant Retama raetam can be used as a low cost adsorbent for the removal of methylene blue from water.


Introduction
The textile industry is one of industrial waste water source.This contaminated water is very toxic for the humans and animals [1].Methylene blue is used in colouring paper, dyeing cottons, wools, silk, leather and coating for paper stock.Although methylene blue is not strongly hazardous, it can cause some harmful effects, such as heartbeat increase, vomiting, shock, cyanosis, jaundice, quadriplegia, and tissue necrosis in human organisms [2].
Chemical coagulation-fl occulation [3], different types of oxidation processes [4], biological process [5], membrane-based separation processes [6] and adsorption [7] were the treatments used in the purifi cation of waters.The most effi cient method used for the quickly removal of dyes from the aqueous solution is the physical adsorption [8].Biosorbents, such as wood sawdust [9], waste-biomass [10], delonix regia [11], agricultural solid waste [12], are able to remove effi ciently the colour from water.
Retama raetam plants can be used as biosorbent.This species belonging to the Fabaceae family has a very productive vertical and horizontal root system, which can reach 20 m.This, in turn, increases substantially the stabilization of the soil.Moreover, the Retama species contributes to the biofertilisation of poor grounds, because of their aptitude to associate with fi xing nitrogen bacteria Rhizobia.Therefore, the genus of Retama is included in a re-vegetation program for degraded areas in semi-arid Mediterranean environments [13].
Retama raetam is a common plant in the North African and East Mediterranean region.In Algeria, it is located in Sahara and Atlas regions and is used in folk medicine under the common name "R'tam" to reduce the blood glucose and skin infl ammations, while in Lebanon it is used as folk herbal medicine against joint aches and in Morocco against skin diseases.Previous pharmacological studies on the plant have revealed its various medicinal properties: antibacterial, antifungal, antihypertensive, antioxidant, antiviral, diuretic, hypoglycaemic, hepatoprotective, nephroprotective and cytotoxic effects.Retama species have been reported to contain fl avonoids and alkaloids [14].
However, there are no reported studies on the adsorption of cationic dyes by Retama raetam.This work aims to understand the potential of Retama raetam for removal of methylene blue dye from simulated aqueous solution in batch mode.The adsorption effi ciency of methylene blue was investigated in order to optimize the experimental parameters.The statistical functions were used to estimate the error deviations between experimental and theoretically predicted adsorption values, including linear and non-linear method.The optimization procedure required a defi ned error function in order to evaluate the fi t of equation to the experimental data.

Materials
Methylene blue (3,7-bis (Dimethylamino)-phenazathionium chloride tetramethylthionine chloride, C 16 H 18 N 3 SCl•3H 2 O, Mw =373.9 g/mol, Figure 1) used in the present study was purchased from Merck (Germany), being selected from the list of dyes normally used in Algeria.Retama raetam plants were collected in Mostaganem region (Algeria), washed several times with deionized water to remove the color and dried at 105°C for 5 h in a convection oven.The residual organics and lipids were respectively removed by methanol and petroleum ether.After this procedure, Retama raetam was washed again with distilled water.

