Preparation and characterization of magnetic

Industrial Crops & Products 105 (2017) 93–103

Contents lists available at ScienceDirect

Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop

Research Paper

Preparation and characterization of magnetic biosorbent based on oil palm empty fruit bunch fibers, cellulose and Ceiba pentandra for heavy metal ions removal

MARK

⁎

S. Daneshfozouna, M.A. Abdullahb, , B. Abdullaha a b

Department of Chemical Engineering, Universiti Teknologi PETRONAS, 32610 Seri Iskandar, Perak, Malaysia Institute of Marine Biotechnology, Universiti Malaysia Terengganu, 21030 Kuala Terengganu, Terengganu, Malaysia

A R T I C L E I N F O

A B S T R A C T

Keywords: Oil palm empty fruit bunches Cellulose Ceiba pentandra Magnetic biosorbent Heavy metal ion removal

This study prepared, characterized and developed agro-based magnetic biosorbents (AMBs) from Ceiba pentandra (RKF), oil palm empty fruit bunches (EFB) and celluloses (CEL) extracted from EFB, using a novel, simple and cost-effective preparation technique for the removal of Pb(II), Cu(II), Zn(II), Mn(II) and Ni(II) ions from aqueous solutions. There has been no report on the methods to prepare and the use of magnetic biosorbent based on these biomaterials. The morphological, chemical and magnetic characterization suggested successful preparation of AMBs with good dispersion of magnetic nanoparticles on the surface of the base materials with clear magnetic properties. Optimum sorption was achieved between pH 5–7, and increase in initial ion concentration and solution temperature resulted in increased ion uptake. AMBs regeneration was successfully performed for 5 adsorption/desorption cycles. The magnetic biosorbent based on kapok showed the best Pb(II) removal efficiency of 99.4% and 49 mg/g adsorption capacity compared to 98.2% for cellulose and 97.7% for EFB. The magnetic biosorbents exhibited 10.3% higher removal efficiency than the raw sorbents.

1. Introduction Heavy metal ions are considered as environmental health hazards, as they are placed in the top 10 in the directory of “Agency for Toxic Substances and Disease Registry Priority List of Hazardous Substances”, based on the poisonousness of the material and its potential exposure from contaminated air, water and soil. A number of international agencies that include Centre for Disease Control (CDC) (Yantasee et al., 2007), World Health Organization (WHO) (Chen et al., 2013; Sharma et al., 2012), Food and Agricultural Organization (FAO) and International Agency for Research on Cancer (IARC) are dealing with the hazardous impacts of heavy metal exposure. This has made detection and bio-monitoring crucial components of environmental pollution, control and remediation strategies (Aragay et al., 2011). Some heavy metals are essential for many biological activities at the micronutrient level, but higher concentrations have the ability to produce a range of toxic effects. In actual fact, lead, cadmium, chromium, arsenic, and mercury are even considered as toxic compounds at low concentrations (Bagal-Kestwal et al., 2008). Although harmful effects of these substances have been known for a long time, indiscriminate disposal of heavy metals are still continuing and even

⁎

Corresponding author. E-mail addresses: [email protected], [email protected] (M.A. Abdullah).

http://dx.doi.org/10.1016/j.indcrop.2017.05.011 Received 10 November 2016; Received in revised form 3 May 2017; Accepted 7 May 2017 Available online 13 May 2017 0926-6690/ © 2017 Elsevier B.V. All rights reserved.

increasing, especially in developing countries. Among various contaminants present in the ground and surface water, inorganic heavy metals are the most difficult to remove due to their small ionic size, complicated state of presence, low concentration in high volume and competition with other non-poisonous inorganic kinds (Bailey et al., 1999). Adsorption is among the most common methods for heavy metal ions removal due to its simplicity, feasibility and effectiveness. It is also widely utilized for both inorganic and organic material removal (Ali, 2012; Ali et al., 2012; Faust and Aly, 2013). In most cases, highly porous materials such as adsorbents can provide sufficient sorption surface area. The major bottleneck, however, is the intraparticle diffusion which causes reduced available space, leading to a decreased adsorption capacity. As the adsorption property of an adsorbent is highly affected by its intrinsic properties, modification or functionalization will significantly change the sorbent surface area, pore size distribution and surface functional groups. It is therefore necessary to develop biocompatible materials with large surface area, active surface sites and low intraparticle diffusion rate (Mahdavian and Mirrahimi, 2010). Adsorption via magnetic biosorbents (MBSs) has attracted much attention and reported as effective and applicable for field application


S. Daneshfozoun et al.

rials can resolve the continued-use issues, create high surface area-tovolume ratio and increase adsorption capacity and efficiency (Feng et al., 2010). Surface functionalized magnetic nanoparticles (MNPs) and the composite materials can further improve the costly synthesis of MNPs for wider application. The composite with core–shell nanostructures are more effective since the shell prevents the core from particle–particle combination and increases the dispersion stability of nanostructures in the suspension medium (Gómez-Pastora et al., 2014). The objectives of this study were to develop, prepare and characterize the magnetic biosorbent based on OPEFB fibers, celluloses extracted from OPEFB and kapok fibers. To the best of our knowledge, there has yet to be any report on heavy metal ion removal using magnetic biosorbents based on these agro-materials. The OPEFB magnetic biosorbent was studied for Pb(II), Cu(II), Zn(II), Ni(II) and Mn(II) ions removal from aqueous solution. The Pb(II) ion removal efficiency was compared with Cellulose and kapok. The parameters optimized included the contact time, pH, effects of initial ion concentration and temperature and the sorbent reusability.

