Solvent-free synthesis of the cellulose-based hybrid beads for adsorption of lead ions in aqueous solutions

In this work, the adsorption of Pb 2+ onto maleic anhydride modi ﬁ ed cellulose/diatomite beads (MCDBs) was investigated. Instead of the general process for esterifying the cellulose beads, a solvent free synthesis which needs no catalyst or solvent was used. An appropriate amount of calcium carbonate was added during the formation of MCDBs to increase pore structure after being removed under an acidic condition. The synthesized adsorbent was characterized by FTIR, SEM, and BET. The degree of the carboxyl group of MCDBs was found to be 0.450 mmol g (cid:1) 1 based on colloid titration. The e ﬀ ects of pH, temperature, contact time, and the concentration of Pb 2+ on adsorption were studied in batch mode. The results indicated that the MCDBs had a good adsorption capacity toward Pb 2+ with the maximum adsorption at 44 mg g (cid:1) 1 . The experimental kinetic data ﬁ t the pseudo-second order model very well. Moreover, the adsorption process for Pb 2+ was better described by the Langmuir isotherm model. The regeneration of the MCDBs could be readily accomplished using an HCl (1 M) treatment without lowering the adsorption capacity signi ﬁ cantly.


Introduction
Environmental problems, especially aquatic environments, are a main theme of scientic research now and in recent years. The removal of toxic heavy metal ions from industrial effluents, water supplies, and mine waters has received much attention. Heavy metal ions, such as Cu, Cd, Ni, and Pb, released into the environment affect ecological life, owing to their tendency to accumulate in living organisms, and are highly toxic when absorbed into the body. 1 Heavy metals usually have long residence times. Lead toxicity has been reported to decrease kidney functions and enzymatic activities and cause neuromuscular difficulties. 2 There are also many lead contamination incidents with drinking water reported in some countries. 3 Thus, it is necessary to remove heavy metal ions from wastewater. 4 Various adsorbent materials have been explored for the removal of heavy metal ions from aqueous solutions. Of all the adsorbent materials, activated carbon has been widely used for the removal of heavy metals at trace levels. 5 However, activated carbon is costly as an absorbent in water treatment. Therefore, the production of low-cost materials, such as cellulose and diatomite, has been a hotspot of research. Cellulose, as one of the natural and green materials, constitutes the most abundant natural polymer resource. However, the adsorption capacity is limited, and the selectivity is low when directly using natural cellulose as an adsorbent. Because there are many hydroxyl groups in the polymer structure, 6 hydrogen bonds form between the molecular chains. 7 Chemical modication, such as esteri-cation, halogenation, oxidation, and etherication, can be used to vary certain properties of cellulose, such as its hydrophilic or hydrophobic character, elasticity, water sorbency, adsorptive or ion exchange capability. 8 Magali et al. 9 synthesized new adsorbents containing carboxyl groups by graing acrylic acid onto sawdust, which was used for the sorption of Cu 2+ from aqueous solution. James et al. 10 milled aspen wood, then the Milled aspen wood was thermochemically modied with citric acid to improve the copper (Cu 2+ ) ion sorption capacity of the wood. Nada et al. 11 modied bagasse bres using three different chemical methods to remove heavy metal ions from waste water. Very recently, Wang et al. 12 reported an eco-friendly sugarcane cellulose-based adsorbent with very high sorption capacities towards Pb 2+ , Cu 2+ and Zn 2+ . The application of such adsorbents was also extended to binary component systems; and the adsorption behaviour was found to be well described by the competitive Langmuir isotherm model. Zhou et al. 13 synthesized a low-cost absorbent by a cost-effective chemical modication for dyes and heavy metal ion removal. Yu et al. 14 prepared cellulose nanocrystals (CNCs) from cotton, and then the CNCs were chemically modied with succinic anhydride to introduce carboxyl groups on the adsorbents. However, a pyridine reux was used during the chemical modication, which was not conducive to the development of the environment.
Diatomite, a so lightweight rock available in large deposits around the world, is a highly porous structure and low-cost material. 15 Attributed to its physical and chemical properties, it has been used as an adsorbent in wastewater treatment. 16 Al-Degs et al. 17 improved the intrinsic exchange properties by modication with manganese oxides, and the resulting adsorbent showed a high tendency for adsorbing lead ions from solution. Khraisheh et al. 18 modied diatomite by NaOH solution and manganese oxide to improve the adsorption capacity of the diatomite for the removal of heavy metals. However, the powdered diatomite and cellulose might lead to difficulty in the application of the adsorbents. Cellulose-based beads facilitate the easy separation of hybrid adsorbents in batch operation. Yang et al. 19 prepared carboxymethyl cellulose hydrogel beads using epichlorohydrin (ECH) as a crosslinking agent and found that the additional carboxyl groups afforded a higher sorption capacity to metal ions as well. Recently, Yu et al. 20 successfully prepared porous magnetic cellulose beads via a one-pot synthesis, which created a new platform to prepare the cellulose-based functional nanocomposites.
Thus, the aim of the present study was to investigate modi-ed cellulose beads in conjunction with alkali-treated diatomite (MCDBs) as a low-cost material for the removal of Pb ions from aqueous solutions. To enhance the absorption capacity of bres and to facilitate the recovery of adsorbent, the cellulose bres were dissolved and regenerated, and then reacted with maleic anhydride. During the formation of adsorbent particles, calcium carbonate was added to further increase the porosity of the adsorbent aer being removed under acidic conditions. The effects of solution pH, temperature, initial concentration, and contact time during the adsorption processes were also evaluated.

