Modification and Application of Albizia lebbeck Sawdust For The Sorption of Lead(II) and Copper(II) From Aqueous Solutions

Sawdust of Albizia lebbeck (locally known as koroi) was modified by treatment with sodium methylate followed by reactions with epichlorohydrin and n-dodecylamine to afford the sorbent which was termed as DDA-SD. The sorbent was characterized with the help of FTIR (ATR), SEM and chemical test. It was then used for the sorption of Pb2+ and Cu2+ from aqueous solutions and the effects of solution pH, dosage of sorbent, initial metal ions concentration and contact time were enumerated. The sorption data were tested for Langmuir and Freundlich isotherm models and found that the process followed Langmuir model and the maximum sorption capacities were calculated as 17.37 mg/g and 9.4 mg/g for Pb2+ and Cu2+ ions respectively. The sorption data were also tested for pseudo-first order and pseudo-second order kinetic models and the data fitted well with pseudo second order kinetics. Exhausted DDA-SD adsorbent can be regenerated using 0.1M HCl or 0.1M HNO3 followed by wash with deionized water and reused for further adsorption purpose.


INTRODUCTION
Disposal of untreated industrial wastes is the major source of heavy metals contamination, which causes severe water, soil and air pollution.
The persistent nature of heavy metals results their accumulation in plants and animal tissues and ultimately enter the food chain. When heavy metals enter into human body through contaminated foods, they are bio-amplified and cause serious threat to the public health [1][2] . From various industrial activities lead and copper are exposed to the environment and cause detrimental effects such as malfunction of brain, cognitive disorder to the children, renal diseases, anemia and loss of reproductive power 3 . Many techniques such as precipitation, ion-exchange, membrane separation, electrodialysis, extraction etc. are reported for removing heavy metals 4,5 . However, these techniques are not attractive because of high cost and difficulties associated with the handling of toxic sludges produced during the processes 1,4,6 . Adsorption processes have drawn much attention due to their high efficacy in removing heavy metals 4 and availability of a wide variety of adsorbents such as activated carbon 7 , clay minerals [8][9] , zeolites 10 , metal oxides [11][12] , agricultural wastes [13][14][15][16][17][18][19][20] , biomass 21 polymeric materials 22 , which demonstrated significant adsorption capacities for heavy metals. Although activated carbon possesses very high adsorption capacities for various organic compounds and dyes, it is not that effective in removing heavy metals. Moreover, activated carbon is too expensive and thus has found less application in removing heavy metals 4 . On the other hand, the use of untreated plant wastes were reported to have poor sorption capacities and releases soluble organic substances and make the treated water unsuitable for uses 23,24 . During chemical treatment of plant materials most of the soluble organic materials are removed and simultaneously the metal sorption capacity is enhanced 4 .
Sawdust is a by-product of wood industries and can be obtained at negligible cost. Earlier sawdust of different plants were chemically modified using a variety of reagents and applied for sequestration of metals ions, e.g. alkali treated sawdusts demonstrated adsorption capacities of 6.92 mg/g and 15.8 mg/g of Cu 2+ and Zn 2+ respectively 25 , while acid treated sawdust showed Cu 2+ adsorption capacities of 3.6-13.95 mg/g [26][27][28] . Few research groups adopted treatment method using formaldehyde in sulfuric acid to enhance the adsorption capacity of sawdust but ended with poor results [29][30][31] . Recently we observed that sawdust impregnated with nano-zerovalent iron showed significant sorption of As(III) from aqueous solution 32 . Other recent studies showed that sawdust 33 , nano-sawdust 34 , sawdust-kaolinite composite 35 could be used as adsorbents for heavy metals. Modification of pure cellulose were reported to have good to excellent adsorption capacities of heavy metals, e.g etherification of cellulose with epichlorohydrin followed by epoxide ring opening by polyethyleneimine afforded an adsorbent that showed moderate adsorption capacities of 25 mg/g and 15 mg/g of Pb 2+ and Cu 2+ respectively 36 . An adsorbent prepared by oxidation of pure cellulose with periodate and hydroxamic acid demonstrated excellent Cu 2+ sorption capacity of 246 mg/g 37 .
To the best of our knowledge no report is available on the development of adsorbent for heavy metal by modification of sawdust using etherification followed by ring opening by an alkyl amine strategy. In the present investigation modification of sawdust of Albizia lebbeck was carried out by treatment with sodium methylate, followed by etherification using epichlorohydrin and oxirane ring opening by n-dodecylamine to afford a material, which was applied for the sorption of Cu 2+ and Pb 2+ ions.