due to its simple operation (Nalbandian et al., 2016). MBSs can be applied to adsorb pollutants from aqueous or effluents, which can later be separated from the medium by using a magnetic field for possible reuse for several cycles. MBSs can be designed as nano-sorbent materials for water pollution removal such as heavy metals and dyes. The advantages include a high number of active surface sites, large surface area, low intraparticle diffusion rate, high adsorption capacities and most importantly reusability, which make their utilization for water treatment more economical (Nassar, 2010; Tan et al., 2012). In practical engineering applications, having consistent supply of raw materials at large industrial scale is of paramount importance. As environmental protection is of increasing global concern, effort to develop greener remediation technique based on biosorption has gained traction. Biosorption involves the property of certain biomolecules to bind and concentrate selected ions or other molecules from aqueous solutions (Volesky, 2007). Agricultural-based biosorbents are therefore attractive for economics reason and ease of applications, as these are mainly made up of cellulose, hemicellulose and lignin with high contents of hydroxyl groups and admixture of other functional groups like carboxyl, sulfhydryl carboxyl, acetamido, phenolic, structural polysaccharides, amino groups, ester and alcohols that could all enhance heavy metals ion adsorption (Okoro and Okoro, 2011). These functional groups substitute hydrogen ions for metal ions or contribute an electron pair to form complexes with the metal ions in the solutions. These binding groups make agricultural lignocellulosic wastes a promising source of adsorbent materials to remove heavy metal ions from water and wastewater (Jiménez-Cedillo et al., 2013). Natural materials such as water bamboo husk (Asberry et al., 2014), Eucalyptus pulp (Squissato et al., 2017), phthalate-functionalized sugarcane bagasse (do Carmo Ramos et al., 2016), oil palm empty fruit bunches (Daneshfozoun et al., 2014a, 2016), cellulose (Daneshfozoun et al., 2014b), Ceiba pentandra (Afzaal et al., 2014; Abdullah et al., 2015), chitosan (Tran et al., 2010), hydroxyapatite (Feng et al., 2010) and moss peat (Bulgariu et al., 2008), are among biosorbents developed for heavy metal ion sorption. Conversion of agro-wastes into valuable products such as composite materials (Abdullah et al., 2016) or as adsorbent to remove inorganic pollutant can resolve both environmental problems: 1) reducing or recycling the wastes and 2) remediating the environment. Oil palm (Elaeis guineensis Jacq), a kind of monoecious plant, has been a cash crop in Malaysia since 1917. Palm oil has become a major contributor to the total oil and fat production in the world − increasing from 13% in 1990 to 28% in 2011 (MPOC, Malaysian Palm Oil Council 2012). Each plant can produce approximately 150 kg of fresh fruit bunches (FFB) per year and the weight of the FFB may vary from 10 to 40 kg, depending on the number of compactly bound fruitlets in the bunch (Yusoff, 2006). After the fruit has been detached from the FFB for oil extraction, the rest is called an oil palm empty fruit bunch (OPEFB). A total of 77.2 million tonnes of biomass are generated from oil palm processes, out of which, the 19.8 million tonnes on wet basis or 6.93 million tonnes on a dry basis, are OPEFB (Foo et al., 2011), making them the most abundant wastes from palm oil industries. Ceiba pentandra (L.) Gaertn. or Kapok, from the Bombaceae family, is another agro-fibers with a high potential to be developed as sorbent materials (Abdullah et al., 2010). A mature Kapok tree bears hundreds of pods that are up to 15 cm long and filled with fibrous seeds. In contrary to cotton, which is lignified and not attached to the seed grains, kapok fiber is made up of single-celled plant hairs (Zheng et al., 2015). Kapok fibers are light brown/yellowish silky fibers which are moisture resistant, quick-drying, and buoyant, and have found applications for oil sorption (Abdullah et al., 2010) and palm oil mill effluent (POME) treatment (Afzaal et al., 2014; Abdullah et al., 2015). Natural biopolymers such as oil palm fibers, kapok and cellulose provide high capacity and selectivity for environmental remediation applications due to their different functional groups representation in the structure (Gómez-Pastora et al., 2014). The magnetic lignocellulose-based mate-

2. Materials and methods 2.1. Raw materials and chemicals The raw OPEFB fibers used in this study were obtained from the FELCRA Nasaruddin Palm Oil Mill, Bota, Perak, Malaysia. The purely extracted cellulose was obtained from OPEFB fibers as reported earlier (Abdullah et al., 2016; Nazir et al., 2013). Raw kapok fiber was collected from Telok Belanja Village in Dungun, Terengganu, Malaysia. Fe2O3 nano-particles (5–50 nm sizes) were purchased from Merck (USA). Stock solutions of Pb(II), Cu(II), Zn(II), Ni(II) and Mn(II) were prepared from the required amount of Pb(NO3)2, CuCl2, ZnSO4, NiCl2 and MnCl2 (Merck) and dissolved in distilled water at room temperature. Other chemical reagents such as NaOH and HCl were all of analytical grade. 2.2. Preparation of magnetic biosorbent The raw OPEFB and cellulose fibers were homogenized to the size of 0.2 mm by bench top ring mill (Rocklabs LTD, New Zealand). The finesized OPEFB samples were obtained by two cycle grinding of the shredded OPEFB using a grinder and a hammer mill (Janke & Kunkel). Subsequently, the ground fibers were subjected to a Planetary Mono Mill (FRITSCH GmbH) to finally achieve the size ranging from 0.005–0.02 mm. In the case of a fine-sized raw kapok fiber (RKF) preparation, the fibers were separated out from the seeds, and the visible dust particles were removed manually. The fibers were dried in an oven at 70 °C until the moisture content was reduced to less than 1%. The dried fibers were then cut into small part, and every 3 g of the fibers were milled for 30 min at the speed of 450 rpm, in order to obtain an average size of 0.1–0.002 mm using the Planetary Mono Mill (FRITSCH GmbH). A mixture of 50 mg of Fe2O3 nanoparticles and 100 mL of fine-sized base-materials (5 mg/ml) in deionized water was sonicated using an Ultrasonic Homogenizer machine (150VT, biologist, USA) for 5 min and then shaken on an orbital shaker for another 10 min before it was sonicated again. The mixture was subjected to this cycle for 3 times and placed on a permanent magnet (4000 G) to recover the magnetic sorbents. The recovered MBSs were washed with deionized water and stored at room temperature. 2.3. Characterization 2.3.1. Morphology Morphology, shape and size of the biosorbent were analyzed using FESEM (VP- FESEM, Carl Zeiss, Supra 55VP). An Energy-dispersive Xray (EDX) joined to the FESEM was used for elemental analysis. As the 94



Fig. 1. FESEM images of a) OPEFB; b) Fe2O3@OPEFB (5Kx magnification); and c) EDX of Fe2O3@OPEFB.

electron beam interpenetrated the sample, the created X-ray was accumulated, detected and analyzed by Silicon Drift Detectors (SDD) to record the X-ray map. Further structure study on a nano-scale was done using a TEM (Carl Zeiss, LIBERA 200 FE). The samples were suspended in isopropyl alcohol (IPA) and sonicated for 1 h with 8 min intervals. A drop of the solution was placed on a carbon tap, dried, and gold coated using the sputtering method and used for structural and morphological investigation

the uniform magnetic field can magnetize the sample. A stronger uniform film creates larger magnetization. Around the sample, a magnetic stray field is created by the magnetic dipole model. When the sample is vibrating, the magnetic stray field changes with time, sensed by pick-up coils. An electric field is induced in the pick-up coils with a current that is proportional to the magnetization of the sample. With the measurement of this induced current, and the assistance of a monitoring and controlling software, the magnetic properties of the sample can be detected.