Materials
The cellulose raw material was from lter paper (Qualitative P4, porosity: Medium-Fine; Flow Rate: Slow, Fisher Scientic). Diatomite, hydroxylamine hydrochloride, dithizone, Pb(NO 3 ) 2 , CaCO 3 powder (#30 mm), maleic anhydride, and urea were all purchased from Sigma-Aldrich, Canada. NaOH is powder purchased from Fisher Scientic, and, NaOH (0.01 M) and HCl (0.01 M) are solution purchased from Fisher Scientic. Ammonium citrate tribasic and potassium sodium L-tartrate tetrahydrate were obtained from Alfa Aesar. The standard Pb (1.000 g L À1 ) solution was prepared by dissolving appropriate amounts of Pb (NO 3 ) 2 in distilled water.

Preparation of cellulose/diatomite beads (CDBs)
The diatomite under a 150 mesh was washed with distilled water to remove nes and other adhered impurities, and then desiccated. 21 Chemical modication was accomplished by treating the diatomite with NaOH (5 mol L À1 ): 150 g diatomite was mixed with 1 L NaOH, then stirred at 100 C for 2 h. The resulting samples were washed with distilled water, and desiccated in an oven at 100 C, then stored in glass bottles.
Cellulose solution (4%) was prepared by dissolving lter paper with NaOH (7 wt%)/urea (12 wt%) at À10 C; homogenizer (Stanfen, Germany) was used during this process. Then, the treated diatomite and CaCO 3 were added at different weight ratios. The mixtures were dropped into HCl (1 mol L À1 ) using a syringe aer being dispersed by the homogenizer. The cellulose/diatomite beads (CDBs) were then washed with distilled water to remove CaCl 2 aer the reaction between CaCO 3 and HCl was accomplished and transferred into acetone immediately. Aer 24 h, the CDBs were desiccated in a vacuum oven (Fig. 1). 22 The synthesized absorbents were named MCDBs-0, MCDBs-10, MCDBs-20, MCDBs-30, and MCDBs-40 according to the different amount of CaCO 3 .

Preparation of maleic anhydride-modied CDBs (MCDBs)
The MCDBs (5 g) were treated with maleic anhydride dissolved in acetone (5 g/40 mL acetone), then the acetone was evaporated at 50 C in a vacuum oven for 1 h and the residual mixture was heated in an oven at 100 C for 3 h. The products were washed with distilled water, ethanol, and nally with acetone, and then desiccated in a vacuum oven (Fig. 1). 11 The degree of substitution (DS) of hydroxyl groups in the cellulose due to esterication was determined.

Degree substitution of hydroxyl groups (charge density)
The surface charge density (degree of substitution of the hydroxyl groups) was determined by colloid titration until zero charge point by a particle charge detector (Mütek PCD-03, Germany). In the test, 0.05 g of MCDBs and 0.01 N poly-DADMAC solution (5 mL) were added in the measuring cell at room temperature with constant stirring for 10 min. Then, titration analysis was conducted for both control and bre samples by adding the PVSK (0.01 N) into the mixture automatically. The point of zero charge detected by the Mutek titrator corresponded to the amount (i.e. volume) of PVSK neutralizing the excessive amount of PDADMAC. 23 The experiment was carried out three times simultaneously. The Fig. 1 The solvent-free approach to preparing the cellulose hybrid beads.
concentration of carboxylic functions [C COOH (mmol g À1 )] was calculated by eqn (1) where V 2 is the volume of PVSK dropped into the solution with MCDBs (mL), V 1 is the volume of PVSK dropped into the blank solution (mL), C is the concentration of PVSK and poly-Dadmac, and mMCDBs is the mass of the MCDBs (g).

