Sorption studies of manganese and cobalt from aqueous phase onto alginate beads and nano-graphite encapsulated alginate beads

G Model

JIEC-1885; No. of Pages 10 Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

Sorption studies of manganese and cobalt from aqueous phase onto alginate beads and nano-graphite encapsulated alginate beads Moonis Ali Khan a, Woosik Jung b, Oh-Hun Kwon b, Young Mee Jung c, Ki-Jung Paeng d, Seung-Yeon Cho b, Byong-Hun Jeon b,* a

Advanced Materials Research Chair, Chemistry Department, College of Science, King Saud University, Riyadh 11451, Saudi Arabia Department of Environmental Engineering, Yonsei University, 1 Yonseidae-gil, Wonju 220-710, Republic of Korea Department of Chemistry, Kangwon National University, 1 Kangwondaehak-gil, Chuncheon 200-701, Republic of Korea d Department of Chemistry, Yonsei University, 1 Yonsei dae-gil, Wonju 220-710, Republic of Korea b c

A R T I C L E I N F O

Article history: Received 15 December 2013 Accepted 23 January 2014 Available online xxx Keywords: Nano-carbon beads Thermal activation Sorption Desorption Regeneration

A B S T R A C T

Comparative sorption study of dissolved manganese and cobalt ions onto alginate beads (ABs) and thermally activated nano-carbon beads (NCBs) was performed. Acidic functionalities dominate over sorbent surface. Elemental analysis confirmed that divalent calcium replacement with heavy metal ions might be a possible sorption mechanism. Optimum metal uptake was observed at pH 8. Most of the metal ions (80–92%) were sorbed within 4 h, followed by a slower sorption stage. Mn(II) and Co(II) recovery was greater than 99% with 0.1 N HCl, and NCB could be repeatedly utilized for Mn(II) and Co(II) sorption with negligible loss in sorption capacity. ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

1. Introduction Divalent heavy metals such as manganese Mn(II) and cobalt Co(II) in trace quantities serve as essential elements and are required to promote biological activities in both human beings and plants. However, these heavy metals can have a deleterious impact if present in greater than trace amounts. Unlike organic compounds, heavy metals are non-degradable and persistent; therefore, it is essential to remove or minimize them to a permissible amount. Manganese is widely used in dry batteries, ceramics, ferroalloys, smelting, welding and electrical coils. Though it is an essential nutrient for human metabolism, it also acts as a potent neurotoxin at higher exposure levels [1]. Epidemiological studies revealed that the over exposure of human beings to Mn(II) can adversely affect physiological systems such as nervous, immune, and reproductive systems in adults [2]. During the perinatal period, both a deficiency and an excess of Mn(II) are problematic. Owing to the hazardous effects of Mn(II), it is essential to stringently assess and control its presence in environmental resources. The maximum permissible levels for Mn(II) as set by World Health Organization (WHO) are 0.1 [2] and 0.05 mg/L [3] in drinking and ground water, respectively, while the United States Environmental Protection

* Corresponding author. Tel.: +82 33 760 2446; fax: +82 33 760 2571. E-mail address: [email protected] (B.-H. Jeon).

Agency (USEPA) has set 0.05 mg/L as a maximum permissible level for Mn(II) in drinking water [4]. The applications of Co(II) in electronic, metallurgical and paint industries have been well defined [5]. Cobalt is essential for human health under permissible limits as it is one of the constituents of vitamin B12 [5]. However, Co(II) may cause bronchial infection, diarrhea, cardiomyopathy and dermatological disorders if present in amounts greater than the permissible limits [6], which are set at 0.05 and 0.002 mg/L for irrigation and drinking water, respectively [7]. Conventional treatment techniques such as chemical precipitation, reverse osmosis, electrolytic removal, ion-exchange and adsorption have been engineered for the abatement and remediation of heavy metals from water systems. Operational ease, economic feasibility and effectiveness even at lower (<10 mg/L) sorbate concentrations make adsorption a preferred method. Additionally, the related sludge production and disposal issues are minimal during adsorption. Adsorbents such as Shewanella putrefaciens [8], calcareous soils [9], brown seaweeds [10], natural clinoptilolite [11], activated carbon [12], vermiculite [13], poly (protoporphyrin-co-vinylpyridine) [14], hierarchical CaCO3-maltose meso/macroporous hybrid material [15], lemon peel [16], red macroalgae [17], EDTA-modified chitosan-silica hybrid materials [18], Rhytidiadelphus squarrosus [19], Garcinia mangostana L. fruit shells [20] and granular activated carbon [21] have been recently reported for Mn(II) and Co(II) removal from aqueous phase.

http://dx.doi.org/10.1016/j.jiec.2014.01.043 1226-086X/ß 2014 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 2

M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

High surface area to mass ratio and porosity, imparts excellent adsorbent characteristics to nano-materials. The advent of nanotechnology has provided immense scope and opportunities to fabricate the desired nano-sized adsorbents specifically for water decontamination. Nanosorbents such as Pleurotus ostreatus nano-particles [2], oxidized multi-walled carbon nanotubes [22], carbon nanotubes [23] and magnetic multi-walled carbon nanotube/iron oxide composites [7] have been reported for the removal of Mn(II) and Co(II) from aqueous media. Though these materials showed high metal removal rate, separation from the aqueous phase is an issue of concern due to their nano-scale size. Utility of nano-sorbents on an industrial scale could cause choking and fouling of industrial columns, limiting their application. We have encapsulated nanographite carbon in an alginate matrix (referred to as nano carbon beads or NCBs) and utilized them for cobalt [Co(II)] removal in a previous study. Slow sorption kinetics and low Co(II) sorption capacity were the major demerits observed [24]. In the present study, thermal activation of both ABs and NCBs was performed through drying, to overcome the problems observed in the previous study. A comparative study was carried out for the sorption of Mn(II) and Co(II) on both ABs and NCBs. Kinetics, isotherms and thermodynamics parameters were studied. The economic feasibility of the sorption process was confirmed by batch mode desorption and regeneration studies.

after 24 h of drying. Further increase in drying time to 48 h showed no significant effect on weight loss. Therefore, 24 h-dried ABs and NCBs were utilized for sorption studies. The diameter of the dried beads was approximately 1.0 mm. Activated beads were sealed in polybags and were stored in a desiccators to avoid exposure to dirt and moisture.