2.3.2. Chemical characterization Functional groups of raw and magnetic biosorbents were investigated using FTIR (Spectrum One, Perkin Elmer). A mixture of the sample and KBr at 1:1 ratio was pressed into a disc shape, which was then placed in the FTIR for scanning for 50 times at 4 cm−1 resolution transmission between 4000 and 450 cm−1 wavelength to produce the spectra. The structure of the adsorbents was characterized using the X-ray diffraction (D8-Advance Bruker-AXS). XRD is a powerful and nondestructive tool for the analysis of structure phases and material characteristics. The samples were analyzed at 2θ scan at a wide-angle range of 2–80°, with a step width of 2° per second at 25 °C. The thermal stability of raw and magnetic biosorbents was examined by using TGA Instruments (TGA Q500, USA). The samples were heated from room temperature up to 800 °C with a heating rate of 30 °C/min and a nitrogen flow of 100 mL/min.

2.4. Batch heavy metal ion adsorption by magnetic biosorbent Batch adsorption studies were performed at different temperatures (298–338 K) where 0.5 g of the sorbents was added separately to 50 mL of each metal ion stock solutions at different concentration (100–1000 ppm). The mixture was continuously shaken at 150 rpm for different pre-treated time and pH. The sorbents were then filtered out or removed by using a magnet and syringe filter (Model Whattman 0.45 pm,UK). Free metal ions were determined by using an Atomic Absorption Spectroscopy (AAS, Hitachi Z-5000) for analyses of the solution before and after adsorption. Each experiment was performed in triplicate and the average values were reported. The effect of contact time on adsorption capacity and heavy metal ions removal efficiency by magnetic biosorbents at 0.5 g/50 mL dosage was investigated for 5–150 mins at pH 5–5.5, and 20–22 °C. The effect of pH was investigated at pH 2–10. The pH was adjusted using 0.1 M NaOH or/and 0.1 M HCl acid, and measured with a pH meter (Mettles Toledo, Ross FE 20, USA). The initial adsorbate concentration was fixed at 500 mg/L with an adsorbent dosage of 0.5 g/50 mL at room temperature and the optimum contact time established previously. The effect of metal ion initial concentrations was tested from 100 to 1000 mg/L at an optimum pH and contact time as established before.

2.3.3. Magnetic properties The magnetic properties of each magnetic nanosorbents were analyzed by using a vibrating sample magnetometer, VSM (LakeShore 7307, USA) at room temperature. The sample was placed in a uniform magnetic field to be magnetized. In the case of magnetic material, the alignment of the magnetic domains or individual magnetic spin with 95



Fig. 2. TEM images of (a,c) raw OPEFB; and (b,d) Fe2O3@EFB biosorbents.

experiments, washed with deionized water and reused for the next adsorption cycle. This adsorption–desorption experiments were repeated for five cycles.

During the experiment, different temperatures within the range of 25–75 °C with the step of 10 °C was applied, using the temperature control system of a water bath shaker (HaakeWia Model, Japan). The adsorption capacity and removal efficiency (μ) of adsorbents was calculated as follows:

(C − Ce ) × V qe = i W μ=

(Ci − Ce ) × 100 Ci

3. Results and discussion

(1)

3.1. Characterization 3.1.1. Morphology Fig. 1(a and b) suggests that the raw OPEFB had similar morphological characteristics as reported earlier (Nazir et al., 2013; Abdullah et al., 2016). The OPEFB magnetic biosorbent (Fe2O3@EFB) showed that the aggregated magnetic nanoparticles were seen as bright spots within the fibers. The EDX plot (Fig. 1c) confirmed that there were no residuals in the final product and that the ratio of elements was in good agreement with the presence of Fe2O3 in magnetic biosorbent. The TEM images of Fe2O3@EFB proved that the aggregated nature of the hexagonal-shaped nanoparticles on the OPEFB base was with good stability and homogeneity (Fig. 2). The agglomeration of the nanoparticles can be clearly observed in the magnetic biosorbent of cellulose (Fe2O3@ CEL) (Fig. 3b) and kapok fiber (Fig. 3d).

(2)

where Ci and Ce are the initial and equilibrium metal ion concentration (mg/L), respectively; V is the volume of metal ion solution (L); and W is the weight of the biosorbent (g). 2.5. Reusability Desorption of metal ions from sorbents was studied using a 0.1 M HCl solution. Different sorbents were first equilibrated with metal ions in a solution at an optimum condition. Then, 0.5 g of metal ion-loaded sorbents were added to 50 mL of the desorption solution. The mixtures were shaken on an orbital shaker at 140 rpm and 25 °C for 1 h and filtered. The samples were removed from the solution in order to analyze the amount of desorbed metal ions in the solution. For reusability test, sorbents were separated after regeneration

3.1.2. Chemical The FTIR spectra in Fig. 4a shows the transmittance bands of OPEFB 96



Fig. 3. TEM images of a) Cellulose; b)Fe2O3@ CEL; c) RKF; and d) Fe2O3@RKF biosorbents.