Characterization
FT-IR spectra were recorded using an NEXUS 470 spectrophotometer (Nicolet Thermo Instruments, Canada) aer the MCDBs were ground with KBr. BET adsorption was carried out using an Autosorb instrument (Belsorp-Max BEL Inc, Osaka, Japan). Scanning electron microscopy (SEM) was performed on a JEOL JSM-6400 SEM instrument (Japan).

Batch experiments
Pb 2+ was chosen to evaluate the effectiveness of synthesized absorbents for water treatment. All the solutions with various concentrations were obtained by successive dilution. The pH was adjusted by adding either 0.01 M HCl or 0.01 M NaOH. Adsorption experiments were carried out using 10 mL of lead ion solution at the desired concentration (10 mg L À1 ) at an initial pH of 7.0, and an adsorbent dosage of 0.01 g per 10 mL in an agitation speed of 130 rpm on a temperature-controlled shaker (SWB25, Thermo Electron Corporation, Germany). In the preliminary experiment, this speed was found to be suitable to reach equilibrium. The shaking time, temperature, pH, and the concentration of CaCO 3 were investigated, respectively. Residual Pb 2+ concentration in the ltrate was determined using UV (Genesys 10-s, Thermo Electron Corporation). 24 The percent metal ion removal R (%) was calculated by eqn (2). 25 where C 1 (mg L À1 ) and C e (mg L À1 ) were the initial and nal concentrations of the metal ions, respectively. The adsorption capacity (Q) was calculated by eqn (3).
where Q is the adsorption capacity (mg g À1 ), V is the volume of solution (L) and m is the mass of adsorbent (g) used.
2.6.1 Adsorption kinetics. The pseudo-second-order kinetic model was used to describe the adsorption kinetic data of Pb 2+ measured on MCDBs. A non-linear form and a linear form are presented in eqn (4) and (5), respectively. 26 where q t (mg g À1 ) and q e (mg g À1 ) are the amounts of the Pb 2+ adsorbed at time t (min) and at equilibrium, respectively. K 1 (g mg À1 min À1 ) is the rate constant of the adsorption process.
2.6.2 Adsorption isotherms. The isotherms were carried out by shaking the MCDBs (0.01 g) with 10 mL of metal ion solution at different initial concentrations. The corresponding adsorption isotherms for lead ions were described by tting the experimental data to the Langmuir, Freundlich, 27 and Temkin isotherms, 28 respectively. The Langmuir equation used was the linear form eqn (6): 29 C e /q e ¼ aC e /K L + 1/K L where q e and C e are the solution (mg g À1 ) and surface concentrations (mg L À1 ) for the adsorbate at the equilibrium, respectively, and K L (L g À1 ) and a (L mg À1 ) are the isotherm constants. K L can be obtained from the relationship between C e /q e and C e . The constant, a, corresponds to the energy of the adsorption process. A dimensionless separation factor R L is the essential characteristic of the Langmuir equation, which is described by eqn (7). 30 where C 0 is the highest initial Pb 2+ concentration (mg L À1 ), the value of R L indicates the nature of the interaction and the isotherm type: unfavourable (R L > 1), linear (R L ¼ 1), favourable (0 < R L < 1), or irreversible (R L ¼ 0).

FTIR spectra and colloid titration
The FTIR spectra of CDBs before and aer modication with maleic anhydride are shown in Fig. 2 -OH stretching adsorption bands at 3442 cm À1 do not signicantly change aer modication. Because only the surface hydroxyls that are available, maleic anhydride can be graed during the chemical reaction. 34 According to the results obtained by colloid titration, the concentration of carboxyl groups on the surface of the MCDBs was 0.45 mmol g À1 , which corresponds to the surface charge density of MCDBs.