2. Experimental

The sorption performance of NCBs was studied in batch mode and compared with ABs using Mn(II) and Co(II) as model sorbates. The studies were performed in 50 mL capped glass tubes. The sorbate solutions (20 mL) of desired concentrations were equilibrated with 0.05 g of sorbents (ABs and NCBs) in a shaker at 100 rpm. The suspensions were filtered after attaining equilibrium and the residual metal concentration in the filtrate was determined using inductively coupled plasma (ICP, Thermo Jarrell Ash, USA) spectrometry. The sorption capacity at equilibrium (qe, mg/g) was determined using Eq. (1):

2.1. Chemicals and reagents The chemicals and reagents used were of analytical reagent grade or as specified. Sodium alginate was obtained from Sigma– Aldrich and was used without further purification. Stock solutions of Mn(II) and Co(II) (1000 mg/L) were prepared with reagent grade manganese sulfate (Mn(SO4)4H2O, Sigma–Aldrich) and cobalt chloride (CoCl26H2O, Sigma–Aldrich) in deionized (DI) water, respectively and were further diluted as desired. Calcium chloride (CaCl2), HCl and NaCl were obtained from Sigma–Aldrich.

2.3. Characterization of adsorbent For spectroscopic characterization of ABs and NCBs, the samples were well-dried and ground. Surface morphology of dry ABs and NCBs was assessed through variable pressure field emission scanning electron microscopy (VP-FE-SEM, Carl Zeiss, Germany). Energy dispersive X-ray analysis (EDX, Hitachi, SU-70) was employed to confirm Mn(II) and Co(II) sorption over AB and NCB surfaces. Fourier transform infrared (FT-IR) spectra were obtained using a Perkin-Elmer spectrometer (BUCKS-HP9 2FX, UK). The IR absorbance was recorded in the range 4000–400 cm1 using single reflection and a resolution of 4 cm1. ASAP2010 (Micro metrics, USA) was used to determine the pore area and porosity of ABs and NCBs. The active sites present on the sorbent surface were determined by Boehm’s acid-base titration experiments [26]. 2.4. Batch sorption/desorption studies

qe ¼ 2.2. Preparation of adsorbent Graphite nano-carbon (GNC) was prepared following a previously reported procedure [25]. In brief, a plastic electrolytic cell containing de-ionized (DI) water as an electrolyte was used for GNC preparation. Three graphite electrodes (two anodes and a cathode) were immersed in the electrolyte. A two-stage working process consisting of activating the anodes in the initial stage and generating GNC in the final stage was used. A graphite nano-carbon colloidal solution (pH 3) with nano-sized graphite carbon (3000 mg/L) was obtained through electrolysis for 15 min. Fifty grams of sodium alginate (5%, w/v) was solubilized in a GNC colloidal solution (1000 mL) under magnetic stirring for 12 h to obtain a homogeneous colloidal suspension. The suspension was then dropped into 1000 mL of 0.05 M CaCl2 solution using a 5.0 mL syringe through a 1.7 mm diameter needle to form NCBs. The beads were incubated for 24 h in a CaCl2 solution under ambient temperature conditions to complete the cross-linking reaction, and the NCBs were further rinsed several times with DI water to remove residual GNC particles and non-cross linked calcium ions from the surface. To compare the adsorption performance between non-impregnated and GNC-impregnated beads, pure ABs were prepared separately as a control. Wet NCBs and ABs were thermally activated at 323  5 K in oven by varying the drying time as thermal activation promotes hydrophobicity over the carbonaceous sorbent surface. Initially, wet ABs and NCBs weighed approximately 0.210 g. A drastic decrease in AB and NCB weight to 0.018 g (91.43%) and 0.015 g (92.86%), respectively, was observed

ðC 0  C e ÞV m

(1)

where C0 and Ce are the supernatant initial and equilibrium concentrations (mg/L),respectively, V is the volume of solution (L) and m is the mass of sorbent (g). The pH studies were conducted in a pH range of 2–8 under ambient temperature (298  2 K) conditions. The initial pH of the sorbate solution (C0 – 50 mg/L) was adjusted using 0.1 M HCl and 0.1 M NaOH solutions. Contact time studies were carried out at a 20 mg/L initial sorbate concentration under ambient temperature conditions, while the isotherm and thermodynamic studies were performed using a wide range of sorbate concentrations. For desorption studies, the sorbates (C0 – 25 mg/L) were initially sorbed over ABs and NCBs. At equilibrium, the sorbent was separated from the sorbate system and washed with DI water to remove the unsorbed metal traces from the surface. The sorbed metal ions were eluted using 0.1 N HCl, 0.1 N HNO3 and 0.1 N H2SO4. Regeneration studies were conducted for five consecutive cycles using 0.1 N HCl as an eluent. 3. Results and discussion 3.1. Characterization of sorbents Morphologically, ABs and NCBs were found to be spherical. Alginate beads appeared pale yellow, while NCBs were black in color due to the encapsulation of GNC. Scanning electron microscopic observations were obtained to characterize the porous structure and rough surface of ABs and NCBs. Microscopic images