and Fe2O3@EFB at 3412.44 cm−1 and 2921.41 cm−1, which are attributable to the celluloseeOH and CeH groups stretching vibrations, respectively. The bands at 1424 cm−1 and 1632 cm−1 represent the bending vibrations of eCH2, CeH, and CeO of cellulose in raw OPEFB which are also present in the Fe2O3@EFB. The band at 1739 cm−1 in raw OPEFB spectra is attributable to the waxy C]O acetyl group of hemicellulose ester or carbonyl ester of the p-coumaric monomeric lignin unit, and the band at 1248 cm−1 is due to the CeOeC of arylalkyl ether in lignin (Alemdar and Sain, 2008). The bands at 657 and 567 cm−1 in Fe2O3@EFB are assigned to Fe-O vibration bonds, confirming the existence of Fe components (Norouzian and Lakouraj, 2015; Sun et al., 2014). Fig. 4b shows the band at 3300–3450 cm−1 which is related to the stretching vibrations of hydroxyl groups of cellulose in Fe2O3@CEL that is becoming stronger and deeper, representing hydrogen bonding interaction between the groups of Fe2O3 nano-particles and cellulose (Liu et al., 2008). Iron oxides are often composited in situ within fibers, forming pigments and changing the color of fibers completely due to the high stability of iron oxides in the fibers (Kongdee and Bechtold, 2004). The stability of Fe2O3 nanoparticles in cellulose fibers is very important for excellent properties of the magnetic biosorbent. In the case of RKF and Fe2O3@RKF (Fig. 4c), the strong and broad band at 3410 cm−1 was assigned to the OH stretching vibration. The strong band around 2915 cm−1 is related to the asymmetric and symmetric stretching vibration in CH2 and CH3, becoming broader in Fe2O3@RKF due to the hydrogen bonding interaction between the groups of Fe2O3 nanoparticles and kapok fibers. This broadening of the band was similarly observed in Fe2O3@EFB (Fig. 4a) suggesting that the

degree of H-bonding interaction may be lesser compared to the pure cellulosic fibers (Fig. 4b). The absorption band at 1739 cm−1 was assigned to the CeO stretching vibration of ketones, carboxylic groups and esters in lignin and acetyl ester groups in xylan (Wang et al., 2012), and bands around 1373 and 1245 cm−1 within the range of the CeH and CeO bending vibration, respectively, also became broader due to the interaction with Fe2O3 nanoparticles. The band at 899 cm−1 is attributed to the CeH group removed, while the bands at 658 and 570 cm−1 represent the FeeO vibration bond (Fig. 4c). Fig. 5 shows the TGA curves with weight losses of Fe2O3 nanoparticle, raw materials (OPEFB, Cellulose, RKF) and magnetic biosorbents (Fe2O3@EFB, Fe2O3@CELand Fe2O3@RKF). The Fe2O3 clearly showed high thermal stability, while the raw sorbents were the least stable of the three. The first weight loss was observed within the range of 30–150 °C and was due to the loss of residual water that is quite small and related to the removal of absorbed physical and chemical water. The second loss was due to the decomposition of raw materials within the region of 200–380 °C which was considerable due to complete decomposition. The higher thermal stability of magnetic biosorbent than that of raw materials suggests the interaction between raw materials and Fe2O3 nanoparticle (Zhu et al., 2010). 3.1.3. Magnetic properties Magnetization measurements are mainly applied to investigate the use of magnetic nanocomposite as nano-sorbent in magnetic separation. The field dependent magnetization plots of magnetic nanosorbent (Fe2O3@EFB) exhibits superparamagnetic behavior without magnetic hysteresis reflecting the room temperature condition. Neither coercivity 97



Fig. 5. TGA curves of a) OPEFB, b) Cellulose and c) RKF; for Fe2O3 nanoparticles ⋯⋯; raw sorbents —; and magnetic biosorbents − · − · −.

Fig. 4. FTIR spectra of a) OPEFB and Fe2O3@EFB fibers; b) Cellulose and Fe2O3@CEL fibers; c) RKF and Fe2O3@RKF fibers.

Pb > Cu > Zn > Ni > Mn, and the selectivity of Pb(II) removal over other ions is due to higher atomic weight and electronegativity (Lee et al., 2015). Pb(II) has the largest atomic weight and radius, but the atomic weight and radius of Cu(II) is close to Ni(II) and Zn(II), but lower than Mn(II) (Shannon, 1976). It is, however, not clear as to why Pb(II) and Cu(II) had almost similar adsorption capacity trends, while that of Zn(II), Mn(II) and Ni(II) were far lower. Only Pb(II) belongs to the “other metal” class, and the rest belong to the same first (3d) dblock series of the transition metals in the periodic table. Although metal ion adsorption isotherms may be case-specific (Park et al., 2016), our finding was similar to the metal adsorption trend of goethite reported, which is in the order of Cu > Pb > Zn > Cd > Co > Ni > Mn (Schwertmann and Taylor, 1989). The adsorption amounts of magnetic biosorbent reached equilibrium between 45 and 60 min, much faster than the raw OPEFB sorbent which was between 90 and 120 mins (data not shown). This is to be expected as the composition of magnetic biosorbent with enhanced surface area should increase the available interaction sites and fast interaction between sorbent and sorbate (Gupta et al., 2011). An adsorption capacity of 47.8 mg/g with about 95.6% removal

nor remanence was observed, suggesting that the prepared magnetic biosorbent was superparamagnetic (Fig. S2). The maximum magnetizations of magnetic bioosorbent was found to be 19.8 emu/g. Such magnetic properties suggest that sorbents have strong magnetic properties and can be separated easily from the solution with external magnetic force (Wu et al., 2009). The prepared sorbent that dispersed within the water body can be recovered using external magnetic field, so that it can be readily re-dispersed and re-used, suggesting its potential application as an economical and effective biosorbent. 3.2. Batch heavy metal ion adsorption 3.2.1. Effect of contact time Fig. 6a shows the effect of contact time on Mn2+, Ni2+, Cu2+, Zn2+ and Pb2+ adsorption amount onto Fe2O3@EFB for 5–150 min at 500 ppm initial concentration, room temperature and pH 5.5 and sorbent dosage of 1 g/50 mL. The adsorption of Pb(II) ions was much higher than other metals due to the difference in atomic weight, ionic radius and the types of interaction between metal ions and sorbent (Park et al., 2016). The results followed the decreasing order of 98



Fig. 6. The effect of contact time on adsorption amount of a) different metal ions onto Fe2O3@EFB; b) Pb(II) ions onto three different magnetic biosorbents.