Methods
The Retama raetam was characterized by pH measurement of the pH PZC (point of zero charge).The pH PZC of an adsorbent is a very important characteristic that determines the pH, at which the adsorbent surface has net electrical neutrality [16].
The pH PZC of Retama raetam was measured by pH drift method: 0.1 mg of Retama raetam is added to 100 mL of water with varying pH from 2 to 12 and stirred for 24 h.Final pH of the solution is plotted against initial pH of the solution and shown in Figure 2 [17].The value of pH PZC for Retama raetam was determined as pH 6. Adsorption isotherms are important for the description of how adsorbates interact with an adsorbent being also critical in optimizing the use of adsorbent.Thus, the correlation of equilibrium data using either a theoretical or empirical equation is essential for interpretation of the adsorption data and prediction, as well.Several mathematical models can be used to describe experimental data of adsorption isotherms.Two famous isotherm equations, the Langmuir and Freundlich, were employed for further interpretation of the obtained adsorption data.
Adsorption kinetics of methylene blue onto Retama raetam was studied in a batch system.The effects of pH and equilibrium time were examined.The adsorption parameters were optimized.In each experiment pre weighed amount of adsorbent (0.04 g) was added to 200 mL of dye solution (20 mg/L) taken in a conical fl ask of 250 mL and 0.1 M NaOH or 0.1 M HCl were added to adjust the pH value.This solution was agitated at 300 rpm and centrifuged.The methylene blue concentration in solution was determined at λ max = 665 nm by using UV-1700 PHARMA SPEC SHIMADZU spectrophotometer.The adsorbed amount of methylene blue per mass unit of adsorbent at time t, q (mg/g), (Eq.( 1)) and the dye removal effi ciency (R, %) (Eq.( 2)) were calculated as: where C 0 is the initial concentration of methylene blue (mg/L), C is the dye concentration at time t, V is the solution volume (L) and M is the adsorbent mass (g) [18].
D. Badis et al. / Chem.J. Mold. 2016, 11(2), 74-83 The effect of pH was evaluated by mixing 0.2 g of adsorbent with 1 L of methylene blue simulated aqueous solution of 20 mg/L.The pH value of solution was varied from 2 to 13, by adding 0.1M NaOH or 0.1M HCl solutions.The suspension was shaken for 24h at 25°C.Kinetic experiments were performed by mixing 200 mL of dye solution (20 mg/L) with 0.04 g of adsorbent for different time (5, 10, 30, 60, 90, 120, 150, and 180 min).The initial pH for each dye solution was set at 8. Methylene blue concentration in the supernatants was determined and the adsorbed amount of methylene blue was calculated.

Results and discussion
For studying the effect of every parameter, it is necessary to fi x the values of other ones.The elimination of pollutant from simulated aqueous solution by adsorption is extremely infl uenced by the medium of solution, which affects the nature of the adsorbent surface charge, the ionization extent, the aqueous adsorbate species speciation and the adsorption rate.The adsorptive process through functional groups dissociation on the adsorbate and adsorbent were affected by a pH change [19].The adsorption of methylene blue augments with increasing the pH of the solution.According to the data presented in Figure 3, the best value of adsorption capacity, q e = 9.938 mg/g, was recorded at pH 8. From this study, it is obvious that in the basic medium, the negatively charged species tend to dominate leading to a more negatively charged surface.In this case, the adsorbent surface is negatively charged.The methylene blue adsorption increases due to the enhancement of electrostatic attractions between the negative charge of Retama raetam particles and the positive charge of methylene blue species.The experimental data for methylene blue adsorption on Retama raetam were analyzed with the Freundlich and Langmuir equations.Equations of these models [20] are presented in Table 1, where q is the equilibrium dye concentration on adsorbent (mg/g), q m is the monolayer capacity of the adsorbent (mg/g), C is the equilibrium dye concentration in solution (mg/L), K L is the Langmuir adsorption constant representing the energy constant related to the heat of adsorption, n and K F are Freundlich constants related to adsorption intensity of the adsorbent and adsorption.A non-linear and linear fi tting procedure using Excel and Origin software were used, respectively.The constants of all models were given in Table 2.

Table 1
Adsorption isotherms models and their linear and non linear forms [20].