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

3

Fig. 1. FT-IR spectra of ABs and NCBs before and after Mn(II) and Co(II) sorption.

of NCBs after GNC encapsulation revealed unevenness and shrinkage over the sorbent surface, resulting in pore creation (Figures S1a and b). Similar results were observed form our previous research, FE-SEM and TEM results confirmed that the graphite nano carbon (GNC) incorporated into alginate beads, resulted in decrease of porosity and an increase of surface area [24]. Surface areas measured using BET analysis were 2.368 and 3.171 m2/g for ABs and NCBs, respectively. The average pore diameters for ABs and NCBs were 10.1 and 10.6 nm, respectively. FT-IR spectra of ABs and NCBs before and after Mn(II) and Co(II) sorption are shown in Fig. 1. A broad and strong band characteristic of all natural polysaccharides (due to –OH stretching) was observed between 3000 and 3600 cm1 on both ABs and NCBs. Bands in the 2990–2945 cm1 region were due to the –CH stretching vibration in –CH and –CH2. Bands between 2373 and 2310 cm1 were due to CBBC stretching vibrations. Bands at 1610 and 1430 cm1 were assigned to C5 5C stretching vibrations in carboxylic and –CH2 groups, respectively. A band at 1398 cm1 was due to C–H deformation vibrations in alkanes. Broad and strong bands in the 1010–1184 cm1 region were due to C–O stretching vibrations in phenolic and carboxylic groups. Infra-red studies confirmed the presence of carboxylic, phenolic and carbonylic functionalities over ABs and NCBs. Shifting of bands after the sorption of Mn(II) and Co(II) indicated the active involvement of these functionalities in the sorption process and the replacement of calcium [Ca(II)] ions with divalent metal ions leading to a change in charge density, radius and atomic weight of the cation. Acid–base titration studies showed the dominance of acidic groups on both ABs and NCBs, favoring the sorption of

cations. An increase in carboxylic groups was observed on NCBs (Table S1a). Elemental analysis data confirmed the replacement of Ca(II) from ABs or NCBs by Co(II) or Mn(II) during the sorption process (Table S1b). 3.2. Effect of pH The effect of pH on Mn(II) and Co(II) sorption onto ABs and NCBs was determined with in a pH range of 2–8. Precipitation of Co(II) as Co(OH)2 occurs in a pH range of 7.3–12, and it is predominate at pH 8.3 [27]. In coal mine drainage, precipitation of Mn(II) as MnOx was observed when aqueous phase alkalinity increased due to the addition of sodium hydroxide. Thus, to avoid the impact of precipitation/coprecipitation mechanisms over physical sorption, the aforementioned pH range was selected along with experimental control [28]. At pH 2, minimum sorption of metals was observed on both the sorbents. The protonation of binding sites under highly acidic conditions results in a competition between H+ and cationic metal ions for occupancy of the binding sites, leading to low sorption. Increasing pH from 2 to 4 improved the metal sorption, which might be due to the decrease in the number of H+ ions competing with cationic metal ions to occupy the sorption sites. Further increases in pH from 4 to 8 showed a slow but constant increase in sorption (Fig. 2). Maximum metal ion uptake was observed at pH 8, which was beyond the pHPZC (6.8) of the NCBs [24]. At pH > pHPZC, the net negative charge density increases over the sorbent surface, promoting electrostatic interaction between metal ions and the sorbent surface as a metal binding mechanism. Higher Co(II) sorption compared to Mn(II) was observed on both

Fig. 2. Effect of pH on Co(II) and Mn(II) sorption onto ABs and NCBs. Conditions: Co – 50 mg/L; temp. – 298  2 K; sorbent dose – 0.05 g; equilibration time – 24 h.

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 4

M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

NCBs and ABs. Manganese and Co(II) showed a sorption of 20.92 and 26.68 mg/g, respectively, on NCBs at pH 8. Smaller ionic radius (A˚) and hydrolysis constant of Co(II) than Mn(II) probably resulted in higher Co(II) sorption capacity in pH range 3–8 [29]. Ions with small ionic radii are easily sorbed in the pores of the sorbent without size exclusion, and a high fraction of hydrolyzed ions under the same pH conditions facilitate a relatively rapid and stable sorption onto the sorbent surface [29]. The differences in sorption capacities between ABs and NCBs for Co(II) rather than Mn(II) appeared almost constant, showing approximately 6.8 and 4.3 mg/g in a pH range of 4–8, respectively. The sorption of both ions on NCBs was significantly affected between pH 2 and 3; however, the differences were not observed at pH 2. This indicates that the active sites of impregnated carbon for both ions are more sensitive to acidic pH conditions than those of alginate beads. The control experiment results revealed no precipitation/coprecipitation of both Mn(II) and Co(II) for an assigned pH range. This indicates that the removal of both Mn(II) and Co(II) occurred only due to the adsorption. In addition, these results were also confirmed by MINTEQA2 calculation. 3.3. Effects of contact time and sorption kinetics Single metal system time course studies were carried out to determine the contact time required for sorption of Mn(II) and Co(II) on AB and NCB. Rapid Mn(II) and Co(II) sorption on both AB and NCB during the initial 4 h contact time was observed (Fig. 3A). During this initial sorption phase 80–92% of metals sorption on AB and NCB was accomplished. The initial rapid phase was followed by a slower phase, and finally the equilibrium phase was reached. A similar equilibration time of 16 h was observed for both modeled sorbates, which suggests that GNC impregnation had little influence on the sorption rate of either ion. It is hinted that the equilibration time for Co(II) sorption was comparatively faster than our previous study [24] where wet ABs and NCBs were utilized for sorption. Drying process apparently removed water in bead, which would decrease weight of bead significantly. This showed that thermal activation of the beads significantly influences sorption kinetics. The higher sorption rate during the initial phase was due to the large number of strong affinity binding sites, resulting in an increased concentration gradient between the divalent metal ions in the solution and the sorbent surface. Once the high affinity binding sites were occupied, residual binding sites with lower affinity were occupied, leading to slow achievement of equilibrium. The equilibration time for sorption of Mn(II) and Co(II) along with experimental conditions are compared with previously reported studies in Table 1. Kinetic models were used to identify a relationship between the sorption rate and equilibrium time [30]. Kinetics data were interpreted by pseudo-first-order, pseudo-second-order kinetic and intra-particle diffusion models. To determine the accuracy of applied models, a Chi square (x2) analysis was also performed. The pseudo-first-order kinetics model, as proposed by Lagergren, is given as [31]: qt ¼ qe ð1  ek2 t Þ