Fig. 7. The effect of solution pH on a) metal ion adsorption amount onto Fe2O3@EFB; b) Pb(II) ion adsorption amount on three different magnetic biosorbents.

efficiency was achieved for Pb(II) and 46.68 mg/g with 93.4% efficiency for Cu(II) suggested good potential of Fe2O3@ EFB sorbent. This is better than 90.5% removal efficiency of Pb(II) by magnetic chitosan at room temperature and same pH (Tran et al., 2010). It was, however, lower than 98% efficiency for Cu(II) removal by amino-functionalized magnetic nanoparticles at the same temperature (Hao et al., 2010). Ni (II) removal at 77% by Fe2O3@OPEFB after 45 mins was comparable to magnetic chitosan at 75% efficiency. There was fast initial adsorption within 2 min for about 81%, 74.5%, 67.4%, 62.2% and 51.4% removal efficacy of Pb(II), Cu(II), Zn(II), Ni(II) and Mn(II), respectively, which suggests the abundance of the available active sites on the surface of adsorbents for metal ion sorption (Fytianos et al., 2000). The effect of time on Pb(II) adsorption was examined with Fe2O3@CEL and Fe2O3@RKF and compared with Fe2O3@EFB. The adsorption capacity achieved its peak in 45 min for Fe2O3@CEL with about 96.1% removal efficiency, and around 60 min in the case of Fe2O3@RKF and Fe2O3@EFB (Fig. 6b). More than 5% increase in the adsorption capacity was achieved with Fe2O3@EFB and Fe2O3@CEL compared to the raw material, while only 2.7% increase in the removal efficiency was achieved with Fe2O3@RKF compared to RKF. The agglomeration observed on the surface of Fe2O3@RKF could have covered some parts of the surface area, blocking the sorption of Pb(II) ions, thus reducing the efficiency.

Cu(II) and M(II) were 45.6 and 39.9 mg/g, respectively, at pH 6. The highest 92.2% removal efficiency of Cu(II) with Fe2O3@EFB was10.2% higher than the EFB raw sorbent. In the case of Fe2O3@RKF, the maximum Pb(II) ions sorption was 49 mg/g at pH 4 with about 98% removal efficiency, and 48.7 mg/g for Fe2O3@CEL at an optimum pH of 5–6 (Fig. 7b). The pH at the zero-point charge (pHzpc) of Fe2O3@EFB was 5.5. At pH > pHzpc, the surface charge of the sorbent became negative and the electrostatic interactions between the metal ions and adsorbent became stronger (Rao and Ikram, 2011). Carboxylic groups (–COOH) of agricultural sorbents participate in the adsorption process, and the ones on the surface of the agro-based sorbents have pKa values of 3–5. At low pH, the surface of the adsorbents exists in a carboxyl form, and the high concentration of the H+ and H3O+ ions at low pH also lead to competition with positive cations for the adsorption sites, resulting in low adsorption and lower removal efficiency (Daneshfozoun et al., 2016). With the increase in solution alkalinity, the carboxyls turn into carboxylate anions and the adsorption increases steadily. At pH higher than 5, the acidic groups of –COOH start to dissociate, resulting in more interaction between positively charged metal ions and the negatively charged carboxylate ions, thus leading to higher removal efficiency. At pH around 6, the divalent cations could interact with the hydroxide (–OH) groups on the sorbent surface and be adsorbed by the hydrogen bonding and ion exchange mechanisms, resulting in greater adsorption. When the carboxyls have completely turned into carboxylate anions, there will be no change in adsorption (Ge et al., 2012). The adsorption of metal ions can therefore be assumed to be affected by hydrogen bonding, ion exchange and surface complexation.

3.2.2. Effect of pH Fig. 7a shows the effect of pH changes from 3 to 10 on the adsorption of different metal ions onto Fe2O3@EFB at 500 ppm initial concentration, at optimum time, and with sorbent dosage of 1 g/50 mL. The adsorption capacity increased with the increasing pH, reaching plateau at 5–7, and dropping beyond pH 7. The Pb(II), Zn(II) and Ni(II) sorption were 48.4, 43.51 and 31.77 mg/g, respectively, at pH 7, while

3.2.3. Effect of initial ion concentration and temperature The effect of the initial concentrations on metal ion adsorption was 99



Fig. 8. The effects of initial concentration and temperature on the removal efficiency of a) Pb(II), b) Cu(II), c) Zn(II), d) Mn(II), e) Ni(II) on Fe2O3@EFB.

endothermic and the removal efficiency was almost constant at 338 K (Fig. 9). The efficiency was already high at 97–98% and 98–99%, respectively, that the change was only less than 0.5% because the temperature was increased from 298 to 328 K. At 338 K, Fe2O3@EFB achieved maximum adsorption capacity of 98.0 mg/g at 1000 ppm Pb (II) with 98% removal efficiency. This is equivalent to the one achieved with Fe2O3@CEL (98.2% removal efficiency) at 1 g/50 mL, pH 6 and 45 min of contact time. The maximum removal efficiency of Fe2O3@ RKF was 99.4% and exhibited an endothermic nature of Pb(II) sorption with about 1.13% increase in efficiency when the temperature was increased to 338 K. The endothermic sorption has been reported for Pb (II) and Ni(II) sorption onto functionalized magnetic graphenes (Guo et al., 2014), which is slightly contradictory to our findings. However, the Pb(II) sorption onto cellulose/Fe3O4/activated carbon composite was reportedly exothermic, with an insignificant temperature effect on

tested between 100 and 1000 mg/L at an optimum pH and time. The temperature varied from 298 to 338 K for different initial concentration. An increased adsorption capacity of Fe2O3@EFB was observed with the increase in the initial concentrations for all samples, but the removal efficiency decreased slightly (Fig. 8). At high ion concentrations, there will be greater competition for accessible sites for adsorption. For the same number of sorption sites, there will be reduced efficiency with high ion concentrations as the sorbed ions leave lower available sites for sorption (Daneshfozoun et al., 2014). For Fe2O3@ EFB, the removal efficiency of all ions, except Ni(II), slightly increased with the increasing temperature, indicating the endothermic nature of the reactions. However, the Ni(II) sorption process showed exothermic reaction, which was similar to the raw EFB sorbent (results not shown) where the removal efficiency decreased with increasing temperature. The Pb(II) adsorptions onto Fe2O3@CEL and Fe2O3@RKF were also 100



signifying the absence of irreversible sites on the sorbent surface (Fig. S3). The Fe2O3@EFB could therefore preserve good metal removal efficiency with only about 3–7% reduction in efficiency as compared to 20% efficiency reduction with the raw OPEFB sorbent (results not shown) after five consecutive adsorption–desorption processes. The removal efficiencies of Fe2O3@RKF and Fe2O3@CEL after 5 cycles were maintained at more than 98% and 95% of their original Pb (II) adsorption capacity, which is comparable to the performance of raw sorbents. These suggest the better potential for reusability and recoverability, thus confirming the advantages of using magnetic biosorbent.

