Removal of methylene blue by orange and uvaia seeds

In this study, the adsorption behavior of methylene blue (MB) was investigated using orange seed (OS) and uvaia seed (US) as low cost adsorbents. These materials were characterized using FTIR (Fourier transform infrared spectroscopy), elemental chemical analysis (CHNO), thermogravimetric analysis (TGA), zeta potential, specific surface area and a test for determination of acid sites. The parameters evaluated in the kinetic study of adsorption were contact time, initial dye concentration, mass of adsorbent and pH. The adsorption of MB onto OS and US samples could be explained by Elovich ́s kinetic model. The experimental isotherms data, carried out at the temperatures of 25°C, 35°C, 45°C and 55°C, were better represented by Sips. It was verified that the adsorption was spontaneous and endothermic according to the thermodynamic parameters ΔG°, ΔH ° and ΔS ° evaluated. The OS and US maximum adsorption capacity at the temperature of 25°C was 38 mg g -1 and 48 mg g -1 , respectively. The reuse tests performed at three subsequent times showed that there was no significant decrease regarding the materials efficiency, therefore emphasizing its viability as biosorbents.


INTRODUCTION
As the population has been growing continuously, most common environmental issues from industrial processes, agriculture, mining, and domestic activities have been observed [1]. For instance, the textile dyeing industry consumes large quantities of water and produces large volumes of wastewater from different steps. These substances are very rich in colors and it is generally applied at food, cosmetics, leather, and pharmaceuticals among other industries segments [2].
Despite the high applicability, most dyes contain chemicals that can harm human health and the environment and the biological stability of the molecules makes it hard to be degraded by conventional treatment systems used by industries [3].
Due to the intense coloration, the passage of solar radiation may be restrict, decreasing natural photosynthetic activities and letting to changes in the aquatic biota [4]. Dye methylene blue belongs to the thiazines class and it is widely used as a model molecule for pollution indicator as well as studies of mesoporosity and functional groups [5].
The chemical, biological, and electrochemically assisted photocatalytic degradation of reactive dyes [6,7], the precipitation processes and photo-Fenton [8,9] are examples of varieties of several decontamination methods for the removal of pollutants from effluents. However, adsorption is one of the most studied methods since it produces a high quality treated effluent. It also has low cost, it is ease to operate and there are possibilities of reusing the water and the adsorbent that can be recycled afterwards [10].
Activated carbon is the most used adsorbent by the industry. However, it has some disadvantages related to production costs and low regenerability [11]. Therefore, researches have been conducted regarding the development of alternative adsorbent materials. For example, agroindustrial wastes such as corncobs [12], cane bagasse [13], coconut fibers [14], coffee beans [15] have been studied with the aim of employing efficient, low cost regenerable adsorbents.
Orange cultivation plays a very important economic role, being one of the most cultivated and consumed fruit in the world. The Brazilian citriculture stands out as one of the main economic activities, driving the country's economy expressively. The uvaia is another fruit which presents high industrial potentialities, once they are used in the manufacture of juices, liquors, jellies, and others. However, most part of these fruits, including shells, bagasse and seeds, are not reused in the process. In this way, this work aims at studying the efficiency of orange seed (OS) and uvaia seed (US) as adsorbent materials in the adsorption of MB dye, potentially adding value to these wastes.

Preparation of adsorbents
The samples of OS and US were dried for 16 h at 40° C, ground and sieved to pass through a 0.425 mm sieve (35 Tyler series).

Characterization of adsorbents
An Elementar Analysensysteme vario MICRO cubeTM was employed to determine the percentages of C, H, N, S and O (by difference) in OS and US. Thermogravimetric analyses were performed using a Shimadzu model DTG-60AH thermomechanic analyzer (Shimadzu model DTG-60AH) and were carried out under a nitrogen atmosphere in the temperature range of 25-900 °C at a heating rate of 10 °C min−1. Fourier transform infrared spectroscopy (FTIR) in the 400 to 4000 cm -1 range a Bruker Vertex 70V. Microscopic observations and electron micrographs were made using a Nano Technology Systems model Evo® 40 VP SEM. For the purpose of determining zeta potentials, adsorbents were ground (< 37 μm particle size), the suspensions adjusted to the required pH (in the range of 2-11) and sedimented/conditioned for 2 h at 22 o C in 250 mL erlenmeyer flasks containing sodium nitrate solution (0.002 mol L −1 ) as supporting electrolyte. Potentials were measured using a Zeta Meter System 3.0+ ZM3-D-G instrument. The applied tension varied between 75 and 200 mV, and zeta potentials were expressed as the average values of 20 repetitions. Acidity was determined using 0.1 g of the adsorbents and 20 mL of potassium hydroxide solution (0.01 mol L -1 ). The system was maintained at a resolution of 4 cm -1 , using KBr pellets (300 mg of KBr to 3 mg of was carried out sample). The analyses were carried out using at 25 o C on an orbital shaker (50 rpm) for 3 hours, followed by titration with hydrochloric acid solution (0.01mol L -1 ).