Applied model
Non linear form Linear form The coeffi cient of correlation indicated that Freundlich isotherm fi tted the experimental data better than Langmuir isotherm.Good agreement between the experimental isotherms and the Freundlich model was found in the case of systems: pentachlorophenol/(M)Al-MCM-41 [21], and toluene/activated carbon [22].
The optimization procedure required a defi ned error function in order to evaluate the fi t of equation to the experimental data.The best-fi tting equation is determined using the well-known special functions to calculate the error deviation between experimental and predicted data.The mathematical equations of these error functions were illustrated in Table 3.  [30] where n is the number of experimental data points, q calc is the predicted (calculated) quantity of methylene blue adsorbed onto Retama raetam, q exp is the experimental data, p is the number of parameters in each kinetic model, ARED is the average relative error deviation (dimensionless parameter), ARE is the average relative error (dimensionless parameter), ARS is the average relative standard error (dimensionless parameter), HYBRID is the hybrid fractional error function (dimensionless parameter), MPSD Marquardt's is the percent standard deviation (dimensionless parameter), MPSED Marquardt's is the percent standard deviation (dimensionless parameter), SAE=EABS is the sum of absolute error (mg/g), SSE is the sum of the squares of the errors (mg/g) 2 , and Δq(%) is the normalized standard deviation (mg/g).The constants of all error analysis are represented in Table 4.The data of adsorption isotherm are essentially required for designing the adsorption systems.In order to optimize the design of a specifi c sorbate/sorbent system for removal of methylene blue from aqueous solution, it is important to establish the most appropriate correlation for the experimental kinetic data.Applicability of some statistical tools to predict the optimum adsorption isotherms of methylene blue onto Retama raetam after linear regression analysis showed that the highest R 2 value and the lowest ARED, ARE, SAE, ARS, MPSD, Δq, SSE, MSPED and HYBRID values can be suitable and meaningful tools to predict the best-fi tting equation models.
The best fi tting is determined based on the use of these functions for calculation of the error deviation between experimental and predicted equilibrium adsorption isotherm data, after linear analysis.Hence, according to Table 4, it seems that the linear Freundlich model was the most suitable mode to describe satisfactorily the studied adsorption phenomenon.Therefore, based on the mentioned results, the best useful error estimation statistical tools point out the non linear Freundlich model, followed by linear Freundlich model, as the best-fi tting models.
In order to better understand the effect of temperature on the adsorption of methylene blue onto Retama raetam, the free energy change (ΔG °, J mol -1 ), enthalpy change (ΔH °, J mol -1 ) and entropy change (ΔS °, J K -1 mol -1 ) were determined (such parameters refl ect the feasibility and spontaneous nature of the process) using Eqs.( 3)- (5).
The combination of Eqs.( 3) and (4) gives Eq.( 5): where R is the universal gas constant (8.314J K -1 mol -1 ), T is the absolute temperature (Kelvin) [31].Experiments were performed using 20 mg/L dye solutions with 0.2 g of Retama raetam for 24 h at various temperatures.The apparent equilibrium constant K c of the adsorption is defi ned as Eq.( 6) [20]: The enthalpy and entropy can be obtained from the slope and intercept of the linear plot of lnK c versus 1/T.The obtained thermodynamic parameters are given in Table 5.A negative enthalpy value of −4.51110 5 kJ/mol indicates that adsorption was exothermic.A negative entropy value of -62.687J/mol and a negatively decreasing Gibbs free energy indicates the increase in the randomness in the solid-liquid interface and adsorption spontaneity [32].
Figure 4 illustrates the effect of contact time on decolorization (dye adsorption) with Retama raetam.The plot (simulated aqueous solution) can be divided in three zones: (i) 0-30 min, which indicate the fast adsorption of methylene blue, suggesting a rapid external diffusion and surface adsorption; (ii) 30-60 min, show a gradual equilibrium, and (iii) 60-180 min, indicate the plateau of the equilibrium state.The adsorption was rapid at the initial stage of the contact, but it gradually slowed down until the equilibrium.The fast adsorption at the initial stage can be attributed to the fact that a large number of surface sites are available for adsorption.After a lapse of time, the remaining surface sites are diffi cult to be occupied.Adsorption is a complex process that is infl uenced by several parameters related to adsorbent and to the physicochemical conditions, under which the process is carried out [33].For understanding the mechanism of the adsorption process, the following equations: pseudo-fi rst order (Lagergren Model) [2], pseudo-second order [34], Esquivel [35], pseudo-third order [36], and Elovich [37] were selected to fi t the experimental kinetic data.Equations of these models are presented in Table 6.

Table 6
Adsorption kinetics models and their linear and non linear forms.

Applied model Non Linear form Linear form Reference Pseudo-fi rst order
Pseudo-fi rst order (type 1) ) 1 (