3.4. Comparison of EFB, cellulose and Ceiba pentandra magnetic biosorbents Comparing the three different raw and magnetic biosorbents (Fig. 10), Fe2O3@RKF showed the best Pb(II) removal efficiency with as high as 99.4% and the highest adsorption capacity of 49 mg/g. The Fe2O3@CEL and Fe2O3@EFBshowed 98.2% and 97.7% Pb(II) ion removal, respectively, which can be equally considered as highly efficient sorbents. The magnetic biosorbents in general exhibited 10.3% higher removal efficiency than the raw sorbents. The existence of magnetic nanoparticles on the sorbent surface could have enhanced the number of active sites, in addition to the available pore volume for chemisorption. Plant fiber consists of cellulosic and non-cellulosic constituents. Cellulose, hemicelluloses and lignin are common constituents that contribute towards hydrophobic properties of plant surface, which regulates the moisture uptake of natural fibers. The non-cellulosic constituents that include proteins, amino acids, other nitrogen compounds, wax, pectin substances, organic acids, sugar, inorganic salts, and small amount of pigments are located principally in the cuticle, in the primary cell wall, and in the lumen. Cellulose is a predominant constituent of lignocellulosic material and forms structural part, as organic polymeric macromolecules have a linear, long chain of a hundred to a thousand 1-4-β-D-anhydroglucose units or glucosidic linkage (Abdullah et al., 2011). OPEFB fibers contain cellulose (40–50%), hemicellulose (20–30%) and lignin (15–20%) of the biomass (Mahjoub et al., 2013). Kapok fiber comprises of 64% cellulose, 13% lignin, and 23% pentosan (Fengel and Przyklenk, 1986), and another report suggests 35% cellulose, 21.5% lignin, and 22% xylan, with a high proportion of syringyl/guaiacyl units (4–6%) and acetyls (13%) (Huang and Lim, 2006). This difference can be attributed to the difference in kapok sources and preparing systems. The complex lignocellulosic composite structure of cellulose, hemicellulose and lignin are held together through CeC, ether, ester and Hbond (Douglas et al., 2012). Each type of interacting force is a characteristic of polymer molecule such as hemicellulose possessing ether bond, lignin showing carbon–carbon and ether bond, and cellulose exhibiting H-bond. The three types of polymeric chain have different types of bond connectivity in complex lignocellulose:- a) the interaction between hemicelluloses-lignin polymeric chains is ether and ester linkage; b) cellulose-lignin polymeric chains has ether linkage; and c) H-bond involves in holding cellulose-hemicellulose, ligninhemicellulose and cellulose-lignin (Faulon et al., 1994). The complex network of fibers with the availability of functional groups could bind the metal ions, resulting in the net increase in efficiency. In addition, the waxy esters and OH− layers and the hollow lumen network of kapok fibers (Abdullah et al., 2010) could have assisted in achieving greater sorption of metal ions than the OPEFB or cellulose. The magnetic properties ensure that the sorbent can be separated from the medium easily and reused many times. From a practical application’s point-of-view, OPEFB magnetic biosorbents can be considered as the best choice due to the large abundance of waste materials from oil palm plantations compared to kapok, while cellulose needs additional extraction and processing steps that will consequently incur extra cost.

Fig. 9. The effects of initial concentration and temperature on the removal efficiency of Pb(II) on a) Fe2O3@CEL, and b) Fe2O3@RKF.

Fig. 10. Comparison between raw and magnetic biosorbent of three different agro-based materials.

the removal efficiency (Zhu et al., 2011). Hence, the effect of temperature on the ion removal efficiency is very much dependent on the type of metal ions and the adsorbent-adsorbate system.

3.3. Reusability of magnetic biosorbents Metal ions desorption and adsorbent re-generation are two important issues when considering adsorbent re-usability. The important aspect of regeneration process is to restore sorbent adsorption capacity and recover metal components. To estimate the reusability of magnetic biosorbent and also recovery of metal ions, regeneration and desorption studies were done with HCl acid and the adsorption–desorption processes were run for five consecutive cycles. The adsorption capacity of the magnetic biosorbent remained almost constant for the five cycles 101