Preparation of dye solutions
The adsorption tests were carried out using the methylene blue dye (MB) (VETEC) as adsorbate. The different concentrations tested in the experiments were diluted from a 2 g L -1 stock solution of MB. The natural pH values of the stock solution was 5.5, and it was adjusted (when necessary) by adding either potassium hydroxide 0.01 mol L −1 or hydrochloric acid 0.01 mol L −1 . The structural formula of MB, properties and characteristics are described in Table 1.

Adsorption experiments
A monitored 24-hour analysis was carried out to determining the equilibrium time. For this, 10 mL of methylene blue solution at an initial concentration of 25 mg L -1 at natural pH was placed on contact with 0.1 g of each adsorbent (US and OS).
To establish the influence of the initial concentration, solutions at concentrations corresponding to 25, 50 and 100 mg L -1 were used at an adsorbent mass/adsorbate volume ratio of 1:100 (0.1 g of adsorbent).
In order to study the adsorbent mass/adsorbate volume ratio, 10 mL of methylene blue solution, at natural pH were placed in contact with 0.2, 0.1 and 0.05g of adsorbent, which led to ratios of 1:50, 1:100 and 1:200, respectively.
The last parameter analyzed was the influence of pH. For that, a solution of methylene blue was prepared at an optimized initial methylene blue concentration, adjusting the pH to 3, 4, 7, 8 and 9, using acetic acid or potassium hydroxide solution 0.1 mol L -1 .
The resulting mixtures were maintained at room temperature (25±1 o C) on an orbital shaker (200 rpm). The supernatants were then separated by centrifugation (5 min at 1540 x g) and diluted (when necessary) so that the remaining concentrations of MB could be determined at 665 nm, using a Femto model 800 XI UV-vis spectrometer.
The dye removal percentage (%) was calculated by using the following equation 1: where Co is the initial dye concentration (mg L −1 ) and Ct is the dye concentration (mg L −1 ) at time any time. All the samples were analyzed in duplicate to ensure data reproducibility.

Adsorption isotherms
The adsorption isotherms were constructed using optimized parameters. For this, solutions of adsorbate (MB) at concentrations ranging from 10 to 2000 mg L -1 were prepared. The quantity of dye adsorbed by mass of OS and US at equilibrium was determined from the equation: J a n u a r y 0 3 , 2 0 1 5 where Qe (mg g -1 ) is the quantity of MB adsorbed by the mass of OS and US at equilibrium, Co is the initial and Ce is the concentration at equilibrium of MB (mg L -1 ), V is the volume of the solution (L) and m is the mass of adsorbent (g). All experiments were carried out in duplicate.

Thermodynamics of the adsorption process
The temperature effect on the adsorption of MB on OS and US was studied at temperatures of 25 o C, 35 o C, 45 o C and 55 °C. The thermodynamic parameters enthalpy (∆H°), entropy (∆S°) and Gibbs free energy (∆G) can be calculated from equations 3, 4 and 5.
where KL represents the Langmuir equilibrium constant (L mol -1 ), R the universal gas constant (8.314 J mol -1 K -1 ) and T the temperature in Kelvin (K). The values of ΔH o and ΔS o were obtained from the curve fitting coefficients of the graphic ln KL versus 1 / T.

Reuse test
The desorption of the amount of MB adsorbed on OS and US samples was carried out using HCl 0.1 mol L -1 solution and an adsorbent mass (g), in a ratio of 1:10. The systems were kept under agitation for 2 hours at 120 rpm and at the end of each experiment the supernatants were collected and the amount of dye removed was then determined by UV-Vis spectroscopy at 665 nm. The adsorbents were water-washed, vacuum filtered and placed in an oven at 50 °C for 2 h for drying. The regenerated adsorbents were used in three subsequent adsorption-desorption cycles.

