2.1 MATERIALS

KC used was collected from the Redemption Camp at Mowe, Ogun state, Nigeria. After collection, particulate matters such as stones and other metal particles were separated from KC which was further purified using method described by Moore and Reynolds. All chemicals used in this study were purchased from Sigma-Aldrich (Brazil). Experimental solutions containing heavy metal ions were prepared by dissolving appropriate amount of nitrates (Cd(NO₃)₂ and Pb(NO₃)₂) in deionized water to desired concentrations.

2.2 MODIFICATION OF KC

Mixture of chloroacetic acid (0.06 mole), thionylchloride (0.08 mole) and chloroform (50 mL) was heated to 75^oC for 30 min to form chloroacetylchloride (Scheme 1). A 20 g of KC was added to the chloroacetylchloride (after the removal of excess thionylchloride under reduced pressure) which was allowed to react for 6 h under constant stirring at 90^oC and finally cooled in ice. Ethylenediamine (100 mL) was added to the solid mass left and stirred for 12 h at 100^oC under reflux. The unreacted ethylenediamine was gently removed; cold deionized water was added to the product left and centrifuged thrice for 10 min at 8,500 rpm to remove excess ethylenediamine. The MKC obtained was oven dried at 60^oC for 48 h (Scheme 2) with about 98% yield.

Scheme 1 Synthesis of chloroacetylchloride. 

Scheme 2 Preparation of MKC. 

2.3 CHARACTERIZATION

The functional groups in KC and MKC were determined using FTIR (Perkin Elmer, spectrum RXI 83303). KC and MKC were blended with KBr, pressed into pellets and analyzed in the range of 400 - 4500 cm⁻¹.

Elemental analysis was achieved using Perkin Elmer series II CHNS/O analyzer (Perkin Elmer, 2400, USA) while information of the surface morphology was obtained using SEM-EDS (FEI quanta 200 3D, dual beam FEG, model EDAX EDS brand, Bruker) with powdered samples being coated with gold using the sputtering technique in order to increase electrical conductivity and the quality of the micrographs. The surface area of KC and MKC was determined by nitrogen adsorption using the BET method in a Quantachrome Autosorb IQ₂ instrument. X-ray diffractometer (XRD-7000X-Ray diffractometer, Shimadzu) with filtered Cu Kα radiation operated at 40 kV and 40 mA was used to obtain the X-ray diffraction pattern. The XRD pattern was recorded from 10 to 80° (2θ), with a scanning speed of 2.00°/min. Particle size distribution and zeta potential were recorded using a zeta potential analyzer (DT1200, Dispersion technology) at 25^oC while observing general calculation model for irregular particles. Several measurements were taken using Dispersion technology-AcoustoPhor Zeta size 1201 software (version 5.6.16).

2.4 EQUILIBRIUM STUDY

Batch adsorption equilibrium study was carried out by contacting 0.5 g of KC and MKC separately with 250 mL varying concentration of Cd²⁺ and Pb²⁺ (10 - 100 mg/L) solutions in different 500 mL beaker at 298 K and 200 rpm for 3 h. Several agitations at 298 K and 200 rpm were repeated in order to establish the equilibrium time. Equilibrium concentration of Cd and Pb were determined by withdrawing clear samples at different interval and analyzing using Atomic Absorption Spectrometer (Varian AA240FS).

2.5 EFFECT OF KC AND MKC WEIGHT ON ADSORPTION OF Cd ²⁺ AND Pb ²⁺ IONS

Effect of KC and MKC weight was evaluated by varying the weight of KC and MKC from 0.1-1.0 g in 250 mL of 100 mg/L solution for 3 h. Concentrations of Cd²⁺ and Pb²⁺ were established after several equilibrium studies. Clear supernatant were withdrawn at different time interval and analyzed using Atomic Absorption Spectrometer (Varian AA240FS).

