Triazole containing magnetic core-silica shell nanoparticles for Pb²⁺, Cu²⁺ and Zn²⁺ removal

2.1 Materials

Ferric chloride hexahydrate (FeCl₃·6H₂O, 98%), Ferrous chloride tetrahydrate (FeCl₂·4H₂O, 99%), 3-aminopropyltriethoxysilane (H₂N(CH₂)₃Si(OC₂H₅)₃, APTES, >99%), Tetraethylortho silicate (Si(OC₂H₅)₄, TEOS > 99%), 1, N, N′-dicyclohexylcarbodiimide (DCC), 4-dimethylaminopyridine (DMAP), propargyl acid, sodium ascorbate, copper sulfate pentahydrate (CuSO₄·5H₂O), sodium hydroxide (NaOH, >98%), and hydrochloric acid (37%) were purchased from Sigma-Aldrich (Prochima Sigma, Tlemcen, Algeria).

The solvents were of commercial grade quality and were dried and distilled before use. Ethanol HPLC grade (99%), dichloromethane (DCM, 99%) and tert-Butanol (⩾99.0%) were purchased from Sigma-Aldrich (Prochima Sigma, Tlemcen, Algeria). Toluene (99%) was obtained from Biochem, Algiers.

CuCl₂·4H₂O, Pb(NO₃)₂, Zn(NO₃)₂·4H₂O were purchased from Sigma-Aldrich (Prochima Sigma, Tlemcen, Algeria) and were used as source of Cu(II), Pb(II) and Zn(II) ions. Deionized water was used for all experiments.

4-azido-1-nitrobenzene was prepared according to Nolting and Michel method (Noelting and Michel, 1893) with a yield of 79%.

2.2 Preparation of magnetite (Fe₃O₄)

The magnetite nanoparticles (MNP) were prepared by the conventional co-precipitation method corresponding to the following reaction path:

FeCl₃·6H₂O (11.68 g) and FeCl₂·4H₂O (4.30 g) were dissolved in 200 mL of deionized water under nitrogen atmosphere with vigorous stirring. The pH of the solution is increased by adding dropwise 12.5 mL of NH₄OH solution (28%) into the above mixture under continuous stirring. The color of the solution turned black immediately. The reaction mixture was heated at 80 °C for 2 h at 300 rpm. After cooling the suspension to room temperature, the resulting Fe₃O₄ nanoparticles were separated using a magnet and then washed with deionized water to remove the base excess until the pH becomes neutral. The black precipitate was then washed three times with ethanol (3 × 50 mL) and dried in vacuum.

2.3 Functionalization of magnetic nanoparticles with APTES (MNP-NH₂)

250 mg of prepared magnetite was dispersed in 100 mL of ethanol and sonicated for 15 min. The resulting dispersion was diluted with 300 mL of ethanol and then sonicated for further 30 min. 1.8 eq of APTES solution (8.62 g) was added to the dispersion under nitrogen atmosphere and the reaction mixture was stirred vigorously at room temperature for 16 h.

The resulting amino functionalized (Fe₃O₄-NH₂) nanoparticles were magnetically recovered and washed with ethanol and DCM several times and then dried under vacuum

2.4 Preparation of alkyne functionalized MNP (MNP-Alky)

300 mg of MNP-NH₂ was dispersed in 10 mL of dichloromethane using an ultrasonic bath for 10 min. (223 mg, 31.8 mmol) of propargyl acid, (220 mg, 31.8 mmol) of DCC and (13 mg, 3.18 mmol) of DMAP were then added and the mixture was magnetically stirred for 16 h under nitrogen atmosphere at room temperature. The solid product was recovered with magnetic field and washed with ethanol (8 × 10 mL), then with a mixture of H₂O/t-BuOH(1/1) (4 × 4 mL) and finally with dichloromethane (several times) to ensure the removal of the urea by-product formed during the reaction. The resulting product was dried under vacuum and stored in a desiccator.