(2)

where qt and qe are the amounts of heavy metal cations sorbed (mg/g) at equilibrium and at time t (h), respectively, and k1 (1/h) is the pseudo-first-order sorption rate constant. The pseudo-second-order kinetics model as proposed by Ho and Mc Kay is given as [32]: qt ¼

q2e k2 t 1 þ q2e k2 t

(3)

where k2 (g/mg h) is the pseudo-second-order sorption rate constant.

Chi square (x2) was used as a statistical error function during the study and is expressed as: x2 ¼

X ðqe;exp  qe;cal Þ2 qe;cal

(4)

where qe,exp and qe,cal are the experimental and calculated sorption capacities at equilibrium, respectively. The smallest x2 value represents the best fitted model. Experimental sorption capacity (qe,exp) was different from the theoretically calculated sorption capacity (qe,cal) derived from the pseudo-first-order kinetics model, while qe,cal derived from the pseudo-second-order kinetics model was very close to qe,exp for both Mn(II) and Co(II) sorption on ABs and NCBs (Table 2a). The regression coefficient (R2) for the pseudo-second-order model was comparatively higher than the pseudo-first-order model. The qe,exp coincided with the pseudo-second-order kinetics qe,cal (Fig. 3A). The x2 for the pseudo-second-order model was also comparatively lower than that of the pseudo-first-order model. These results implied that the sorption of Mn(II) and Co(II) onto ABs and NCBs adheres to pseudo second-order kinetics, suggesting chemisorption as the rate-controlling step [32]. Weber and Morris’ intra-particle diffusion model was also applied to the kinetics data. The model is an empirical functional relationship, common to most sorption processes. It is given as [33]: qt ¼ kdi t 1=2 þ C i

(5)

where kdi (mg/g h1/2) and Ci are the intra-particle rate constant and the intercept of stage i, respectively. The intercept gives an indication of the boundary layer thickness (i.e., the larger the intercept, the greater the boundary layer effect). According to Weber and Morris’ model, the following steps are involved during the sorption process: (1) bulk diffusion involving transport of sorbate molecules from the bulk solution to the boundary layer surrounding the sorbent; (2) film diffusion, i.e., sorbate diffusion from the boundary layer to the external sorbent surface; (3) pore or intra-particle diffusion, where sorbate diffusion proceeds from the surface to the sorbent internal pores; (4) sorbate uptake/binding by the active sites over the sorbent surface (via chemisorption, physisorption, ion-exchange, complexation, precipitation, or chelation). Intra-particle diffusion plots for Mn(II) and Co(II) sorption on ABs and NCBs are illustrated in Fig. 3B. Multi-linear plots with deviation of the straight line from the origin indicated involvement of other mechanisms along with intra-particle diffusion for the sorption process. The initial (external surface) sorption stage was rapid, as revealed by the steep slope, possibly due to high initial heavy metal ions concentration in the aqueous phase. The sorption gradually slowed in the second stage, during which time the sorbate had gone through internal diffusion, where it was transferred to the interior of the particle by diffusion of sorbate molecules through macropores, wider and smaller mesopores and micropores [34]. In this stage, intra-particle diffusion is the ratecontrolling step. Equilibrium was accomplished in the third stage, as demonstrated by the lines parallel to the contact time axis. At this stage, the intra-particle diffusion slows because of the saturation of the sorbent surface and low metal concentration left in the aqueous phase. The intra-particle diffusion constants follow an order kd1 > kd2 > kd3 (Table 2b), while the boundary layer follows the order C1 < C2 < C3. This is because the concentration of metal ions left in the aqueous phase gradually decreases. 3.4. Effects of concentration and sorption isotherms The sorption capacities of Mn(II) and Co(II) on AB and NCB as a function of initial sorbate concentration were evaluated. Increases

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

5

Fig. 3. (A) Kinetics plots for the sorption of Co(II) and Mn(II) on ABs (a and b) and NCBs (c and d). Conditions: Co – 20 mg/L; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 6  0.2. (B) Intra-particle diffusion plots.

in sorption were observed with an increase in initial divalent metal concentration on both ABs and NCBs (Fig. 4). High initial sorbate concentration possibly provides driving forces to overcome mass transfer resistance between solid/solution interfaces. Increases in initial Mn(II) concentration from 52.28 to 891.5 mg/L increased sorption on ABs and NCBs from 18.07 to 63.7 mg/g and 23 to 78 mg/g, respectively. The sorption of Co(II) on ABs and NCBs increased from 19.3 to 71.5 mg/g and 30.15 to 89.5 mg/g,

respectively, with an increase in its initial concentration from 62.83 to 1046.5 mg/L. Co(II) and Mn(II) have a greater sorption capacity on NCBs, it might be due to the impregnation of GNC provided available sorption sites by nano-sized carbonaceous materials [35]. The efficacy of the sorption process is determined by sorption isotherms. Sorption data were modeled by two-parameter (Langmuir and Freundlich) and three-parameter (Redlich–Peter-