Materials and reagents
Sawdust of Albizia lebbeck (Local name Koroi) obtained from a sawmill located at Mirpur in Dhaka, Bangladesh was washed with water and sundried for 3 days followed by drying in an oven at 50 o C for 72 h and cooled in desiccators. The dried material was sieved and the fraction with a particle size between 0.2 and 0.6 mm was used for the preparation of chemically modified adsorbent. Chemical analyses of dried sawdust were carried out using literature methods 38 and found that it contained 7.25% extractible materials, 68.7% cellulose and 17% lignin.
Sodium methylate, dimethylsulfoxide (DMSO), epichlorohydrin, ethanol, n-dodecylamine, copper sulfate pentahydrate, lead nitrate, sodium hydroxide and nitric acid were purchased from Aldrich and used as received. Stock solutions of the metals were obtained by dissolving required amounts of CuSO 4 .5H 2 O and Pb(NO 3 ) 2 in deionized (DI) water. A small volume (10-20 µl) of 0.1M NaOH or 0.1M HNO 3 were added to the metal ions solutions to adjust the pH.

Preparation of the adsorbent
Modification of sawdust was carried out following a literature procedure 36 . In brief, sawdust (5.0 g) was suspended in DMSO (100 mL), purged with N 2 and stirred in a hot plate magnetic stirrer for 2 h at 60 o C. The mixture was cooled to room temperature, treated with sodium methylate (15.0 g) and stirred for 1 h in N 2 atmosphere. The resulting solids were separated by filtration and washed with DI water (200 mL) and dried at 50 o C for 3 h to afford the material NaO-SD (4.5 g). To the suspension of NaO-SD (3.0 g) in DMSO (50 ml) epichlorohydrin (30 mL) was added and stirred for 2 h at 50 o C in N 2 atmosphere. Solid materials thus obtained were separated by filtration and washed with DI water and dried at 50 o C for 3 h to afford ECH-SD (3.12 g). ECH-SD (1.4 g) was suspended in DMSO (30 mL), n-dodecylamine (3.2 g) was added in portions and heated at 100 o C with stirring for 18 hours. The final material was separated by filtration and washed with DI water and dried in an oven at 50 o C for 3 h and cooled in desiccators to afford the sorbent DDA-SD (1.4 g).

Instruments
The functional groups present in the sorbent (DDA-SD) were identified by FTIR-ATR (Shimadzu, Japan) and the surface morphology of it was determined using a scanning electron microscope (Philips XL30, FEI, USA). The surface area was determined by BET analysis, which was performed on a Quantachrome Autosorb-1. For BET experiment DDA-SD sample were out gassed at 120°C for 24 h before analysis and Krypton gas was used as the adsorbent. To obtain a value of surface area, 11 points were measured in the standard range of (P/Po) = 0.05 to 0. 3

Batch sorption experiments
Metal uptake capacities of the DDA-SD were determined by adding 100 mg of DDA-SD into 100 mL of metal ion solution taken in a conical flask. The experiments were carried out at 25 o C under shaking at 180 rpm in an orbital shaker. After completion of the experiments the solutions were filtered with Whatman 41 filter paper and the filtrates were then analyzed with AAS. The sorption capacity, q e was determined according to the equation (1).
Where Ci and Ce stand for initial and equilibrium concentrations of the metal ion (g/L), m represents mass of the adsorbent (g) and V is the volume of the solution (L).
To observe the effect of pH on the sorption process 100 mL of a metal ion solutions were taken in different conical flasks, pH were adjusted 1.9-6 for Pb(II) and 3-10.2 for Cu(II) by adding dilute nitric acid or sodium hydroxide and 100 mg of DDA-SD was added to each flask and left under shaking for 1 hour. Afterwards sorbents were separated by filtration and sorption capacities were calculated using the equation (1).
Effect of dosage was investigated by adding different amounts of the DDA-SD (50-350 mg) to 100 mL metal ion solutions of fixed concentration (25.0 ppm) in different conical flasks, pH were adjusted 5.8 and 7.1 for Pb 2+ and Cu 2+ respectively. Effect of contact time was monitored by adding fixed amount of the sorbent (500 mg) to metal ion solutions of concentrations of 23.5, 38.0 ppm for Cu 2+ and 9.5, 28.5 ppm for Pb 2+ and pH were maintained at 5.8 for Pb 2+ and 7.1 for Cu 2+ and left under shaking. Aliquots of 2 mL were drawn at different time intervals and metal ion concentrations were measured by AAS.

Desorption study
Desorption study was conducted by soaking 20 mg of the exhausted DDA-SD in 20 mL of 0.1M HNO 3 or 0.1M HCl and left under shaking for 3 hours. An aliquot of 0.5 mL was drawn in every 30 min and metal ion concentrations were analyzed using AAS.