2.6 EFFECT OF PH ON ADSORPTION OF Cd ²⁺ AND Pb ²⁺ IONS BY KC AND MKC

A 0.5 g of KC and MKC was separately agitated with 100 mg/L solution of Cd²⁺ and Pb²⁺ ions. This was adjusted over a pH range of 1.70 - 6.60 using 0.1 M HCl and 0.1 M NaOH and stirred at 200 rpm in a 500 mL beaker for 3 h. Clear samples were withdrawn at different intervals and analyzed using Atomic Absorption Spectrometer (Varian AA240FS).

2.7 THERMODYNAMIC STUDY

Thermodynamics of the adsorption of Cd²⁺ and Pb²⁺ on KC and MKC was carried out at different initial concentrations (10 to 100 mg/L) and temperatures ranging from 298 to 333 K. The solutions were stirred at 200 rpm in a 500 mL beaker for 3 h. Clear samples were withdrawn at different intervals and analyzed using Atomic Absorption Spectrometer (Varian AA240FS).

3.1 CHARACTERIZATION

The XRD results in Fig. 1 reveals prominent peaks at 7.19 Å, 4.11 Å, 3.58 Å, 3.35 Å, 2.98 Å, 2.29Å, 1.82 Å, and 1.49 Å which correspond to kaolinite, quartz, tridymite, kaolinite, quartz, quartz + alunite, quart and kaolinite.

Fig. 1 XRD of KC and MKC 

Both spectra of KC and MKC looks similar suggesting that the d-spacing of the crystal lattice of KC was not significantly altered as similar peaks were found in both KC and MKC. This observation suggests the modification to be a surface phenomenon.

Structure of kaolinite minerals is expected to have a sheet of corner-sharing tetrahedral linked with a plane of oxygen and hydroxyl groups which are referred to as inner hydroxyls (^{Saikia & Parthasarathy, 2010}). The OH stretching for kaolinite were seen in both KC and MKC at 3697, 3620 and 3452 cm^-1 as reported by Farmer (1974); this is presented in Fig. 2. Band at around 3620 cm^-1 was attributed to the inner hydroxyls while band at 3697 cm^-1 was considered to be vibrational frequencies of the external hydroxyls. The band observed at 3452 cm^-1 in KC was taken to be the OH vibrational mode for natural hydrous silicates (^{Saikia & Parthasarathy, 2010}). This peak at 3452 cm^-1 in KC became broader in MKC which appeared at 3439 cm^-1 suggesting an overlap of the hydroxyl peak with the amine peak formed in MKC. The peaks at 1033 cm^-1 in KC and MKC may be due to the interference from alunite and or quartz while peak at 914 cm^-1 could be due to Al-OH vibrational bending. The doublet at 792-777 cm^-1 in KC and MKC may be due to Si-O-Si inter tetrahedral bridging bonds (^{McMillan, Wolf, & Poe, 1992}). The Al-O-Si and Si-O-Si deformations were seen at 536 and 466 cm^-1; respectively while the Si-O frequency was seen at 694 cm^-1 in both KC and MKC. The CHN analysis results showed the hydrogen content to increase from 2.90 % in KC to 3.96% in MKC. Carbon and nitrogen were found to be 8.88 and 1.20 %; respectively in MKC. The increase in hydrogen content and the presence of carbon and nitrogen in MKC suggests that modification had taken place.

Fig. 2 FTIR results of KC and MKC. 