Pseudo-second order
Pseudo-second order (type 9)   [39] Pseudo-second order (type 10) where k 1 is pseudo-fi rst order rate constant (min -1 ), k 2 is pseudo-second order rate constant (g/(mg min)) , k 3 is pseudo-third order rate constant (g 2 /(mg 2 min)), K E is Esquivel rate constant (min), k 4 is Elovich rate constant (mg/(g min)), k 5 is extent of surface coverage and activation energy of the process (g/mg), k 6 is extent of surface coverage and activation energy of the process (g/mg), k 7 is Elovich rate constant (mg/(g min)), q e is amount of adsorption at equilibrium (mg/g), and θ is dimensionless parameter (=q/q e ).For the non-linear and linear fi tting procedures Excel and Origin software were used, respectively.The constants of all models were given in 4.167 0.899 q = 0.240*ln(t) + 8.756 Table 7 shows that q e , k 2 and R 2 values obtained from the two linear forms of pseudo-secondorder expressions were the same.The value of q e and k 2 were calculated to be, respectively, 9.932 mg g -1 and 1.926 g mg -1 min -1 for linear pseudo-second-order and 9.908 mg g -1 and 1.97 g mg -1 min -1 for non linear pseudo-second order biosorption.The constants of all error analysis are represented in Table 8.
Adsorption kinetic data are the basic requirements for the design of adsorption systems.In order to optimize the design of a specifi c sorbate/sorbent system to remove methylene blue from aqueous solution, it is important to establish the most appropriate correlation for the experimental kinetic data.Applicability of some statistical tools to predict optimum adsorption kinetics of methylene blue onto Retama raetam after linear regression analysis showed that the highest R 2 value and the lowest ARED, ARE, SAE, ARS, MPSD, Δq, SSE, MSPED, and HYBRID values could be suitable and meaningful tools to predict the best-fi tting equation models.
The best fi tting is determined based on the use of these functions to calculate the error deviation between experimental and predicted equilibrium adsorption kinetic data, after linear analysis.Hence, according to Table 4, it seems that the linear pseudo-second order type 9 and type 10 models were the most suitable models to describe satisfactorily the studied adsorption phenomenon.Therefore, based on these mentioned results, the best useful error estimation statistical tools should point out the linear pseudo-second order type 9 followed by linear pseudo-second order type 10 as the best-fi tting models.D. Badis et al. / Chem. J. Mold. 2016, 11(2), [74][75][76][77][78][79][80][81][82][83] In the most studied adsorption systems, the pseudo-fi rst-order model does not fi t well over the entire adsorption period and is generally applicable over the fi rst 20-30 min of the sorption process.The pseudo-second-order model is based on the biosorption capacity of the solid phase and it generally predicts the "chemisorption" behaviour over the whole time of adsorption [20].
Obtained results, presented in Table 4 show that the pseudo-fi rst-order model data do not fall on straight lines indicating that this model was less appropriate.In contrast, the pseudo second order kinetics have shown very low ARED, ARE, SAE, ARS, MPSD, Δq, SSE, MSPED, and HYBRID and high R 2 values for type 9, 10 linear pseudo-secondorder, and non linear pseudo-second-order expressions suggest that it is appropriate to use the pseudo-second-order model, suggesting that it is applicable to the adsorption kinetics.This suggests that, the biosorption of methylene blue onto Retama raetam is a chemisorption process involving exchange or sharing of electrons mainly between the dye ions and the sorbent functional groups [20].Using linear method it was found that a theoretical pseudo-second order model represents well the experimental kinetic data of adsorption of methylene blue onto Retama raetam based on a Type 9 and 10 pseudo-second-order kinetic expression.
Studies regarding the use of Retama raetam as biosorbent are in progress.More technical and experimental optimisations and treatments should be realised to improve the adsorption capacity of Retama raetam.For example, use of more effective pre-treatment methods and reduction in particle size (larger specifi c adsorption area, m 2 /g) may further improve the rate and the extent of adsorption of methylene blue onto Retama raetam.Besides, the methylene blue -loaded biomass itself has to be treated, in order to avoid a pollution transfer.Indeed, one of the more common questions aroused by biosorption processes involves the fate of the biosorbent after the process.Care must be taken that solving one problem, not to create another.The sorbed methylene blue can be recovered by extraction from the biomass in order to be concentrated and then stored, reused, or eliminated.Also, the decontamination of the methylene blue -loaded biomass by biodegradation is a very interesting approach.

Conclusions
Retama raetam plant was used for the adsorption of methylene blue in simulated aqueous solution.In batch mode, the adsorption was highly dependent on two operating parameters (pH, contact time).The obtained results revealed the following optimal conditions: pH value of 8 and 120 min of contact time, which lead to 90.38 % methylene blue removal.
Kinetics data correlated well with the pseudo second order kinetic model (type 9 and type 10), whereas equilibrium study was best described by non linear Freundlich isotherm model.

Figure 2 .
Figure 2. Point of zero charge (pH PZC ) of the Retama raetam used for the adsorption experiments.

Figure 3 .
Figure 3.Effect of the initial pH of solution on equilibrium adsorption capacity of Retama raetam.