SEM
The surface structure of cellulose beads obtained was revealed by SEM and the images are shown in Fig. 3. The CDBs obtained exhibited good spherical shape together with a porous structure ( Fig. 3(a)). More pores can be observed in Fig. 3(c) and (e) than in Fig. 3(b). The application of CaCO 3 was conducive to the pore structure. The pores may be formed from the reaction between CaCO 3 and HCl, during which the CaCO 3 rich regions were transformed into pores due to the release of carbon dioxide and dissolution of CaCO 3 . The alkali-treated diatomite exists in CDBs independently (Fig. 3(d)), which can preserve the adsorption capacity of adsorbate. Aer modication, maleic anhydride was graed onto cellulose, which leads to decreasing pores in CDBs (Fig. 3(c) and (f)), due to the graed maleic anhydride which might ll up some pores on the spheres (Fig. 3(g) and (h)). This can be proved with the results of BET.

Brunauer-Emmett-Teller (BET) adsorption
The specic surface areas of the MCDBs were evaluated by BET analysis; and the results are presented in Table 1. The specic surface area of MCDBs increased signicantly when the concentration of CaCO 3 changed from 10% to 30%, at the same time, the total pore volume was changed from 5.78 Â 10 À3 to 9.42 Â 10 À3 m 3 g À1 , which indicates the application of CaCO 3 is conducive to the pore structure. Aer modication with maleic anhydride, the specic surface area of MCDBs reduced to 0.73 m 2 g À1 , meanwhile, the total pore volume changed from 9.42 Â 10 À3 to 4.89 Â 10 À3 m 3 g À1 , implying the graed maleic anhydride might ll up some pores. However, the adsorption capacity of MCDBs (30% CaCO 3 ) is much lower than that MCDBs-30 (aer modication), which means the esterication is the main reason for the increase in the adsorption capacity of Pb 2+ onto MCDBs. MCDBs-40 does not have any results from BET, which might be due to the collapse of the pores. In addition, the specic surface area of MCDBs-0 could not be detected by BET, which indicate that MCDBs-0 had very limited or no micro or mesoporous structures.

Comparative test
The Pb 2+ removal rate (%) of different absorbents, including raw materials, CDBs and MCDBs, are shown in Fig. 4. The adsorption capacity of raw materials is extremely low, which is 44.95% for lter paper and 9.30% for alkali-treated diatomite, respectively. The CDBs have a higher removal rate compared with the raw materials. Furthermore, the adsorption capacity of Pb 2+ signicantly increased aer the modication by maleic anhydride. These results demonstrated that the MCDBs are good carriers or adsorbents toward Pb 2+ . The effect of CaCO 3 concentration was studied at room temperature and neutral pH, and the results are also shown in Fig. 4. The removal rate (%) increased signicantly with the initial concentration of CaCO 3 increase from 0% to 30%, and the removal rate was negatively affected when further increasing the concentration of CaCO 3 up to 40%, because the CaCO 3 at an over-dosage can induce the collapse of CDB pores. Therefore, CaCO 3 play a role in the pore structure; the optimum concentration of CaCO 3 is 30%.

Effect of pH
The removal of Pb 2+ from aqueous solutions by adsorption depends on the pH, which affects the ionization of metal ions and the concentration of counter H + ions on the surface of the adsorbent. To study the effect of H + concentration on metal  MCDBs-0 (before modication) -1.1x10 À3 MCDBs-10 (before modication) 1.00 5.78x10 À3 MCDBs-20 (before modication) 1.21 6.64x10 À3 MCDBs-30 (before modication) 1.53 9.42x10 À3 MCDBs-30 (aer modication) 0.73 4.89x10 À3 MCDBs-40 (before modication) -6.23x10 À3 removal, the solution pH was varied from 2 to 6. The results are shown in Fig. 5. The optimum pH value for the adsorption of Pb 2+ by MCDBs was found to be pH 6.0 (Fig. 5). For MCDBs-30, the percentage removal rate is observed to be 88.89% for Pb 2+ adsorption at this pH. In aqueous solutions, Pb 2+ is dominant at pH 2-6. Other species (PbOH + ) can exist at a higher pH. The removal rate of Pb 2+ is extremely low at a pH below 3. The higher the concentration of H 3 O + , which means a lower pH, the higher the concentration of protons competing with metal ions for the active sites. The surface of absorbents is positively charged, and it is hard to adsorb heavy metals with a positive charge. 35,36 With increasing pH (pH > 3), the concentration of protons decreased, and the surface is negatively charged, which is advantageous for adsorption.

