Biosorption of methylene blue onto Arthrospira ...

Journal of Environmental Chemical Engineering 3 (2015) 670–680

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering journal homepage: www.elsevier.com/locate/jece

Biosorption of methylene blue onto Arthrospira platensis biomass: Kinetic, equilibrium and thermodynamic studies Dimitris Mitrogiannis a, * , Giorgos Markou a , Abuzer Çelekli b , Hüseyin Bozkurt c a b c

Department of Natural Resources Management and Agricultural Engineering, Agricultural University of Athens, Iera Odos 75, Athens 11855, Greece Department of Biology, Faculty of Art and Science, University of Gaziantep, Gaziantep 27310, Turkey Department of Food Engineering, Faculty of Engineering, University of Gaziantep, Gaziantep 27310, Turkey

A R T I C L E I N F O

A B S T R A C T

Article history: Received 12 November 2014 Accepted 11 February 2015 Available online 17 February 2015

In this study, Arthrospira platensis biomass was employed as a biosorbent for the removal of methylene blue (MB) dye from aqueous solutions. The kinetic data were better described by the pseudo-second order model and equilibrium was established within 60–120 min. The intra-particle diffusion was not the only rate-limiting step and film diffusion might contribute to MB biosorption process. The increase of temperature from 298 to 318 K caused a decrease of biosorption capacity. The Langmuir, Freundlich and Dubinin–Radushkevich (D–R) isotherm models described well the experimental equilibrium data at all studied temperatures. The maximum monolayer adsorption capacity (qmax) was 312.5 mg MB/g at 298 K and pH 7.5. According to the results of the thermodynamic analysis and the release of exchangeable cations from the biomass surface, physical sorption and ion exchange were the dominant mechanisms of MB biosorption at lower temperature. Methanol esterification of the dried biomass showed the involvement of carboxyl functional groups in MB chemisorption. The thermodynamic parameters indicated that MB biosorption onto A. platensis was a spontaneous, favorable and exothermic process. The biosorption results showed that A. platensis could be employed as an efficient and eco-friendly biosorbent for the removal of cationic dyes. ã 2015 Elsevier Ltd. All rights reserved.

Keywords: Arthrospira platensis Methylene blue Cationic dye Thermodynamics Biosorption mechanism Cation exchange

Introduction Synthetic dyes are hazardous pollutants which present toxic and aesthetic effects in aquatic environments. Dye effluents, containing colored organic molecules, increase the organic load of water bodies and reduce the sunlight penetration, affecting the photosynthetic activity of phytoplankton and disturbing the ecological balance of the aquatic environments. Moreover, some dyes display carcinogenic and mutagenic activity [1,2]. Potential sources of dyes are textile, leather, paper, printing, plastic, electroplating, food and cosmetic industries. Various physical, chemical and biological methods have been investigated for the treatment of wastewaters contaminated with synthetic dyes [3]. However, each of these technologies has its disadvantages, such as high operational and initial capital costs, low efficiency at low dye concentrations and production of undesirable sludge [4]. Among treatment technologies, adsorption is considered as an effective method for dye removal using low-cost materials.

* Corresponding author. Tel.: +30 6974876236. E-mail address: [email protected] (D. Mitrogiannis). http://dx.doi.org/10.1016/j.jece.2015.02.008 2213-3437/ ã 2015 Elsevier Ltd. All rights reserved.

Although activated carbon is the most commonly used adsorbent and is very efficient to remove dyes from wastewater, it presents high costs of production and regeneration [5]. A number of studies have been made to find cost-effective and eco-friendly methods for treatment of dye wastewaters using cheep biomaterials as adsorbents [3]. Algae and cyanobacteria have gained interest as alternative biosorbents due to their high binding affinity, their higher sorption selectivity for pollutants than commercial ion-exchange resins and activated carbon, and due to their capability of growing using wastewater as cultivation medium [3,4,6,7]. The filamentous cyanobacterium Arthrospira platensis is a potential biosorbent, having several advantages, such as relative high growth rates, high biomass productivity, ease of cell harvesting and biomass composition manipulation [8]. The surface of A. platensis consists of various macro-molecules with diverse functional groups such as carboxyl, hydroxyl, sulphate and phosphate, which are responsible for dye binding [9]. A. platensis has already been studied for the removal of inorganic pollutants such as heavy metals [6,10–12] and organic pollutants such as anionic dyes [9,13–15] and phenol [16,17] from aqueous solutions. To our knowledge, there is lack of published work about the adsorption of cationic dyes onto A.