Synthesis of the adsorbent
The major chemical constituents of sawdust are cellulose, lignin and hemicellulose. The reactions occurred during the modification process are represented in the scheme 1. The primary hydroxyl group of the β unit (1-4 linked D-glucose) of cellulose reacted with sodium methylate to afford Cell-CH 2 -ONa (NaO-SD). During treatment with sodium methylate in DMSO a considerable amount of extractible materials, hemicellulose and lignin were removed from the cellulosic material 38 . Cell-CH 2 -ONa then reacted with epichlorohydrin to afford cellulose derivative containing oxirane moiety (ECH-SD). The oxirane ring was opened by reaction with n-dodecyl amine to afford the sorbent (DDA-SD). method 36 . On the treatment of ECH-SD with sodium thiosulfate the oxirane ring breaks down and liberates an equimolar amount of NaOH which can be quantified by titration with standardized HCl in presence of phenolphthalein indicator. Using this strategy the amount of epoxide ring (mg) per gm of DDA-SD was determined by using the following equation (2). The amount of epoxide moiety was found to be 0.37 mg/g which was in good agreement with the previous result 36 .

Characterization
FTIR Analysis: FTIR analysis of NaO-SD, ECH-SD and DDA-SD were carried out using ATR sampling technique and vibrational frequencies are summarized in the Table 1.

SEM Analysis
SEM images of NaO-SD and DDA-SD revealed similar morphology with rough surfaces in both cases (Figure 2).

Surface Area
The surface area of NaO-SD and DDA-SD were measured by BET analysis and found to be 0.75 m 2 /g and 0.73 m 2 /g respectively, which revealed that the materials have comparable surface areas.

Effect of pH
Depending on the pH of the solution the surface charge of an adsorbent varies significantly, which plays a vital role in the sorption process. The effect of pH on the sorption of Pb 2+ and Cu 2+ was studied by adding a constant dosage of DDA-SD (20 mg/20 mL) into and metal ion solutions of fixed concentrations (50 ppm for Pb 2+ and 5 ppm for Cu 2+ ) in different conical flasks maintaining pH in the range of 1.9 to 10. The result showed that with increasing pH the sorption of Pb 2+ increased dramatically and reached the maximum at pH 5.9. Effect of further increase of pH was not monitored as Pb 2+ precipitated beyond the pH of 5.9. In case of Cu 2+ the maximum sorption was achieved at pH 7.1 and it remained same up to the pH 10. It is assumed that at low pH some of the NH groups of DDA-SD were protonated, which hindered the chelation of metal ions and resulted lower sorption (Scheme 3).

Fig. 3. Effect of pH on sorption of Pb 2+ and Cu 2+ on DDA-SD
It is clear that sorption capacity decreased with the increasing amounts of dosage and maximum adsorption capacity was obtained at a dosage of 0.5g/L (Fig. 4). This is because almost all the active sites of sorbent was efficiently occupied by Pb 2+ ions and at higher dosage although the number of active sites increased a situation was created where a surplus of adsorption sites existed and resulted lower sorption capacity.

Effect of Time
The equilibration time is considered as the optimum contact time, which was determined by measuring the sorption capacities at different time. In a conical flask 500 mL of the metal ion solution was taken, the pH was adjusted (5.9 for Pb 2+ and 7.1 for Cu 2+ ) and 500 mg of DDA-SD was added. The solution was left under shaking and an aliquot of 2 mL was withdrawn at different time intervals, filtered, analyzed by AAS and the results are shown in the Fig. 5. It was evident that for all experiments the equilibration time was 60 minutes. It was also found that the sorption capacity of DDA-SD increased with increasing metal ion concentration, thus sorption capacity of Pb 2+ increased from 7.97 mg/g to 17.47 mg/g as the concentration was raised from 9.5 ppm to 28.5 ppm, while that for Cu 2+ increased from 10.32 mg/g to 17.8 mg/g when the concentration was increased from 23.5 to 38 ppm. The higher sorption capacity at higher concentration is attributed to the higher probability of collision between sorbate and sorbent surface 39 .

Effect of dosages
The effect of dosages on the sorption of Pb 2+ was carried out using DDA-SD dosages ranging from 0.5 g/L to 3.5 g/L at pH 5.85 and at room temperature. The experiments were carried out under shaking in an orbital shaker at rpm 180 calculated and shown in the Fig. 7 and Table 2. The maximum sorption capacities were 17.37 mg/g and 9.4 mg/g for Pb 2+ and Cu 2+ respectively. It was clear that sorption process of both Pb 2+ and Cu 2+ on DDA-SD followed Langmuir isotherm which was attributed to the fact that the sorption occurred through the formation of monolayer at homogeneous surface of the sorbent and there was no transmigration of sorbate particles during the process 40 .   A dimensionless parameter known as separation factor (R L ) was obtained from the Langmuir constant, K L using the equation (4).
Where Ci is the initial concentration of the metal ion (mg/L) and K L is the Langmuir constant (L/mg).