2.5 Preparation of triazole functionalized MNP (MNP-Trz)

The catalytic 1,3-dipolar cycloaddition of alkyne functionalized magnetite with 4-Azido-1-nitrobenzene was carried out under the following conditions (Zhou et al., 2008). 800 mg of alkyne functionalized magnetite and (355 mg, 2.1621 mmol) of 4-azido-1-nitrobenzene were mixed with 6.5 mL of tBuOH/H₂O (1:1). 650 μL of freshly prepared 1 M aqueous solution of sodium ascorbate (0.65 mmol) and 217 μL of 1 M aqueous solution of copper sulfate pentahydrate (0.065 mmol) were added. The mixture was stirred at room temperature for 48 h. The final product was recovered by an external magnet and washed several times with ethanol and subsequently with deionized water and then dried.

2.6 Preparation of silica coated MNP (Fe₃O₄@SiO₂)

To protect MNP from oxidation reaction and improve their dispersion in water, they were coated with a silica layer using the alkaline hydrolysis of TEOS. In a 250 mL round-bottom flask, 100 mg of magnetite (Fe₃O₄), dispersed in a mixture of 80 mL of anhydrous ethanol and 20 mL of deionized water, was sonicated for 15 min. The pH was adjusted to pH = 9 with ammonia solution (28%) and then 200 μL of TEOS was added. The reaction was stirred for 4 h at room temperature under nitrogen atmosphere. Finally, the product was recovered with an external magnet and washed several times with de-ionized water and then with ethanol.

2.7 Preparation of functionalized silica coated MNP (MNPS-NH₂, MNPS-Alky, MNPS-Trz)

These materials were prepared by the same procedure used of Fe₃O₄ particles using as starting compound Fe₃O₄@SiO₂.

2.8 Characterization

The specific surface areas of the modified and functionalized Fe₃O₄ nanoparticles were performed by N₂-adsorption and desorption at −196 °C, using Micromeritics Gemini VII instrument. Before measurement, each sample was degassed at 200 °C for 24 h under N₂ flow. Surface areas were calculated using Brunauer-Emmett-Teller (BET) equation.

Infrared spectroscopies were carried out with a Jasco-4200 Fourier transform infrared spectrometer (Jasco Inc., Easton, MD, USA) using KBr pellets in the 4000–400 cm⁻¹ region. The dry powders of magnetic particles were blended with KBr, ground and pressed into disks for measurements.

Powder X-ray diffraction (XRD) patterns of magnetic nanoparticles were collected with a Bruker-D8 advance AXS diffractometer with CuKα radiation λ = 1.5418A° operating at 40 kV, 40 mA. The scanning rate was 2°/min and the scanning scope of θ was from 2° to 20 at room temperature.

Thermogravimetric analysis of particles was performed with a TG 209 Netzsch apparatus. Alumina crucibles were filled with accurately weighted samples of about 10 mg. The temperature program ranged from 35 °C to 813 °C at a heating rate of 10 °C/min. All experiments were conducted under nitrogen atmosphere. The weight loss was recorded as a function of temperature and time.

Size measurements of aqueous solutions of magnetic nanoparticles were done by Zetasizer nanoseries “zs” of Malvern instruments (England). DTS nano program was used for data evaluation. The following Stokes–Einstein law was used to compute the mean droplet diameter: D_H = K_BT/(3πηD), where D_H, K_B, T, η and D are the hydrodynamic diameter of the particles, Boltzmann constant, temperature in Kelvin, viscosity of the medium and diffusion coefficient, respectively. Zeta potentials of magnetic nanoparticles were determined as function of pH values using the Malvern Zetasizer via electrophoretic mobility measurements. Different pH solutions were prepared using 1 mM NaCl solution and the pH values were adjusted by 1 mM NaOH or 1 mM HCl solution using a pH meter. The samples were diluted in the corresponding solution before measuring the electrophoretic mobility.

Transmission Electron Microscopy, TEM, was performed with a Philips CM120 microscope at the “Centre Technologique des Microstructures” (CTμ) at the University of Lyon (Villeurbanne, France). Before observation, carbon grids were plunged in nanoparticles powder, and the excess was then removed. The samples were observed by TEM under 120 kV acceleration voltages.

Scanning Electron Microscopy, SEM, was performed with a FEI Quanta 250 FEG microscope at the “Centre Technologique des Microstructures” (CTμ) at the University of Lyon (Villeurbanne, France). Powder was deposited on a flat steel holder. The sample was coated under vacuum by cathodic sputtering with copper (10 nm). The samples were observed by SEM under an accelerating voltage of 15 kV.