M., 2010. Adsorption of Cd (II) and Zn (II) from aqueous solutions using magnetic hydroxyapatite nanoparticles as adsorbents. Chem. Eng. J. 162, 487–494. Fengel, D., Przyklenk, M., 1986. Studies on kapok. 2. chemical investigation. Holzforschung 40, 325–330. Foo, Y.N., Foong, K.Y., Yousof, B., Kalyana, S., 2011. A renewable future driven with malaysian palm oil-based green technology. J. Oil Palm Environ. 2, 1–7. Fytianos, K., Voudrias, E., Kokkalis, E., 2000. Sorption-desorption behaviour of 2, 4dichlorophenol by marine sediments. Chemosphere 40, 3–6. Gómez-Pastora, J., Bringas, E., Ortiz, I., 2014. Recent progress and future challenges on the use of high performance magnetic nano-adsorbents in environmental applications. Chem. Eng. J. 256, 187–204. Ge, F., Li, M.-M., Ye, H., Zhao, B.-X., 2012. Effective removal of heavy metal ions Cd 2+, Zn 2+, Pb 2+, Cu 2+ from aqueous solution by polymer-modified magnetic nanoparticles. J. Hazard. Mater. 211, 366–372. Guo, X., Du, B., Wei, Q., Yang, J., Hu, L., Yan, L., Xu, W., 2014. Synthesis of amino functionalized magnetic graphenes composite material and its application to remove Cr (VI), Pb (II), Hg (II), Cd (II) and Ni (II) from contaminated water. J. Hazard. Mater. 278, 211–220. Gupta, V., Agarwal, S., Saleh, T.A., 2011. Chromium removal by combining the magnetic properties of iron oxide with adsorption properties of carbon nanotubes. Water. Res. 45, 2207–2212. Hao, Y.-M., Man, C., Hu, Z.-B., 2010. Effective removal of Cu (II) ions from aqueous solution by amino-functionalized magnetic nanoparticles. J. Hazard. Mater. 184, 392–399. Huang, X., Lim, T.-T., 2006. Performance and mechanism of a hydrophobic?oleophilic kapok filter for oil/water separation. Desalination 190, 295–307. Jiménez-Cedillo, M., Olguín, M., Fall, C., Colin-Cruz, A., 2013. As (III) and As (V) sorption on iron-modified non-pyrolyzed and pyrolyzed biomass from Petroselinum crispum (parsley). J. Environ. Manag. 117, 242–252. Kongdee, A., Bechtold, T., 2004. In-fibre formation of Fe (OH) 3—a new approach to pigment coloration of cellulose fibres. Dyes Pigm. 60, 137–142. Lee, C.-G., Jeon, J.-W., Hwang, M.-J., Ahn, K.-H., Park, C., Choi, J.-W., Lee, S.-H., 2015. Lead and copper removal from aqueous solutions using carbon foam derived from phenol resin. Chemosphere 130, 59–65. Liu, S., Zhang, L., Zhou, J., Wu, R., 2008. Structure and properties of cellulose/Fe2O3 nanocomposite fibers spun via an effective pathway. J. Phys. Chem C 112, 4538–4544. Mahdavian, A.R., Mirrahimi, M.A.-S., 2010. Efficient separation of heavy metal cations by anchoring polyacrylic acid on superparamagnetic magnetite nanoparticles through surface modification. Chem. Eng. J. 159, 264–271. Mahjoub, R., Mohamad Yatim, J.B., Mohd Sam, A.R., 2013. A review of structural performance of oil palm empty fruit bunch fiber in polymer composites. Adv. Mater. Sci. Eng (Article ID 415359). Malaysian Palm Oil Council, 2012. Malaysian Palm Oil: Assuring Sustainable Supply of Oil & Fats into the Future. http:/www.ceopalmoil.com. Nalbandian, M.J., Zhang, M., Sanchez, J., Choa, Y.-H., Nam, J., Cwiertny, D.M., Myung, N.V., 2016. Synthesis and optimization of Fe2O3 nanofibers for chromate adsorption from contaminated water sources. Chemosphere 144, 975–981. Nassar, N.N., 2010. Kinetics, mechanistic, equilibrium, and thermodynamic studies on the adsorption of acid red dye from wastewater by γ-Fe2O3 nanoadsorbents. Sep. Sci. Technol. 45, 1092–1103. Nazir, M.S., Wahjoedi, B.A., Yussof, A.W., Abdullah, M.A., 2013. Eco-friendly extraction and characterization of cellulose from oil palm empty fruit bunches. BioResour 8, 2161–2172. Norouzian, R.-S., Lakouraj, M.M., 2015. Preparation and heavy metal ion adsorption behavior of novel supermagnetic nanocomposite based on thiacalix [4] arene and polyaniline: conductivity, isotherm and kinetic study. Synth. Met. 203, 135–148. Okoro, I., Okoro, S., 2011. Agricultural by products as green chemistry absorbents for the removal and recovery of metal ions from waste-water environments. Cont. J. Water Air Soil Pollut. 2, 15–22. Park, J.-H., Ok, Y.S., Kim, S.-H., Cho, J.-S., Heo, J.-S., Delaune, R.D., Seo, D.-C., 2016. Competitive adsorption of heavy metals onto sesame straw biochar in aqueous solutions. Chemosphere 142, 77–83. Ramos, S.N.C., Xavier, A.L.P., Teodoro, F.S., Gil, L.F., Gurgel, L.V.A., 2016. Removal of cobalt (II), copper (II), and nickel (II) ions from aqueous solutions using phthalatefunctionalized sugarcane bagasse: mono-and multicomponent adsorption in batch mode. Ind. Crops Prod. 79, 116–130. Rao, R.A.K., Ikram, S., 2011. Sorption studies of Cu (II) on gooseberry fruit (emblica officinalis) and its removal from electroplating wastewater. Desalination 277, 390–398. Schwertmann, U., Taylor, R., 1989. Iron oxides. In: Dixon, J.B., Weed, S.B. (Eds.), Minerals in Soil Environments. Soil Sci. Soc. Am., Madison WI, pp. 379–438. Sharma, P., Bihari, V., Agarwal, S.K., Verma, V., Kesavachandran, C.N., Pangtey, B.S., Mathur, N., Singh, K.P., Srivastava, M., Goel, S.K., 2012. Groundwater contaminated with hexavalent chromium [Cr (VI)]: a health survey and clinical examination of community inhabitants (Kanpur, India). PLoS One 7, e47877. Squissato, A.L., Lima, A.F., Almeida, E.S., Pasquini, D., Richter, E.M., Munoz, R.A., 2017. Eucalyptus pulp as an adsorbent for metal removal from biodiesel. Ind. Crops Prod. 95, 1–5. Sun, X., Yang, L., Li, Q., Zhao, J., Li, X., Wang, X., Liu, H., 2014. Amino-functionalized magnetic cellulose nanocomposite as adsorbent for removal of Cr (VI): Synthesis and adsorption studies. Chem. Eng. J. 241, 175–183. Tan, Y., Chen, M., Hao, Y., 2012. High efficient removal of Pb (II) by aminofunctionalized Fe3O 4 magnetic nano-particles. Chem. Eng. J. 191, 104–111. Tran, H.V., Dai Tran, L., Nguyen, T.N., 2010. Preparation of chitosan/magnetite composite beads and their application for removal of Pb (II) and Ni (II) from aqueous