D. Mitrogiannis et al. / Journal of Environmental Chemical Engineering 3 (2015) 670–680

platensis. The only related study to this, uses an artificial neural network to predict the biosorption capacity of methylene blue onto Spirulina sp. [18]. However, there is no literature information about the biosorption kinetics and thermodynamics of a cationic dye on this cyanobacterium and about the contribution of the ion exchange mechanism on dye removal. Although the important role of the ion exchange mechanism in MB removal by various biosorbents is mentioned very often, it has not been widely investigated by detection measures [7]. Methylene blue (MB) is a common cationic dye used for dyeing paper, cotton, wool and silk [7,19]. The harmful effects of MB include: breathing difficulties, nausea, vomiting, tissue necrosis, profuse sweating, mental confusion, cyanosis and methemoglobinemia [5,7]. MB has been widely employed as a model cationic dye in adsorption studies, using low-cost adsorbents such as natural minerals (clays, zeolites and perlite), activated carbon, dead or non-growing microbial biomass, agricultural and industrial wastes [7]. The aim of the present study was to investigate the potential of A. platensis dry biomass to remove MB dye from aqueous solutions. The effect of solution pH, initial MB concentration, contact time, temperature and ionic strength on the biosorption capacity was investigated. Kinetic, isotherm and thermodynamic parameters were estimated to understand the biosorption rate and mechanisms of MB onto A. platensis.

Batch biosorption experiments The biosorption experiments were carried out in batch mode by mixing 12.5 mL aqueous suspension containing 12.5 mg dried biomass with 12.5 mL MB dye solution of known concentration. The final 25 mL solution was placed in a 50 mL plastic flask, which was sealed and agitated with a rotary shaker at 140 rpm. The desired initial pH (range 4–10) of the adsorbate and adsorbent solution was adjusted using 0.1 M HNO3 and/or NaOH before mixing them. Biosorption kinetics were investigated with a biomass concentration of 0.5 g/L at three initial dye concentrations (25, 50 and 100 mg/L) and pH 7.5 0.1. Samples were collected at time intervals (2, 5, 10, 15, 30, 60, 90, 120, 180 and 240 min) and subjected to MB concentration determination. The kinetic experiments were conducted in an air-conditioned room with temperature of 298–300 K. Equilibrium experiments were carried out at 298, 308 and 318 K, placing the flasks and shaker in a temperature controlled incubator and using five different initial MB concentrations (6.25, 12.5, 25, 50 and 100 mg/L), in order to estimate the parameters of isotherm models and thermodynamic equations. The contact time of equilibrium experiments was chosen to be 24 h. The amount of MB adsorbed per unit weight of A. platensis biomass at equilibrium, qe (mg/g), and the percentage dye removal (R%), were calculated with the following equations:

Materials and methods qe ¼ Biosorbent cultivation and preparation The cyanobacterium A. platensis (SAG 21.99) used in this study was cultivated in Zarrouk medium within 10 L plastic cubical photobioreactor, which were kept at 303 2 K in semi-continuous cultivation mode with a dilution rate of 0.11/d [6]. The A. platensis biomass was harvested by filtration and rinsed with deionized (DI) water. The cultivation medium salts were removed by washing the biomass twice by re-suspension in DI water. After that the biomass was separated with centrifugation (5000 rpm for 5 min) and dried overnight in an oven at 353 K. The dried biomass was milled (IKA Labortechnik, A10), sieved through a metal sieve (100 mesh, particle diameter pHzpc the biosorbent surface is negatively charged due to the deprotonation of functional groups such as carboxyl, amino,


673

phosphate and hydroxyl [13,21], and thus electrostatic attraction can occur between the negatively charged functional groups of biosorbent surface and the positively charged cationic dye [11]. In contrast, at pH < pHzpc the biosorbent surface is positively charged and electrostatic repulsion occurs between MB cations and A. platensis surface. At acidic pH, the H+ ions compete with MB cations for available binding sites onto A. platensis [3]. However, the remarkable qe at pH < pHzpc where the most of the binding sites are protonated, suggests that hydrophobic interactions also contributed to MB removal [26]. In addition, based on typical deprotonation constants for short chained carboxylic groups (4 < pKa < 6), the increased MB binding in the pH range of 4–6 may be also attributed to the deprotonation of carboxyl groups [21]. This was confirmed by the chemical modification of dried cells and the esterification of surface carboxyl groups, which resulted to the decrease of the biosorption capacity (see Biosorption mechanisms). The decrease of qe at pH > 8 is difficult to be explained. Similar result was observed at pH 9.5–11 for MB adsorption on cedar sawdust [27]. Some of the reasons for the biosorption decrease at high pH values might be the involvement of other adsorption mechanisms such as ion exchange or chelation, or the hydrolysis of the biosorbent surface which creates positively charged binding sites [27]. In this study, it was observed that the equilibrium pH (pHe) of the samples at initial pH 9 and 10 decreased by 0.85– 1.23 units, indicating that an exchange mechanism of H+ ions with MB cations occurred (Fig. 1b). However, other dye–dye interactions such as an increased formation of MB aggregates at higher pH, which are unable to enter into the pores of A. platensis, may be responsible for the decreased qe at pH 9 and 10 [28].