Adsorption isotherms
The relationship between the amount of sorbate attached on the surface of a sorbent and its concentration in the equilibrium solution at a constant temperature is represented by sorption isotherm. Fig. 6 shows the equilibrium sorption data for Pb 2+ and Cu 2+ on DDA-SD.
The relationship between the equilibrium sorption capacities, q e (mg/g) with the metal ion concentrations at equilibrium, C e (mg/L) can be obtained by using Langmuir (equ.3) and Freundlich (equ. 4) isotherm models.
Where C e is the equilibrium concentration of sorbate (mg. L -1 ); qe is the equilibrium sorption capacity (mg. g -1 ); qmax is the maximum monolayer sorption capacity (mg. g -1 ) and K L is Langmuir constant (L. mg -1 ).
From the slope and intercept of the Langmuir plot the values of K L and qmax were R L values indicate the feasibility of the sorption process, thus R L = 0, 0<R L <1, R L = 1 and R L >1 indicate that the process is irreversible, favourable, linear and unfavourable respectively. The values of R L in the present study were between 0 and 1 ( Table 3) and indicated that the sorption process was favourable.
The sorption data were then evaluated for Freundlich model using the equation (5).
The Freundlich constant, K F indicates sorption capacity, while 1/n is a function of the energy of sorption process. It is evident from the Fig. 8 that the sorption process did not follow Freundlich isotherm model indicating that the processes did not proceed through the formation of a multilayer on heterogeneous surface 40 .
concentration as well as the number of available active sites of the sorbent. The sorption capacities of DDA-SD were calculated from the pseudo-second-order kinetics plots and listed in the Table 4 and found that the sorption capacities were similar to those obtained from Langmuir isotherm model.

Sorption Kinetics
To understand the kinetics of the sorption of Pb 2+ and Cu 2+ on DDA-SD, the pseudo first order (equ. 6) and pseudo second order (equ. 7) kinetic equations were tested for the sorption data. (6) According to the equation 6, log (q e -q t ) versus t was plotted and observed that the sorption data for neither Pb 2+ nor Cu 2+ (not shown) matched with the first order kinetics (Figure 9).
The second order kinetics model was tested by plotting t/q t versus t and it was evident that the sorption of both Pb 2+ and Cu 2+ on DDA-SD perfectly followed this kinetic model (R 2 = 0.9966 to 0.9998). This means that the sorption rate depends on the metal ion

Regeneration
In this study, regeneration of the sorbent was carried out by soaking 20 mg of exhausted DDA-SD in 20 mL 0.1M HNO 3 or 0.1M HCl and leaving it for 30 min followed by wash with DI water. The desorption ratio (DR) was calculated as follows.
After 30 min the desorption ratio was 0.97 for both the acid solutions used for regeneration purpose. Further increase in time decreased the desorption ratio, which may be due to the re-adsorption of the metal ions on regenerated DDA-SD.
The regenerated DDA-SD was reused for sorption of Pb 2+ and it was found that in case of 34.0ppm solution the removal efficiency of regenerated DDA-SD was 30.4%, 29% and 27 % for 1 st , 2 nd and 3 rd recycle while removal efficiency of fresh DDA-SD was 49.5%. The decrease of removal efficiency of the regenerated DDA-SD is attributed to the reduction of active site due to protonation of the NH functional group in the acidic solution, which is assumed to be responsible for metal ion sorption.
In the present investigation sawdust of Albizia lebbeck was modified chemically and the sorption characteristics of the material were studied under equilibrium conditions using Pb 2+ and Cu 2+ solutions. The optimum pH for the sorption of Pb 2+ and Cu 2+ were 5.9 and 7.1 respectively. The sorption data fitted well to Langmuir model and did not fit Freundlich model at all. The maximum sorption capacities of Pb 2+ and Cu 2+ on DDA-SD were 17.37 and 9.4 mg/g respectively. The sorption data fitted well with the second order kinetic model. The exhausted DDA-SD was regenerated by treating with either 0.1M HNO 3 or 0.1M HCl followed by washing with DI water and the regenerated DDA-SD showed acceptable sorption capacities.

ACKNOWLEDGEMENT
The research was suppor ted by the Commonwealth Scholarship Commission in the UK. Thanks to the Ministry of Science and Technology, Bangladesh for financial support to one of the researchers. The authors are grateful to the Department of Applied Chemistry and Chemical Engineering, University of Dhaka, Bangladesh and the Division of Materials, Mechanics and Structures, Faculty of Engineering, University of Nottingham, UK.

Conflict of Interest
The authors declare that there is no conflict of interest.