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

6

Table 1 Comparison of equilibration time and maximum monolayer sorption capacity of various sorbents for the removal of Mn(II) and Co(II). Sorbent

Mn(II) ABs NCBs AC immobilized by tannic acid Chitin + Protein(DM-SC20) Ziziphus spina-christi seeds Oxalic acid modified maize husk Kaolinite MgO coated zeolite

Co(II) AB NCB Magnetic AB containing Cyanex 272R MWCNTs/iron oxide composites Alginate–chitosan hybrid gel bead Kaolinite Hydroxyapatite/chitosancomposite

Maximum monolayer sorption capacity, qm (mg/g)

Reference

Experimental conditions

Equilibration time, t (h)

& Co – 20 mg/L; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 5  0.2 & Contact time – 24 h; sorbent dose – 0.02 g; pH – 5  0.2 & Co – 20 mg/L; temp. – 298  2 K; sorbent dose – 0.072 g; pH – 5.2  0.2 & Contact time – 24 h; sorbent dose – 0.02 g; pH – 5  0.2 & Co – 2 mg/L; temp. – 298  2 K; sorbent dose – 2 g/L; pH – 5.4 & Contact time – 60 min; sorbent dose – 2 g/L; pH – 5.4 & Co – 10 mg/L; temp. – 298  2 K; sorbent dose – 5 g/L; pH – 8.6 & Contact time – 72 h; sorbent dose – 5 g/L; pH – 8.6 & Co – 20 mg/L; temp. – 298  2 K; sorbent dose – 0.75 g/L; pH – 4 & Contact time – 180 min; sorbent dose – 0.75 g/L; pH – 4 & Co – 300 mg/L; temp. – 298  2 K; sorbent dose – 0.8 g; pH – 4 & Contact time – 30 min; sorbent dose – 0.8 g/L; pH – 4 & Co – 4.46 mg/L; temp. – 298  2 K; sorbent dose – 10 g/L & Contact time – 2 h; sorbent dose – 10 g/L & Co – 100 mg/L; temp. – 298  2 K; sorbent dose – 2.5 g/L; pH – 6.0  0.2 & Contact time – 120 min; sorbent dose – 2.5 g/L; pH – 6.0  0.2

16

63.7

This study

16

78

This study

& Co – 20 mg/L; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 6  0.2. & Contact time – 24 h; sorbent dose – 0.02 g; pH – 5  0.2 & Co – 20 mg/L; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 6  0.2. & Contact time – 24 h; sorbent dose – 0.02 g; pH – 5  0.2 & Co – 0.88 mg/L; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 5  0.2 & Contact time – 48 h; sorbent dose – 0.072 g; pH – 7.5  0.2 & Co – 4.2 mg/L; temp. – 295  2 K; sorbent dose – 0.005 g; pH – 6.4  0.1 & Contact time – 24hr; sorbent dose – 0.005 g; pH – 6.4  0.1 & Co – 300 mg/L; temp. – 298; sorbent dose – 3 g; pH – 3.5 & Contact time – 3 h; sorbent dose – 0.5 g; pH – 3.5 & Co – 20 mg/L; temp. – 293.15; sorbent dose – 2 g/L; pH – 8.0 & Contact time – 24 h; sorbent dose – 2 g/L; pH – 8.5 & Co – 10 mg/L; temp. – 303; sorbent dose – 3 g/L; pH – 6.0 & Contact time – 40 min; sorbent dose – 3 g/L; pH – 6.0

1

1.73

[40]

72

5.437

[41]

3

172.413

[42]

0.5

9.004

[43]

2

0.446

[44]

2

30.85

[45]

16

71.5

This study

16

89.5

This study

48

32.65

[46]

24

10.60

[7]

3

3.18

[5]

24

51.32

[47]

10.63

[48]

0.67

Experimental conditions: & contact time; & isotherm studies.

son) sorption isotherms. The Langmuir model assumes formation of a monomolecular layer over the sorbent surface without interaction between the sorbed molecules. The model is expressed as [36]: qe ¼

qm K L C e 1 þ K LCe

(6)

where qm (mg/g) and KL (L/mg) are the Langmuir constants for the maximum solid phase loading on sorbent and energy constant related to heat of sorption, respectively. The essential characteristics of the Langmuir model can be expressed by a separation factor (RL), a dimensionless constant expressed as: RL ¼

1 1 þ K LC0

(7)

Table 2a Kinetics data for Mn(II) and Co(II) sorption on ABs and NCBs. Kinetics model

qe,exp (mg/g) Pseudo-first-order qe,cal (mg/g) K1 (1 h) R2

x

2

Pseudo-second-order qe,cal (mg/g) K2 (g/mg h) R2

x2

Mn(II)

where C0 (mg/L) is the highest aqueous phase initial solute concentration. The Freundlich model is an empirical relation that assumes the sorption to occur on heterogeneous surface sites. The application of this model also suggests an exponential decrease in sorption energy with saturation of sorption sites on the sorbent surface [37]. It is expressed as: 1=n

qe ¼ K F Ce

(8)

where KF (mg/g) (L/mg)1/n and n are the Freundlich constant and a heterogeneity factor, respectively. The Redlich–Peterson (R–P) isotherm is a three-parameter empirical relationship. It is a hybrid isotherm featuring both Langmuir and Freundlich isotherm models and is linearly dependent on concentration in the numerator and has an exponential function in the denominator to represent sorption equilibria over a wide concentration range. The R–P isotherm can be applied in homogeneous or heterogeneous systems due to its