Fig. 2. (a) Effect of contact time on MB biosorption onto A. platensis at three different initial MB concentrations (biomass dosage = 0.5 g/L, pH 7.5, temperature = 298 K). Symbols and curves represent the experimental data and the fitted pseudosecond order kinetic model, respectively. (b) Pseudo-second order model linear plots for MB biosorption onto A. platensis biomass.

Biosorption kinetics

Fig. 1. (a) Plot of initial pH versus final pH for the determination of biomass pHzpc, and (b) the effect of initial pH on MB biosorption onto A. platensis (pHe = equilibrium pH).

Biosorption kinetic experiments were carried out at three initial MB concentrations and at temperature of 298 K. As shown in Fig. 2a, the biosorption of MB onto A. platensis was very rapid in the first 2–10 min for all studied concentrations. After the rapid adsorption during the initial stage, the biosorption increased at a slower rate with time and equilibrium was established within 60– 120 min for all initial MB concentrations. Equilibrium capacity did not change significantly up to 24 h (data not shown). The equilibrium time is in agreement with a previous work about MB biosorption by Spirulina sp. [18]. The pseudo-first order model could not describe the kinetic data, because the plot of log (qe qt) versus t (Eq. (3)) presented very low values for R2 ( 0.988. The applicability of this model suggests that the biosorption rate was controlled by chemisorption [29], involving exchange or sharing of electrons between the MB cations and functional groups of the biomass surface [30]. For the

674


Table 1 Kinetic and diffusion model parameters for MB biosorption onto A. platensis. Initial dye concentration (mg/L)

qe,exp (mg/g) Pseudo-first order model R2

25

50

29.48

54.94

0.355

0.241

100 82.95

t0:5 ¼

1 ðk2 qe Þ

(13)

0.337

Pseudo-second order model 27.40 qe,calc (mg/g) k2 (g/mg min) 0.0247 h (mg/g min) 18.52 1.479 t0.5 (min) R2 0.998 CFEF 3.46 x2 4.84

55.56 0.0134 41.32 1.344 0.998 18.49 29.36

80.65 0.0214 138.89 0.581 0.988 6.39 6.40

Intra-particle diffusion model: Whole time data 0.307 kid (mg/g min0.5) I (mg/g) 23.25 R2 0.583 CFEF 0.52 x2 0.54

0.197 59.11 0.269 0.39 0.39

1.220 67.66 0.517 3.88 3.70

Intra-particle diffusion model: Second linear section kid,2 (mg/g min0.5) 0.562 I (mg/g) 22.05 R2 0.646 CFEF 0.48 x2 0.50

1.024 54.59 0.944 0.15 0.15

2.866 58.94 0.961 0.94 0.91

pseudo-second order kinetics, the calculated qe values (qe,cal) agreed well with the experimental q values (qe,exp) (Table 1). However, the nonlinear analysis of the kinetic data for the initial MB concentration of 50 mg/L showed relative high CFEF and x2 values (Fig. 2a), which are due to an underestimation of the early time data (first 30 min) by the kinetic model [6]. The biosorption capacity (qe) at equilibrium, calculated from the pseudo-second order model, increased with increasing initial MB concentration (Table 1). However, the pseudo-second order rate constant (k2) decreased slightly when the initial MB concentration increased from 25 to 100 mg/L, but its values [0.0134–0.0247 g/(mg min)] demonstrated a same magnitude for all studied concentrations (Table 1). A decreasing value of k2 suggests that the biosorption equilibrium capacity was established slower at higher MB concentrations due to the limited quantity of binding sites at the biosorbent surface [25]. In addition, the nonlinear relationship between the rate constant values and initial MB concentrations suggest that various mechanisms involved in the biosorption process, such as ion exchange, chelation and physisorption [31]. The initial adsorption rate h (mg/g min) at 298 K was calculated from the pseudo-second order model parameters with the following equation [32]: h ¼ k2 q2e

The half adsorption time or half-life, t0.5 (min), expresses the time required for the biosorbent to remove the adsorbed amount of dye at equilibrium to its half, and is calculated from the pseudosecond order model parameters with the following equation [33]:

As shown in Table 1, the estimated values of t0.5 decreased from 1.479 to 0.581 min when the initial MB concentration increased, indicating a faster biosorption [33]. This parameter is used as a measure of adsorption rate and to understand the operating time of an adsorption system [33]. Fig. 3 shows the behavior of the intra-particle diffusion model of Weber–Morris at three initial MB concentrations and 298 K. This model was applied to the kinetic data in order to determine the biosorption process mechanism and the rate controlling step. As shown in Table 1, the values of R2 obtained from the linear regression plots of qt versus t0.5 for the whole time data of the sorption process, were low ( 0.963). Although the Langmuir and Freundlich isotherm models presented satisfactory and similar determination coefficients (R2 > 0.950 and 0.960, respectively), the Freundlich model could better describe the experimental data than the Langmuir model due to the lower CFEF and x2 values (Table 2). Thus, the good and similar agreement of the three applied isotherm models with the experimental data show that the MB sorption was a complex process, involving more than one mechanism [4]. Both the monolayer biosorption and surface heterogeneity of biosorbent affected the removal of MB from the solution [4], and no clear biosorption saturation was occurred in the studied range of MB concentration [34]. Table 2 Isotherm parameters values of MB biosorption onto A. platensis at different temperatures. Solution temperature (K) 298

308

318

89.56

82.18

65.70

312.50 117.42 0.0109 0.478–0.936 0.950 10.82 8.96

204.08 86.94 0.0126 0.442–0.927 0.989 4.02 3.24

80.65 59.31 0.0414 0.195–0.794 0.952 3.71 3.62

x2

99.75 4.766 1.319 0.967 2.86 2.89

82.55 3.512 1.291 0.981 1.50 1.50

64.95 5.003 1.641 0.960 1.42 1.59

Dubinin–Radushkevich qs (mol/g) BD (mol2/kJ2)

0.0048 6.05 109

0.0042 5.85 109

0.0017

E (kJ/mol) R2 CFEF

9.09 0.974 4.98 106

9.25 0.986 4.73 106

10.77 0.963

x2

5.42 106

4.46 106

5.40 106

qe,exp (mg/g) Langmuir qmax (mg/g) qe,cal (mg/g)a KL (L/mg) RL (range) R2 CFEF

x2

Freundlich qe,cal (mg/g)a KF ((mg/g) (L/mg)1/n) n R2 CFEF

4.31 109

R% 298 K R% 308 K

4.93 106

R% 318 K

Fig. 4. Effect of initial MB concentration on the percentage removal of MB and the biosorption capacity of A. platensis at different temperatures.

a

qe,cal corresponds to C0 = 100 mg/L.

676


Fig. 5. Linear plots of (a) Langmuir and (b) Freundlich isotherm model for the MB biosorption onto A. platensis at different temperatures.

The Langmuir model assumes a monolayer adsorption onto homogeneous surfaces with finite number of binding sites and no interaction between adsorbate molecule [1,4]. The constants qmax and KL were estimated from the intercept and slope of the linear plot of experimental data of 1/qe versus 1/Ce (Fig. 5a). The maximum monolayer adsorption capacity (qmax) decreased from 312.50 to 80.65 mg/g when the temperature increased from 298 to 318 K (Table 2). However, the Langmuir constant KL increased with the increasing temperature (Table 2), indicating a higher affinity (0.0414 L/mg) of A. platensis biomass for the MB

molecules at 318 K. The values of the dimensionless separation factor, RL, found to be less than unity and greater than zero (0 < RL < 1) at all initial MB concentrations and temperatures, confirming a favorable sorption process. If RL > 1 the adsorption is unfavorable. As shown in Fig. 6, the higher the initial MB concentration, the lower the RL value and the more favorable the MB biosorption. A comparison of the maximum monolayer adsorption capacity (qmax) for MB onto various adsorbents [25,26,35–38] and that obtained onto A. platensis in this work, shows that the cyanobacterium is an efficient biosorbent for the removal of MB from aqueous solutions. According to recent studies, S. platensis presented also a satisfactory biosorption capacity for anionic dyes [9,13,23,39]. The Freundlich model assumes a multilayer adsorption onto heterogeneous surfaces with energetically non-equivalent binding sites and interactions between adsorbent molecules [1]. The constants KF and n were evaluated from the intercept and slope of the linear plot of experimental data of ln (qe) versus ln (Ce) (Fig. 5b). The values of the dimensionless Freundlich constant n related to the adsorption intensity and surface heterogeneity, were higher than 1 and less than 10 (1 < n < 10) (see Table 2), indicating a favorable sorption of MB onto A. platensis biomass at all studied temperatures. No significant difference for n values was observed with respect to temperature. The parameter KF represents a relative measure of adsorption capacity and strength. When the equilibrium concentration Ce tends to be one, then KF reaches the value of qe [4]. As can be seen in Table 2, the values of KF increased slightly with the rising temperature from 298 to 318 K, but decreased between 298 and 308 K. It shows that the multilayer biosorption of MB was enhanced at higher solution temperature. To distinguish between physical and chemical sorption, the mean free energy E (kJ/mol) of MB biosorption was calculated by the following equation: 1 E ¼ pffiffiffiffiffiffiffiffiffiffiffi 2K DR