4. Conclusion The morphological, chemical and magnetic characterization showed successful preparation of magnetic biosorbent from EFB, cellulose and kapok. The optimum sorption was accomplished between pH 5–7, increased initial concentration and temperature. The reusability of sorbents was proven after 5 cycles of adsorption and desorption. The magnetic biosorbent based on kapok showed the highest Pb(II) removal of 99.4% as compared to 98.2% for cellulose and 97.7% for EFB. However, with 78–98% removal efficiency, the EFB magnetic biosorbent may be a better choice as green, eco-friendly and cost-effective sorbent for heavy metal ions removal in the waste water treatment due to larger area of oil palm plantation and distribution, resulting in greater abundance of biomaterials and EFB for re-utilization. Acknowledgment The authors would like to thank Universiti Teknologi PETRONAS for the scholarship to Safoura Daneshfozoun and the research facilities. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.indcrop.2017.05.011. References Abdullah, M.A., Rahmah, A.U., Man, Z., 2010. Physicochemical and sorption characteristics of Malaysian Ceiba pentandra (L.) Gaertn. as a natural oil sorbent. J. Hazard. Mater. 177, 683–691. Abdullah, M.A., Afzaal, M., Ismail, Z., Ahmad, A., Nazir, M.S., Bhat, A., 2015. Comparative study on structural modification of Ceiba pentandra for oil sorption and palm oil mill effluent treatment. Desalin. Water Treat. 54, 3044–3053. Abdullah, M.A., Nazir, M.S., Raza, M.R., Wahjoedi, B.A., Yussof, A.W., 2016. Autoclave and ultra-sonication treatments of oil palm empty fruit bunch fibers for cellulose extraction and its polypropylene composite properties. J. Clean Prod. 126, 686–697. Afzaal, M., Periyasamy, B., Abdullah, M.A., 2014. Continuous heavy metal removal from Palm Oil Mill Effluent using natural Ceiba pentandra (L.) Gaertn packed-bed column. Appl. Mech. Mater. 625, 822–825. Alemdar, A., Sain, M., 2008. Isolation and characterization of nanofibers from agricultural residues–wheat straw and soy hulls. Bioresour. Technol. 99, 1664–1671. Ali, I., Asim, M., Khan, T.A., 2012. Low cost adsorbents for the removal of organic pollutants from wastewater. J. Environ. Manag. 113, 170–183. Ali, I., 2012. New generation adsorbents for water treatment. Chem. Rev. 112, 5073–5091. Aragay, G., Pons, J., Merkoçi, A., 2011. Recent trends in macro-, micro-, and nanomaterial-based tools and strategies for heavy-metal detection. Chem. Rev. 111, 3433–3458. Asberry, H.B., Kuo, C.-Y., Gung, C.-H., Conte, E.D., Suen, S.-Y., 2014. Characterization of water bamboo husk biosorbents and their application in heavy metal ion trapping. Microchem. J. 113, 59–63. Bagal-Kestwal, D., Karve, M.S., Kakade, B., Pillai, V.K., 2008. Invertase inhibition based electrochemical sensor for the detection of heavy metal ions in aqueous system: application of ultra-microelectrode to enhance sucrose biosensor's sensitivity. Biosens. Bioelectron. 24, 657–664. Bailey, S.E., Olin, T.J., Bricka, R.M., Adrian, D.D., 1999. A review of potentially low-cost sorbents for heavy metals. Water Res. 33, 2469–2479. Bulgariu, L., Ratoi, M., Bulgariu, D., Macoveanu, M., 2008. Equilibrium study of Pb (II) and Hg (II) sorption from aqueous solutions by moss peat. Environ. Eng. Manag. J. 7, 511–516. Chen, Y., Wu, F., Liu, M., Parvez, F., Slavkovich, V., Eunus, M., Ahmed, A., Argos, M., Islam, T., Rakibuz-Zaman, M., 2013. A prospective study of arsenic exposure, arsenic methylation capacity, and risk of cardiovascular disease in Bangladesh. Environ. Health Perspect. 121, 832. Daneshfozoun, S., Abdullah, B., Abdullah, M.A., 2014a. Heavy metal removal by oil palm empty fruit bunches (OPEFB) biosorbent. Appl. Mech. Mater. Trans Tech Publ. 889–892. Daneshfozoun, S., Nazir, M.S., Abdullah, B., Abdullah, M.A., 2014b. Surface modification of celluloses extracted from oil Palm empty fruit bunches for heavy metal sorption. Chem. Eng. Trans. 37, 679–684. Daneshfozoun, S., Abdullah, B., Abdullah, M.A., 2016. The effects of oil palm empty fruit bunch sorbent sizes on plumbum (II) ion sorption. Adv. Mater. Res. Trans Tech Publ. 542–546. Faulon, J.L., Carlson, G.A., Hatcher, P.G., 1994. A three-dimensional model for lignocellulose from gymnospermous wood. Org. Geochem. 21, 1169–1179. Faust, S.D., Aly, O.M., 2013. Adsorption Processes for Water Treatment. Elsevier. Feng, Y., Gong, J.-L., Zeng, G.-M., Niu, Q.-Y., Zhang, H.-Y., Niu, C.-G., Deng, J.-H., Yan,

102



Yusoff, S., 2006. Renewable energy from palm oil −innovation on effective utilization of waste. J. Clean Prod. 14, 87–93. Zheng, Y., Wang, J., Zhu, Y., Wang, A., 2015. Research and application of kapok fiber as an absorbing material: a mini review. J. Environ. Sci. 27, 21–32. Zhu, J., Wei, S., Ryu, J., Sun, L., Luo, Z., Guo, Z., 2010. Magnetic epoxy resin nanocomposites reinforced with core- shell structured Fe@ FeO nanoparticles: fabrication and property analysis. ACS Appl. Mater. Interfaces 2, 2100–2107. Zhu, H.-Y., Fu, Y.-Q., Jiang, R., Jiang, J.-H., Xiao, L., Zeng, G.-M., Zhao, S.-L., Wang, Y., 2011. Adsorption removal of congo red onto magnetic cellulose/Fe3O4/activated carbon composite: equilibrium, kinetic and thermodynamic studies. Chem. Eng. J. 173, 494–502.

solution. Mater. Sci. Eng. C 30, 304–310. Volesky, B., 2007. Biosorption and me. Water Res. 41, 4017–4029. Wang, J., Zheng, Y., Wang, A., 2012. Effect of kapok fiber treated with various solvents on oil absorbency. Ind. Crops Prod. 40, 178–184. Wu, Y., Wang, Y., Luo, G., Dai, Y., 2009. In situ preparation of magnetic Fe3O4-chitosan nanoparticles for lipase immobilization by cross-linking and oxidation in aqueous solution. Bioresour. Technol. 100, 3459–3464. Yantasee, W., Lin, Y., Hongsirikarn, K., Fryxell, G.E., Addleman, R., Timchalk, C., 2007. Electrochemical sensors for the detection of lead and other toxic heavy metals: the next generation of personal exposure biomonitors. Environ. Health Perspect 1683–1690.

103