Co(II)

ABs

NCBs

ABs

NCBs

6.394

7.391

6.899

7.932

6.369 0.232 0.814 1.907

7.376 0.259 0.732 3.619

6.862 0.235 0.815 2.296

7.908 0.241 0.723 4.054

6.392 2.102 0.992 0.089

7.387 3.299 0.991 0.021

6.887 1.992 0.989 0.079

7.931 2.113 0.985 0.046

Table 2b Intra-particle diffusion model for Mn(II) and Co(II) sorption on ABs and NCBs. Sorbent/ sorbate

Intra-particle diffusion constants of stage i (kdi, mg/g h1/2)

Thickness of boundary layer at stage i (Ci)

kd1

kd2

kd3

C1

C2

C3

Co(II) ABs NCBs

2.593 3.928

0.818 0.280

0.062 0.133

1.613 2.123

3.945 6.789

6.603 7.313

Mn(II) ABs NCBs

2.453 3.731

0.584 1.213

0.0062 0.0042

2.018 1.457

4.384 4.103

6.365 7.372

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

7

Fig. 4. Isotherm plots for the sorption of Co(II) and Mn(II) on ABs (a and b) and NCBs (c and d). Conditions: equilibration time – 24 h; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 5  0.2.

versatility [38] and is expressed as: qe ¼

AC e 1 þ BCeb

(9)

where A (mg/g) and B (mg/gg) are the R–P isotherm constants and b is the R–P isotherm exponent ranging between 0 and 1. The isotherm will convert to a Langmuir isotherm if b = 1. If BCeb  1, it will be reduced to the Freundlich isotherm. If BCeb  1, it will be reduced to Henry’s law [39]. Fig. 4 and Table 3 show isotherm plots and data for the sorption of Co(II) and Mn(II) on ABs and NCBs, respectively. Isotherm plots showed good agreement with the Freundlich model for Co(II) sorption experimental data (Fig. 4(a and c)), while the Langmuir model fitted well to Mn(II) sorption experimental data (Fig. 4(b and d)) on both ABs and NCBs. Higher regression coefficients (R2) (nearer to unity) and smaller x2 values also supported the aforementioned results. The maximum experimental sorption capacities (qm,exp) for Co(II) on ABs and NCBs were 71.5 and 89.5 mg/g, respectively, while the observed qm,exp values for Mn(II) on ABs and NCBs were 63.7 and 78 mg/g, respectively. The qm,exp values of the present study were compared with those of previously reported studies (Table 1). The values of RL in the present study ranged between 0 and 1, confirming favorable uptake of heavy metal ions on both ABs and NCBs. A heterogeneity factor (n) much greater than 1 indicates that sorption intensity is good over the entire concentration range studied. If n < 1, it indicates that sorption intensity is good at high sorbate concentrations but not as good at lower concentrations [49]. In the

present study, values of n were well above unity (1) for the entire concentration range, indicating that the sorption intensity is good over the entire sorbate concentrations range studied. The values of n were comparatively higher for Co(II) sorption on AB and NCB Table 3 Isotherm data for Mn(II) and Co(II) sorption on ABs and NCBs. Isotherm

qm,exp (mg/g) Langmuir qm,cal (mg/g) KL R2 RL

x2 Freundlich qm,cal (mg/g) KF n R2

x2 R–P qm,cal (mg/g) A B b R2

x2

Mn(II)

Co(II)

ABs

NCBs

ABs

NCBs

63.7

78

71.5

89.5

70.614 0.008 0.981 0.360 0.166

86.581 0.010 0.994 0.325 0.254

66.863 0.0119 0.823 0.279 1.048

71.310 2.394 1.980 0.940 1.213

85.455 4.265 2.237 0.908 1.204

72.895 10.900 3.257 0.999 0.013

90.440 4.542 2.481 0.999 0.039

79.200 0.987 0.036 0.825 0.821 1.654

85.914 0.972 0.064 0.723 0.950 1.209

66.063 0.199 1.525 0.036 0.997 42.180

105.005 0.456 0.091 0.792 0.998 34.639

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

75.780 0.001 0.455 0.718 137.618

G Model

JIEC-1885; No. of Pages 10 8

M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

indicating favorable Co(II) sorption. The values of BCeb , as obtained from the R-P model over the entire studied concentration range, were greater than unity, showing conversion of the model to the Freundlich model [42]. 3.5. Thermodynamic studies The physicochemical behavior of metal ions under environmental conditions can be influenced by temperature. Therefore, thermodynamics studies are essential in the design of a sorption system. The sorption of Mn(II) and Co(II) on both ABs and NCBs over the entire concentration range increased with increasing reaction temperature, confirming endothermic nature of sorption process. Thermodynamics parameters such as Gibbs (or standard) free energy change (DG8), standard entropy change (DS8) and standard enthalpy change (DH8) were evaluated to confirm the nature of metals sorption on sorbents. Gibbs free energy change (DG8, kJ/mol) is given as: 



DG ¼ DH  T DS



(10)

where T (K) is the reaction temperature, DH8 (kJ/mol) and DS8 (J/ mol K) are the standard enthalpy and entropy changes, respectively. The values of DH8 and DS8 were calculated from the slope and intercept of the Van’t Hoff plot (ln Kd vs. 1/T) using the following expression: ln K d ¼

DS  R



DH  RT

(11)

where Kd is the metal distribution coefficient on the sorbent surface and in the aqueous phase, and R (8.314 J/mol K) is the ideal gas constant.