(14)

where KDR (mol2/kJ2) is the constant of Dubinin–Radushkevich isotherm. The parameter E is related to the mean free energy of sorption per molecule of sorbate, assuming that the sorbate is transferred to the biosorbent surface from infinite distance in the solution. Typical values of E for chemical sorption are in the range of 8–16 kJ/ mol, while E < 8 kJ/mol indicates physical sorption [24]. As shown in Table 2, the mean free energy E of MB biosorption onto A. platensis suggests a chemisorption mechanism, because its values are in the range of 8–16 kJ/mol at all studied temperatures. The increasing temperature caused a slight increase of E from 9.09 to 10.77 kJ/mol, indicating an enhancement of the chemisorption at higher temperatures. The biosorption mechanisms are further discussed in Biosorption mechanisms. Biosorption thermodynamics The thermodynamic behavior of MB biosorption onto A. platensis biomass was investigated estimating the thermodynamic parameters of Gibbs free energy change (DG ), enthalpy change (DH ) and entropy change (DS ). The values of these parameters were estimated using the following equations [35]:

Fig. 6. Relationship between initial MB concentration and dimensionless separation factor RL at different temperatures.

DG ¼ RTlnK c

(15)

DG ¼ DH T DS

(16)


DS R

DH

(17)

RT

where R is the universal gas constant [8.314 J/(mol K)], T the absolute solution temperature (K), and Kc (Cad,e/Ce) is the adsorption equilibrium constant, which is the ratio of the MB concentration adsorbed (Cad,e) to the MB concentration (Ce) in solution at equilibrium [38]. The negative values of DG indicates a spontaneous and favorable adsorption process at all studied temperatures and initial concentrations (see Table 3), suggesting that the system required no energy input from outside [23]. Similar thermodynamic behavior in respect to negative DG values has been found for S. platensis dry biomass as a biosorbent of anionic dyes [13,23,39]. For a given initial MB concentration in this work, no significant change of DG was observed with increasing temperature. However, the DG values decreased slightly as the initial MB concentration increased from 50 to 100 mg/L, indicating a more favorable adsorption of MB at lower dye concentration. The values of enthalpy change (DH ) and entropy change (DS ) can be calculated from the slope and intercept of the linear plot of ln Kc versus 1/T, based on the Eq. (17). As shown in Fig. 7, the determination coefficient (R2) of the plots was 0.939 and 0.940 for the two highest initial MB concentrations, respectively, indicating that the estimated values of DH and DS were confident. As can be seen in Table 3, the negative values of DH at all studied initial dye concentrations correspond to an exothermic nature of MB biosorption. Similar results for the cyanobacterium in respect to negative DH values obtained by other studies, which found an exothermic biosorption of anionic dyes [13,23,39] and phenol [17] onto S. platensis dry biomass. There are different results in the literature in respect to the exothermic or endothermic nature of MB adsorption onto various materials, based on the estimated DH values. An exothermic adsorption of MB was found onto cyclodextrin/silica hybrid adsorbent [38] and green algae Ulothrix sp. [31]. On the other hand, an endothermic adsorption of MB was found onto diatomite treated with sodium hydroxide [29], marble dust [19], montmorillonite clay [1], and acid treated kenaf fiber char [5]. The magnitude of enthalpy change can be used to classify the type of interaction between sorbent and sorbate. Values of DH < 30 kJ/ mol indicates a physical sorption such as hydrogen bonding [13]. Other mechanisms of physical sorption such as van der Waals forces usually presents DH values in the range 4–10 kJ/mol, hydrophobic bonds forces about 5 kJ/mol, coordination exchange about 40 kJ/mol and dipole bond forces 2–29 kJ/mol [13]. In contrast, DH > 80 kJ/ mol indicates chemical bond forces and a chemisorption process [13,17,20]. According to the DH values (