The surface morphology and EDS results of KC and MKC are presented in Fig. 3 and Fig. 4; respectively. The surface morphology of KC is different from that of MKC according to the SEM images in Fig. 3. The flaky-like crystal and fluffy appearance of KC was seen in the micrograph while MKC appears to be fluffier with a more consistent texture than KC. The EDX results for KC and MKC are presented in Fig. 4 which reveals their composition. The BET surface area reveals the surface area of KC to be 16 m²/g while that of MKC was found to be 13 m²/g. The particle size distribution is shown in Fig. 5; this was found to be bimodal for KC and monomodal for MKC. The zeta potential for KC and MKC are presented in Fig. 6. As previously reported (^{Adewuyi & Pereira, 2017}), values obtained for zeta potential helped in quantifying the charges at the surfaces of KC and MKC which suggested the degree of stability of KC and MKC and their electrostatic repulsion with other charged particles (^{Hanaor, Michelazzi, Leonelli, & Sorrell, 2012}). In this study, the zeta potential reduced as pH values increased below the pH of 6. In the case of KC, the zeta potential increased as pH increased above 6 but later dropped at pH above 9. In MKC, the zeta potential only slightly increased above pH 6 but dropped immediately. As pH further increased, the zeta potential of MKC kept reducing until pH of 12 when the zeta potential began to increase again. This observation may be due to the presence of the amine groups at the surface of MKC. The distribution of the zeta potential at various pH along the surfaces of KC and MKC has revealed that both KC and MKC are stable in liquid medium, mainly where solid-liquid interaction exists. Moreover, materials with high zeta potential (negative or positive) are known to be better electrically stabilized than materials with low zeta potentials which tend to coagulate or flocculate (^{Hanaor et al., 2012}).

Fig. 3 SEM of KC and MKC 

Fig. 4 EDX of KC and MKC 

Fig. 5 Particle size distribution of KC and MKC 

Fig. 6 Zeta potential of KC and MKC. 

3.2 ADSORPTION

The effect of contact time on the amount of Cd²⁺ and Pb²⁺ ions adsorbed is presented in Fig. 7. The adsorption capacity was seen to increase with time. The amount of Cd²⁺ and Pb²⁺ ions adsorbed in mg/g by KC and MKC were calculated using equation:

Fig. 7 Adsorption capacity of KC and MKC towards Cd and Pb at different concentration (25-100 mg/L) and time (298 K). 

Where q_e is the amount adsorbed in mg/g,Co and Ce are initial and final concentrations (mg/L) of adsorbates in solution; respectively while V and M are volumes (L) of Cd²⁺ and Pb²⁺ ions solution and weight (g) of KC and MKC used. The data obtained from the study were further evaluated for pseudo-first-order, pseudo-second-order, Elovich, intra-particle and liquid-film diffusion models.

Pseudo-first-order model was given as:

On rearrangement Eq. 2 becomes

The linearized form of the kinetic rate expression for a pseudo-second-order model is given as:

where q_e is the amount of Pb²⁺ or Cd²⁺ adsorbed at equilibrium (mg/g), q_t is the amount of Pb²⁺ or Cd²⁺ (mg/g) adsorbed at time t (min) and k₁ (min^-1) and k₂ (g/mg/min) are the rate constants of the pseudo-first-order and pseudo-second-order models, respectively; for sorption of Pb²⁺ and Cd²⁺ ions. Plot of log (q_e-q_t) against time t, gave a linear relationship for pseudo-first-order model from which k₁ and equilibrium sorption capacity (q_e), were calculated from the slope and intercept respectively. For the pseudo-second-order, the parameters h_o which is the initial sorption rate and k₂ were determined from the slope and intercept of a plot of tqt   against t.

Table 1 compares the K_ad, h_o, qe and r² associated with pseudo-first-order and pseudo-second-order equations for sorption of Cd²⁺ ions on KC and MKC. K_ad, h_o and q_e increased with increase in concentration of Cd²⁺ ions. In both KC and MKC pseudo-second-order exhibited a better r² value indicating that the order of the adsorption of Cd²⁺ on KC and MKC follows a pseudo-second-order. Moreover, the qe, calculated values for pseudo-second-order agreed better with the experimental qe values which further supported the claim that the present adsorption process follows pseudo-second-order.

The adsorption capacity of KC and MKC was found higher for Pb²⁺ than Cd²⁺ which may be due to the higher ionic radius of Pb²⁺ (119 pm) than that of Cd²⁺ (95 pm).