Thermodynamic study data is listed in Table S2(a and b). The positive enthalpy change (DH8) suggested that the sorption of Mn(II) and Co(II) on both ABs and NCBs was endothermic. The hydration sheath of metal ions has to be destroyed before sorption to the solid sorbent surface. This process needs energy, and it is favored at high temperature. The provided energy exceeds the exothermicity of divalent metal ions to attach to the solid sorbent surface [7]. The Gibbs free energy (DG8) was negative at lower and positive at higher Mn(II) and Co(II) concentrations. The DG8 decreased with increasing reaction temperature as metal ions are readily hydrolyzed at higher temperature, indicating an increase in sorption efficiency with increasing temperature. The positive entropy change (DS8) implies disorder at the solid/solution interface. 3.6. Desorption and regeneration studies Desorption studies were carried out to elucidate the recovery of Mn(II) and Co(II) from ABs and NCBs using different eluents. Different acid solutions (HCl, HNO3 and H2SO4) and an alkaline NaOH solution of 0.1 N concentrations were used as eluents. Leaching of sorbed metal ions from sorbents was also tested using DI water as a control. Optimum metal recovery (>99%) for both sorbents (Fig. 5) was observed with a 0.1 N HCl solution followed by those in 0.1 N HNO3 and 0.1 N H2SO4. Deformation in the bead structure was observed on application of a 0.1 N NaOH solution as an eluent. To determine the economic feasibility of the sorbent, regeneration studies were carried out using 0.1 N HCl as an eluent. Regeneration studies showed that the sorption of Mn(II) for five consecutive cycles varied between 47.37% and 46.04% on ABs and 54.76% and 53.16% on NCBs (Fig. 6a and b). The sorption of Co(II) on

Fig. 5. Desorption of Mn(II) from ABs (a), NCBs (b); and Co(II) from AB (c) and NCB (d) using different eluents. Conditions: Initial concentration – 25 mg/L; contact time – 24 h; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 5  0.7.

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

9

Fig. 6. Regeneration studies of Mn(II) on ABs (a), and NCBs (b); and Co(II) on ABs (c), and NCBs (d). Conditions: Initial concentration – 50 mg/L; contact time – 24 h; temp. – 298  2 K; sorbent dose – 0.05 g; pH – 5  0.4.

ABs and NCBs for five consecutive cycles varied between 52.78% and 51.54%, and 59.76% and 54.16%, respectively (Fig. 6c and d). After five consecutive regeneration cycles, the recovery of Mn(II) for ABs and NCBs reduced from 99.29% and 99.16% to 96.87% and 98.62%, respectively. The observed loss in Mn(II) recovery for ABs and NCBs after five consecutive cycles was 2.43% and 0.54%, respectively, while Co(II) recovery from ABs was reduced from 99.16% to 96.7% after five consecutive cycles with a loss of 2.30% in recovery; for NCBs recovery was reduced from 99.16% to 96.62% with a loss of 2.54% in recovery.

HCl. Regeneration studies showed Co(II) sorption loss of 2.35% and 9.37%, and Mn(II) sorption loss of 2.81% and 2.92% on ABs and NCBs, respectively, after five consecutive cycles. Our findings reveal that ABs and NCBs have potential for removing Mn(II) and Co(II) ions from aqueous solutions. Further study is necessary to investigate functionalization of NCB to enhance its sorption performance and selectivity using both batch and column experiments.

4. Conclusions and future research

This work was supported by the Global Research Laboratory Project (2010-00248), the Eco-Innovation project (Global-Top project) of the Korea Ministry of the Environment, and the Small & Medium Business Administration (SMBA) under grant C0103527 through Academic-Industrial Common Technology Development Project.

A comparative batch sorption, desorption, and regeneration study on alginate beads (ABs) and nano-carbon beads (NCBs) was performed using divalent manganese and cobalt ions as model sorbates in a single metal system. Thermal activation was a major contributor to improved sorption performance. Spectroscopic analysis showed band shifting after Mn(II) and Co(II) sorption without the disappearance or appearance of new bands, confirming that the sorption occurred by replacement of Ca(II) with Mn(II) and Co(II). Sorption capacities for Mn(II) and Co(II) were found to be dependent on the pH of the solution and the maximum capacities were obtained at pH 8 with a contact time of 16 h, in addition 80–92% of the sorption accomplished in the initial 4 h for immobilized ABs and NCBs. Kinetics studies showed the applicability of a pseudo-second-order model, while multi-step sorption was confirmed by Weber and Morris’ model. Isotherm studies indicated the applicability of a Langmuir model for Mn(II) sorption, while sorption of Co(II) followed the Freundlich model. The sorption process was endothermic and spontaneous in nature. Desorption studies concluded optimum metal recovery with 0.1 N

Acknowledgement

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jiec.2014.01.043. References [1] M. Arora, A. Bradman, C. Austin, M. Vedar, N. Holland, B. Eskenazi, D.R. Smith, Environ. Sci. Technol. 46 (2012) 5118. [2] L. Ma, Y. Peng, B. Wu, D. Lei, H. Xu, Chem. Eng. J. 225 (2013) 59. [3] WHO, Manganese in drinking water: Background document for development of WHO guidelines for drinking water quality report, 2004. [4] US EPA, Effluent guidelines program plan, 2008, http://www.epa.gov/guide/ 304m/2008. [5] T. Gotoh, K. Matsushima, K.I. Kikuchi, Chemosphere 55 (2004) 135. [6] J. Snukiskis, D. Kauspediene, Colloids Surf. A 253 (2005) 27.