Characteristics of the absorbents
The direct (TGA) and differential (DTG) thermogravimetric curves for OS and US show that mass loss occurred at three different stages in both adsorbents, with similar characteristics (Figure 1). The initial loss of mass at approximately 100 o C was associated with the elimination of water and small volatile molecules, while the second reduction occurred between 300 and 350 o C and was caused by the thermal degradation of cellulose and hemicelluloses. The final mass loss at approximately 400 o C may be attributed to the degradation of lignin, which has a much higher thermal stability than either cellulose or hemicellulose polymers. The greatest rate of mass loss occurred at the latter temperature [16]. The elemental composition (Table 2) indicates that the adsorbents materials were rich in oxygen but contained low levels of sulfur. Both seeds presented similar C/ H ratio. However, for C/O, US had a significantly smaller ratio compared to OS. The FTIR spectrum of OS and US ( Figure 2) showed a broad peak at 3400 cm −1 indicating hydroxyl groups, characteristic of OH stretches corresponding to pectin, cellulose, water, hemicellulose and lignin [17,18]. The bands located at 2931 cm -1 and 2924 cm -1 can be assigned to the CH stretching of methyl and methoxy groups. The band identified at about 1744 cm -1 confirms the presence of the carbonyl group (C = O) that extends to vibrations of carboxyl groups present in hemicellulose, pectin and lignin. The band located at 1638 cm -1 was attributed to the stretching of C = C bound constituents of the aromatic benzene molecules or rings in lignin. The bands located between 1300 and 1000 cm  The adsorbent materials were submitted to chemical titration of acid ligand sites. The results indicated that OS had an acidity corresponding to 3.11 mmol g -1 , which was higher than the US (up to 2.78 mmol g -1 ). The high acidity found in the materials can be attributed to the presence of Bronsted acidic groups (carboxylic acid and alcohols) identified in the FTIR analysis.
Zeta potential is a measurement of the superficial charge of the adsorbent at a specific pH and indicates which type of ion would be adsorbed under such conditions. As shown in Figure 3, both OS and US presented negative surface charges throughout the pH range studied (3.0-10.5), meaning that both adsorbents would exhibit a great affinity for cations. The negative charge present on the surface of lignocellulosic materials is associated with acidic entities such as carboxyl and phenolic OH groups.

Influence of physicochemical parameters on adsorption
The rates of adsorption of MB as a function of time of contact with the absorbent are shown in Figure 4. As observed, the equilibrium times for OS and US were 4 and 8 hours, respectively. The difference in the equilibrium times of the adsorbent materials can be attributed to the bio-sorbents chemical composition as well as the adsorption sites accessibility. J a n u a r y 0 3 , 2 0 1 5 The analysis results show that an increase of the initial concentration of the dye reduced the adsorption the adsorbents capacities, but this parameter was most significant for US sample. The lowest removal when higher concentrations were used can be attributed to difficulties in diffusion capacity of the adsorbate molecules or even the competition among the molecules that are being adsorbed on the active surface sites of the adsorbent material [19].
The study of the mass of bio-sorbents ( Figure 5) showed that the optimal amounts for continuing the adsorption process were 0.1 g and 0.05 g for OS and US, respectively.

. Influence of the initial concentration on the adsorption of MB
The influences of variable pH values are shown in Figure 6. It is observed that for both bio-sorbents a lower percentage removal of dye occurred at low pH values (pH 3 and 4), which can be attributed to the association of H3O+ ions with the adsorbents surface, limiting the approximation of MB, which is cationic dye [20]. There was an increase in adsorption capacity at pH 5.5 (natural) and above this there is a significantly decrease in the adsorption process efficiency. J a n u a r y 0 3 , 2 0 1 5

Kinetic characteristics of the adsorbents
The kinetic parameters of adsorption of MB onto OS and US were determined under optimized conditions. The data were analyzed using models represented mathematically by non-linear equations of Table 3.

Models Equations
Pseudo-first order [27] ) 1 ( 1 t k e t e q q   Pseudo-second order [28] t k Q t k q q e e t 2 2 2 1  Intra-particle diffusion [29] C t k q ID t   Elovich [30] ) t ln( Where, qt is the amount of dye removed in time t (mg g -1 ), qe is the amount of dye removed at equilibrium (mg g -1 ), k1 is the rate constant of pseudo first order (h -1 ), t is the contact time, k2 is the constant of pseudo second order (g mg -1 h -1 ), kID is the rate constant of intra-particle diffusion, C is the constant related to the diffusion layer thickness, β is the relation between the surface coverage degree and the activation energy involved in the chemisorption, t0 is the initial time , kAV is the Avrami kinetic constant and nAV is a constant related to the kinetics of adsorption reactions.
It was observed that the experimental data of the adsorbents OS and US were better fitted to the kinetic model described by Elovich (Figure 7 and Table 4). The proposed model describes the adsorption maintained by chemisorption in heterogeneous surfaces in a process relatively slow and justified by the equilibrium times of 4 and 8 hours to OS and US samples, respectively [21]. J a n u a r y 0 3 , 2 0 1 5  Error=0.4770 J a n u a r y 0 3 , 2 0 1 5