Since the ionic radius of Pb²⁺ is higher, the outermost electrons are more freely available for interaction with KC and MKC unlike in the case of Cd²⁺ where the inward pull from the nucleus on the outermost electrons is higher. The adsorption capacity for Cd²⁺ ions increased from 8.01 mg/g in KC to 24.41 mg/g in MKC. The present adsorption capacity by MKC towards Cd²⁺ is higher than values reported for Trichoderma viride biomass (^{Singh et al., 2010}), modified corn stalk (^{Zheng, Dang, Yi, & Zhang, 2010}), and Lentinus edodes (^{Zhang, Zeng, Ma, He, & Falandysz, 2012}) but lower than value reported by ^{Yuan et al (2016)} for biogenic selenium nanoparticles (59.7 mg/g). The results for Pb²⁺ ions are presented in Table 2. The q_e and h_o increased with increase in concentrations. The r² values were also better for pseudo-second-order than pseudo-first-order.

The adsorption capacity also increased from 24.75 mg/g in KC to 36.41 mg/g in MKC. The value obtained for MKC towards Pb²⁺ is higher than values reported for Walnut (^{Bulut & Tez, 2007}), CuO nanostructures (^{Farghali, Bahgat, Allah, & Khedr, 2013}) and Sugar cane saw dust (^{Putra et al., 2014}) but lower than that of Cassia fistula seed carbon (^{Senniappan, Palanisamy, Shanmugam, & Gobalsamy, 2017}). This observed increase in the adsorption capacity after functionalization may be attributed to the presence of the amide function group. This suggests that the amide group may have participated in the adsorption process via cation exchange mechanism as proposed in Scheme 3. It was observed that the BET surface area reduced after surface functionalization whereas the adsorption capacity towards Cd²⁺ and Pb²⁺ increased. This observation revealed that the surface functionalization of KC played an important role in the adsorption process. However, the surface functionalization was effective enough to cover up for the reduction in BET surface area which is about 18.75% (from 16 m²/g to 13 m²/g).

Moreover, the reduction in surface area may not have seriously affected or reduced the amounts of specific active sites on KC which were involved in the adsorption process.

The data were further subjected to the Elovich model equation has expressed below:

Scheme 3 Proposed adsorption mechanism of Cd²⁺ and Pb²⁺ on MKC. 

where α is the initial adsorption rate (mg/g min) and β is the adsorption constant (g/mg). A plot of q_t versus ln t gave a linear trace with a slope of (1/β) and an intercept of 1/β ln(αβ). The values of α and β are presented in Table 3.

The diffusion distribution of Cd²⁺ and Pb²⁺ ions on KC and MKC was evaluated using intraparticle diffusion model; this was considered in term of the total surface area of KC and MKC which may be taken to be made up of both external and internal surfaces. The intraparticle diffusion model was estimated as described by Srivastava et al (^{Srivastava, Swamy, Mall, Prasad, & Mishra, 2006}):

From equation 6, k_id is the intraparticle diffusion rate constant (mg/g/min^0.5) and C (mg/g) is a constant that describes the thickness of the boundary layer. On plotting q_t versus T^0.5, a linear plot was obtained indicating that sorption process was controlled by intra-particle diffusion. The respective K_id and r² values are shown in Table 3. Liquid film diffusion model was also used to study the movement of Cd²⁺ and Pb²⁺ ions from the liquid phase up to the solid phase boundary during the adsorption process.

Where F is defined as:

K_id is the adsorption rate constant and t is time. A linear plot of -In (1-F) versus t was obtained which suggests that the kinetics of the sorption process may also be influenced by diffusion through the liquid surrounding KC and MKC (^{Ekpete, Horsfall & Spiff, 2012a}). The effect of adsorbent weight on the sorption of Cd²⁺ and Pb²⁺ ions are presented in Fig. 8; it was noticed that the adsorption capacity of KC and MKC decreased as their weight increases.

Fig. 9 shows the effect of pH on adsorption process. It was observed that the capacity of KC and MKC to remove Cd²⁺ and Pb²⁺ ions from solution increased with increase in pH.