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043

G Model

JIEC-1885; No. of Pages 10 10

M.A. Khan et al. / Journal of Industrial and Engineering Chemistry xxx (2014) xxx–xxx

[7] Q. Wang, J. Li, C. Chen, X. Ren, J. Hu, X. Wang, Chem. Eng. J. 174 (2011) 126. [8] N. Chubar, T. Visser, C. Avramut, H. Waard, Geochim. Cosmochim. Acta 100 (2013) 232. [9] S. Moharami, M. Jalali, Environ. Earth Sci. 68 (2013) 847. [10] S.H. Lee, C.H. Park, Biotechnol. Bioprocess Eng. 17 (2012) 853. [11] M. Jovanovic, N. Rajic, B. Obradovic, J. Hazard. Mater. 223–224 (2012) 57. [12] J. Lach, E. Okoniewska, E. Neczaj, M. Kacprzak, Desalination 206 (2007) 259. [13] M.G. Fonseca, M.M. Oliveira, L.N.H. Arakaki, J. Hazard. Mater. B137 (2006) 288. [14] T.F. Oliveira, E.S. Ribeiro, M.G. Segatelli, C.R.T. Tarley, Chem. Eng. J. 221 (2013) 275. [15] X. Ma, L. Li, L. Yang, C. Su, K. Wang, S. Yuan, J. Zhou, J. Hazard. Mater. 209–210 (2012) 467. [16] A. Batnagar, A.K. Minocha, M. Sillanpaa, Biochem. Eng. J. 48 (2010) 181. [17] W.M. Ibrahim, J. Hazard. Mater. 192 (2011) 1827. [18] E. Repo, J.K. Warchol, A. Bhatnagar, M. Sillanpaa, J. Colloid Interface Sci. 358 (2011) 261. [19] J. Maresova, M. Pipiska, M. Rozloznik, M. Hornik, Desalination 266 (2011) 134. [20] R. Zein, R. Suhaili, F. Earnestly, A. Indrawati, E. Munaf, J. Hazard. Mater. 181 (2010) 52. [21] A.H. Sulaymona, B.A. Abidb, J.A. Al-Najar, Chem. Eng. J. 155 (2009) 647. [22] P. Ganesan, R. Kamaraj, G. Sozhan, S. Vasudevan, Environ. Sci. Pollut. Res. 20 (2013) 987. [23] A. Stafiej, K. Pyrzynska, Sep. Purif. Technol. 58 (2007) 49. [24] M.A. Khan, E. Gee, J. Choi, M. Kumar, W. Jung, T.C. Timmes, H.C. Kim, B.H. Jeon, Chem. Eng. Commun. 201 (2014) 403. [25] S.H. Kim, K.H. Lee, D.H. Kim, J.P. Park, J.H. Kim, S.Y. Cho, Indoor Built Environ. 21 (2012) 837. [26] S.L. Goertzen, K.D. Theriault, A.M. Oickle, A.C. Tarasuk, H.A. Andreas, Carbon 48 (2010) 1252. [27] F. Granados-Correa, S. Bulbulian, Water Air Soil Pollut. 223 (2012) 4089.

[28] H. Tan, G. Zhang, P.J. Heaney, S.M. Webb, W.D. Burgos, Appl. Geochem. 25 (2010) 389. [29] O. Yavuz, Y. Altunkaynak, F. Guzel, Water Res. 37 (2003) 948. [30] W. Cheah, S. Hosseini, M.A. Khan, T.G. Chuah, T.S.Y. Choong, Chem. Eng. J. 215–216 (2013) 747. [31] S. Lagergren, Handlingar Band 24 (1898) 1. [32] Y.S. Ho, G. Mckay, Water Res. 34 (2000) 735. [33] W.J. Weber Jr., J.C. Morris, J. Sanitary Eng. Div. Am. Soc. Chem. Eng. 89 (1963) 31. [34] Y.P. Teoh, M.A. Khan, T.S.Y. Choong, Chem. Eng. J. 217 (2013) 248. [35] K. Pyrzyn´ska, M. Bystrzejewski, Colloids Surf. A 362 (2011) 102. [36] I. Langmuir, J. Am. Chem. Soc. 40 (9) (1918) 1361. [37] M.A. Khan, M. Ngabura, T.S.Y. Choong, H. Masood, L.A. Chuah, Bioresour. Technol. 103 (2012) 35. [38] O. Redlich, D.L. Peterson, J. Phys. Chem. 63 (1959) 1024. [39] W.S. Wan Ngah, L.C. Teong, R.H. Toh, M.A.K.M. Hanafiah, Chem. Eng. J. 209 (2012) 46. [40] A. Ucer, A. Uyanik, S.F. Aygun, Sep. Purif. Technol. 47 (2006) 113. [41] M.A. Robinson-Lora, R.A. Brennan, Chem. Eng. J. 162 (2010) 565. [42] A. Omri, M. Benzina, Alexandria Eng. J. 51 (2012) 343. [43] A.I. Adeogun, M.A. Idowu, A.E. Ofudje, S.O. Kareem, S.A. Ahmed, Appl. Water Sci. 3 (2013) 167. [44] O. Yavuza, Y. Altunkaynak, F. Guzelc, Water Res. 37 (2003) 948. [45] S.R. Taffarel, J. Rubio, Miner. Eng. 23 (2010) 1131. [46] A.F. Ngomsik, A. Beea, J.M. Siauguea, D. Talbot, V. Cabuila, G. Coted, J. Hazard. Mater. 166 (2009) 1043. [47] K. Li, J. Hu, Z. Liu, L. Chen, Y. Dong, J. Radioanal. Nucl. Chem. 295 (2013) 2221. [48] N. Gupta, A.K. Kushwaha, M.C. Chattopadhyaya, J. Taiwan Inst. Chem. Eng. 43 (2012) 125. [49] G. McKay, Water Res. 14 (1980) 1.

Please cite this article in press as: M.A. Khan, et al., J. Ind. Eng. Chem. (2014), http://dx.doi.org/10.1016/j.jiec.2014.01.043