The magnetization properties of the samples were examined by an automatic bench for magnetic measurement (ABMM) at CNRS-IRC laboratory Lyon. Magnetization of the samples was investigated by decreasing magnetic field (H) from 20,000 to 15,000 Oersted.

2.9 Adsorption and desorption experiments

The adsorption of various metal cations Cu²⁺, Zn²⁺ and Pb²⁺ on magnetic dispersions (functionalized magnetic nanoparticles MNP-Trz and silica coated homologous MNPS) was carried out on the basis of batch equilibrium technique in aqueous medium at 25 °C. In a typical experiment, 2 mg of MNP-Trz or MNPS-Trz is mixed with 2 mL of appropriate aqueous solution containing (50–600 ppm) of metal ions for 4 h at 250 rpm to reach the adsorption equilibrium. The amount of residual cation in the solution was measured with flame atomic absorption spectrometer (Perkin Elmer 700) after magnetic separation of chelated MNP from the supernatant.

The adsorption capacity q_e (mg g⁻¹) of metal ions in the presence of magnetic dispersion at equilibrium state was calculated according to Eq. (2):

For kinetic studies, the initial concentration of metal ion (Cu²⁺, Zn²⁺ and Pb²⁺) was 400 mg.L⁻¹ and the initial pH value was 5.5. The contact time was varied under a stirring rate of 250 rpm at 25 °C. The effect of pH solution on metal ion adsorption at equilibrium state was studied for each initial metal ion concentrations of 400 mg L⁻¹. The pH values varied from 2 to 6 by adjusting the initial metal ion solution with NaOH or HCl. The adsorption studies were carried out at 250 rpm for 4 h at room temperature. Duplicate experiments were carried out for each experiments and the high reproducibility was obtained.

The desorption efficiency studies were conducted with dilute HNO₃ solution concentration range of 0.1–0.2 mol L⁻¹ as desorption medium. Adsorption was firstly conducted using 20 mg of MNP-Trz in 400 mg L⁻¹ Pb²⁺ solution for 4 h at room temperature. The resulting Pb²⁺ sorbed MNP-Trz containing 136 mg g⁻¹ of lead ions was used to carry out the desorption study. 10 mL of HNO₃ solution was added to 10 mg of this material and then stirred at 250 rpm during variable time ranging from 5 min to 250 min at 25 °C. The MNP-Trz particles were collected by an external magnetic field and the concentration of Pb²⁺ ions was determined via supernatant analysis.

The desorption efficiency studies were conducted with dilute HNO₃ solution concentration range of 0.1–0.2 mol L⁻¹ as desorption medium. Adsorption was firstly conducted using 20 mg of MNP-Trz in 400 mg L⁻¹ Pb²⁺ solution for 4 h at room temperature. The resulting Pb²⁺ sorbed MNP-Trz containing 136 mg g⁻¹ of lead ions was used to carry out the desorption study. 10 mL of HNO₃ solution was added to 10 mg of this material and then stirred at 250 rpm during variable time ranging from 5 min to 250 min at 25 °C. The MNP-Trz particles were collected by an external magnetic field and the concentration of Pb²⁺ ions was determined via supernatant analysis.

The desorption efficiency DE (%) was determined using the following Eq (3):

3.1 Synthesis and characterization of functionalized MNP

The synthetic pathway used to prepare MNP functionalized with 1,2,3-triazole units is shown in Scheme 1. Firstly, magnetite nanoparticles NPs (Fe₃O₄) were synthesized by coprecipitation method from Fe²⁺ and Fe³⁺ solutions. The XRD spectrum (Fig. 1) confirms the cubic spinel structure of magnetite with six characteristic peaks 2θ = 30.3°, 35.6°, 43.1°, 53.7°, 57° and 62.7° which may be indexed as (2 2 0), (3 1 1), (4 0 0), (4 2 2), (5 1 1) and (4 4 0) respectively. These data matched well with those of literature data (Wang et al., 2010). The lattice constant was found to be 8.37 $\dot{A}$ which is of the same order of magnitude as the lattice parameter of magnetite (8.39 $\dot{A}$ ) (Mahdavian and Mirrahimi, 2010).