Fig. 8 Effect of weight of KC and MKC on adsorption of CD²⁺ and Pb²⁺ from aqueous solution (100 mg/g) at 298 K. 

Fig. 9 Effect os solution (100 mg/L) pH on the sorption of Cd²⁺ and Pb²⁺ ions by KC and MKC at 298 K. 

3.3 ADSORPTION ISOTHERM

It is paramount to establish the most suitable correlation for the equilibrium adsorption curve for the sorption of Cd²⁺ and Pb²⁺ ions on KC and MKC. To this effect, data obtained were subjected to some isotherm equations, namely; Langmuir, Freundlich and Temkin. Of the three equations, Freundlich and Temkin fitted well for the isotherm plots. The Freundlich isotherm is characteristic of heterogeneous surface, this is expressed as:

The linearized form is given as:

K _f (mg/g) is the Freundlich isotherm constant which indicates adsorption capacity, C_e (mg/L) is the equilibrium concentration of Cd²⁺ and Pb²⁺, q_e (mg/g) is the amount of Cd²⁺ and Pb²⁺ adsorbed at equilibrium and n is the adsorption intensity. Parameters K _f and n play important role in understanding the adsorption process. When n = 1, then the partition between the liquid and solid phases are independent of the concentration; If 1/n = <1, it indicates a normal adsorption but when 1/n = >1, it indicates cooperative adsorption (^{Goldberg, Tabatabai, Sparks, Al-Amoodi, & Dick, 2005}). The results obtained for the Freundlich parameters are presented in Table 4 indicating the values of 1/n for Cd²⁺ and Pb²⁺ ions to be 1 suggesting a linear adsorption isotherm.

The Temkin isotherm contains a parameter that explicitly takes the adsorbate-adsorbent interaction into consideration. This is expressed as:

Note that:

From equations 12 and 13, B (J/mol) is the constant related to heat of adsorption, A (L/g) is the Temkin isotherm equilibrium binding constant, b is the Temkin isotherm constant, R is the gas constant (8.314 J/mol/K) and T is the absolute temperature (K). The parameters were determined from the slope and intercept of a linear plot of q_e against ln C_e. The values of B, A and b are shown in Table 4. MKC had higher heat of adsorption than KC which is higher for Pb²⁺ than Cd²⁺ ions.

3.4 SORPTION THERMODYNAMICS

Experimental data obtained from the effect of temperature on the sorption of Cd²⁺ and Pb²⁺ ions on KC and MKC were analyzed to determine some thermodynamic parameters such as Gibb’s free energy change (∆G^o), enthalpy change (∆H^o) and entropy change (∆S^o). The adsorption equilibrium constant b_o was estimated from the expression (^{Ekpete, Horsfall, & Spiff, 2012b}):

where C_o and C_e are the initial and equilibrium concentrations of Cd²⁺ and Pb²⁺ ions, R is universal gas constant (8.314 J mol^-1 K^-1) and T is the absolute temperature in K. The values of ∆H^o and ∆S^o were calculated from the slope and intercept of the linear plot of ln b_o against reciprocal of temperature (1/T). The values obtained for ∆G^o was positive nature which shows that energy is required to promote the adsorption of these metal ions on KC and MKC. The values of ∆G^o for KC and MKC are presented in Table 5 and Table 6 respectively. In the case of KC the value of ∆G^o increased with increase in temperature for Cd²⁺ ions while for Pb²⁺ ions, the value obtained decrease with increase in temperature. Values of ∆H^o and ∆S^o for KC and MKC are shown in Table 7. ∆S^o is negative for the sorption of Cd²⁺ ion on KC (-0.0055 KJ mol^-1K^-1) and MKC (-0.0001 KJ mol^-1K^-1) whereas in the case of Pb²⁺ ion, the value was positive for KC (0.0016 KJ mol^-1K^-1) and MKC (0.0081 KJ mol^-1K^-1).

∆H^o was found positive for Cd²⁺ ion suggesting an endothermic process while the process was negative in the case of Pb²⁺ ion suggesting exothermic process.