Effect of temperature
The effects of temperature on the adsorption of Pb 2+ were studied by changing the temperature from 15 to 50 C under the optimum pH in 2 h; the results are presented in Fig. 6. The adsorption of Pb 2+ increased signicantly as the temperature changed from 20 to 30 C, which indicates that the adsorption process on MCDBs is an endothermic process. It changes indistinctively aer 30 C. So, the optimum temperature of Pb 2+ adsorption on MCDBs is selected to be 30 C. Experiments were performed at different temperatures (20,30 and 45 C) at pH 6. During short contact times, the adsorption process is very fast (within 10 min more than 50% is adsorbed) due to the numerous active sites on the surface of MCDBs-30; then, the removal rate levelled off. Over 84% adsorption efficiency is obtained at a certain temperature, indicating that Pb 2+ is absorbed by available adsorption sites. The sorption of Pb 2+ onto MCDBs-30 is mainly mediated through complexation with surface functional groups (especially carboxyl groups) which can react with Pb 2+ to form bound complexes. 37 The adsorption process for Pb 2+ is described by a pseudosecond-order model, where R 2 is 0.9959, 0.9950, and 0.9981 at the temperatures of 20, 30 and 45 C, respectively (Fig. 8). The tting results suggested that the adsorption rate was dominated by chemical adsorption, which involved the electron sharing or electron transfer between the adsorbent and the adsorbate. The q e values for Pb 2+ adsorption are in good agreement with the experimental one. This indicates a good agreement that the pseudo-second-order model describes the kinetics of the adsorption. 38    3.7.2 Adsorption isotherms. The results of adsorption isotherms, calculated from different models, are shown in Table 2 and Fig. 9. A Langmuir model ts the experimental data well and the correlation coefficients of the equations (R 2 ¼ 0.9952) indicate that this model can explain the adsorption process satisfactorily. The Langmuir isotherm is monolayer sorption where the adsorption sites are the same and the absorbed particles are independent, and therefore no further sorption can take place aer a saturation value is reached. 39 In this study, the R L value is between 0 and 1, which indicates that the adsorption process is favourable. The estimated q m ¼ 46.04 mg g À1 (K L /a) is very close to the experimentally (44.54 mg g À1 ) obtained maximum metal uptake for Pb 2+ . Monolayer adsorption capacities for the adsorption of Pb 2+ onto MCDBs-30 and other materials are compared in Table 3.
In addition, the thermodynamic parameters, such as enthalpy (DH ), entropy (DS ) and Gibb's free energy (DG ), were estimated using the Van't Hoff equation. Table 4 lists the thermodynamic parameters obtained from the temperature effect on Pb 2+ adsorption, and the values of free energy changes DG are negative, conrming that the adsorption of Pb 2+ onto MCDBs is spontaneous and thermo-dynamically favourable. The DG value decreased as the temperature increased, indicating a lower driving force for the adsorption. The negative values of DH and DS indicate that the Pb 2+ adsorption on MCDBs was exothermic in nature and the stability of the adsorption process, with no structural change at solid-liquid interface.

Desorption of Pb 2+ from MCDBs-30
Desorption of the Pb 2+ was investigated. This procedure is necessary to restore the original adsorption capacity of the adsorbent and it also enables recovering valuable metals from wastewater streams. In this study, Pb 2+ was desorbed from MCDBs-30 using 1 M HCl for 24 h. Then the absorbent was washed with distilled water, ethanol, and nally acetone. The absorbent was reused three times at the same conditions (Pb 2+ concentration: 10 mg L À1 ; adsorbent: 0.01 g; volume of solution: 0.01 L), resulting in the removal rates (%) at 94.96, 90.74, and 87.13, respectively. The removal rate of Pb 2+ is substantially unchanged aer reusing three times (the third time is 87.13%), indicating that the MCDBs-30 are a renewable absorbent for Pb 2+ removal.     42 113.63 Apricot stone 43 22.85

Conclusion
Maleic anhydride modied CDBs (combine cellulose and alkalitreated diatomite) was found to effectively adsorb Pb 2+ from aqueous solutions. The pore structure of the absorbents was enhanced by introducing CaCO 3 during the preparation. Further experiment data revealed that the adsorption process of Pb 2+ on the MCDBs-30 follows a pseudo-second-order kinetics and the equilibrium data can be well tted with a Langmuir isotherm. The maximum adsorption capacities of adsorbent (Langmuir) was 44 mg g À1 from an initial concentration of 800 mg L À1 at pH 6 and at 30 C. The capacity of the adsorbents in Pb 2+ removal remained unchanged aer reusing three times. The MCDBs-30 is a green-based, cost-effective, and renewable absorbent. It could be used to adsorb more heavy metal ions which are positive charged.

Conflicts of interest
There are no conicts to declare.