Adsorption isotherms of the adsorbents
The adsorption isotherms of OS and US for the removal of MB under optimized conditions are shown in Figure 8. The data obtained were fitted to the isotherm models represented, mathematically, using non-linear equations (Table 5)

Time (h)
Experimental dates Pseudo-first order Pseudo-second order Intraparticle diffusion Elovich Avrami Figure 8. Fits of the kinetic models for OS and US J a n u a r y 0 3 , 2 0 1 5

US
Where, qe is the amount of adsorbate adsorbed per unit of adsorbent mass at equilibrium (g g -1 ), qm is the adsorption capacity of the monolayer (L mg -1 ), KL is the equilibrium constant (L mg -1 ), Ce is the adsorbate solution concentration at equilibrium (mg L -1 ), KF is the Freundlich constant (mg 1-1/n kg -1 L 1/n ) , 1/n represents the adsorption intensity, KS is the Sips equilibrium constant related to the adsorption energy (L mg -1 ) 1/m , 1/m is the Sips model exponent that characterizes the system heterogeneity, bDR is the constant of adsorption energy (mol 2 K J -2 ), R is the gases universal constant (kJ mol -1 K -1 ), T is the temperature (K).
The parameters results obtained to adsorption in OS and US are shown in Table 6. 1.4762 J a n u a r y 0 3 , 2 0 1 5 According to the correlation coefficients, it was verified that the data were better fitted to the Langmuir and Sips models to OS and US at all temperatures. Although the determination coefficients were close, the Langmuir model showed the highest standard error value. Thus, the experimental data have were better fitted to the Sips model ( Figure 9). These results confirm the superficial heterogeneity of adsorbent materials when using the methylene blue dye as adsorbate [22,23]. J a n u a r y 0 3 , 2 0 1 5

Thermodynamic studies
The results obtained to the MB adsorption onto OS and US conduced at different temperatures are shown in Table 7. J a n u a r y 0 3 , 2 0 1 5 As it can be seen, the values of qm,exp increased with the increasing temperature, which characterizes an endothermic process in OS and US, justified by the ΔH o positive values for both adsorbents. The ΔHo values between 40 and 120 kJ mol -1 reinforce the chemisorption characteristic. It was observed with increasing temperature that the Gibbs free energy became more negative, indicating an increase in the adsorption process spontaneity for both seeds. The ΔS o positive value to the dye adsorption in both seeds show an increase in entropy (or disorder) in the solid / liquid interface, as a result of MB adsorption [24]. Increasing adsorption with increasing temperature can be attributed to the adsorption sites dilation.
The increase in the adsorbate molecules diffusion in the external boundary layer and also inside the adsorbent inner layer are due to a decrease in the solution viscosity and an increase of the dye molecules mobility, which obtains an appropriate kinetic energy to interact with the OS and US surfaces. These results suggest that the processes involve chemisorption in adsorbents. Similar results have also been observed in other studies performed with different materials [25,26].

Specific surface area of the adsorbents
Values for the specific surface area of the adsorbent were calculated from the following equation [17]: where qm is the maximum adsorption capacity (mg g −1 ), NA is Avogadro's number (6.022 ×1023 mol −1 ), MMMB is the molecular weight of MB (319.85 g mol −1 ) and  is the area occupied by a single adsorbed MB molecule (130Å 2 ).
Considering the experimental results and that the area occupied by an adsorbed molecule of this dye is 130 Å 2 , the specific surface area of the adsorbents OS and US (estimated by Eq. 6) were 91.48 e 116.48 m 2 g -1 , respectively.

Adsorbent materials reusing test: desorption
The results of the three adsorption/desorption cycles showed that the removal efficiency of the dye was reduced by 24% (from 90% to 69%) and 10% (from 97% to 87%) to the OS and US, respectively. The results showed that the US had higher regenerative capacity, allowing it to be a good source suitable to being used as an adsorbent material.

Concluding Remarks
The uvaia seed (US) was more efficient than the orange seed (OS) in the removal of methylene blue. Both materials are very promising in the removal of contaminants from aqueous effluents. Further, after three consecutive reuse cycles proved effective, both emphasize their viability as bio-sorbents. J a n u a r y 0 3 , 2 0 1 5