Close

Scheme 1

A schematic representation for preparation of modified silica coated and uncoated magnetite-triazole nanoparticles.

Export to PPT

Close

Figure 1

Powder X-ray diffraction (XRD) of magnetite Fe₃O₄.

Export to PPT

Three step synthetic strategy was used to surface modification of Fe₃O₄ NPs. Magnetite particles reacted with 3-aminopropyltriethoxysilane (APTES) and were then subsequently added to propargyl acid to introduce alkyne functions using Steglich esterification in the presence of DCC and DMAP. The 1,3-dipolar cycloaddition of the latter compound with 4-nitroarylazide derivative in click chemistry conditions affords the triazole-based NPs (MNP-Trz).

Silica coated magnetite NPs have been also prepared in this work. Such encapsulation should ensure better stability of Fe₃O₄ core toward chemical attacks especially oxidation reactions and favors dispersion of nanoparticles in aqueous solution. The silica shell was introduced by the alkaline hydrolysis of TEOS with ammonia solution followed by functionalization using the same procedure used for Fe₃O₄ NPs.

FTIR analysis was used to confirm the surface modification of MNP for the different synthesis steps (Fig. 2). Magnetite spectrum reveals an intense peak at 569 cm⁻¹ related to Fe—O vibrations. The occurrence of the characteristic peaks of primary amine group at 3427 cm⁻¹ and 1631 cm⁻¹, of C—H stretching vibrations at 2921 cm⁻¹ and 2852 cm⁻¹ and of Si—O stretching at 1027 cm⁻¹ indicates that the APTES has been successfully coupled to the surface of magnetite particles. The spectrum of MNP-Alky shows the appearance of peaks related to the newly formed alkyne and amide functions at 2121 cm⁻¹ (≡C—H stretching) and 1650 cm⁻¹ (C⚌O stretching). After the click reaction, the vibration band related to ≡C—H group at 2121 cm⁻¹ disappeared and several new bands appear between 1374 and 1670 cm⁻¹ indicating the presence of 1,2,3-triazole heterocycle in particular at 1393 cm⁻¹, 1563 cm⁻¹ and 1529 cm⁻¹ related to C—N, C⚌C and N⚌N vibrations. Similar IR characteristics were obtained for MNPS series (Fig. S1, supplementary file). The silica-modification was confirmed in Fe₃O₄@SiO₂ spectrum by the presence of 3 bands at 1056, 794 and 973 cm⁻¹ corresponding to the stretching vibrations of Si—O—Si, Si—OH and Si—O—Fe respectively.

Close

Figure 2

FTIR spectra of (a) Fe₃O₄, (b) MNP-NH₂, (c) MNP-Alky, (d) MNP- Trz magnetic nanoparticles.

Export to PPT

Thermogravimetric analysis was used to estimate the degree of modification for intermediate and final magnetic materials. The relative comparison between weight losses of each sample with the previous one indicates the progress of modification (Fig. 3). The weight loss of Fe₃O₄ MNP in the investigated thermal range was found to be around 4.39% and involves the evaporation of physically and chemically adsorbed water. It occurs on 3 steps: A first weigh loss of 2.15% was observed below 125.5 °C related to the evaporation of adsorbed water followed by a weight gain between 125.5 °C and 151 °C which was also previously reported by Caruntu et al. (2004) and Cao et al. (2009) and was attributed to the oxidation of magnetite Fe₃O₄ to maghemite Fe₂O₃. The final step corresponds to dehydration of OH groups.

Close

Figure 3

TGA thermograms of (a) Fe₃O₄, (b) MNP-NH₂, (c) MNP-Alky, (d) MNP- Trz magnetic nanoparticles.

Export to PPT

The weight losses relative to the crude MNP for MNP-NH₂, MNP-Alky and MNP-Trz are 1.54 wt.%, 6.83 wt.% and 11.94 wt.%, respectively. The quantity of APTES immobilized onto the surface of MNP, determined from the weight loss under the assumption that all the ethoxy group were eliminated during grafting of APTES to Fe₃O₄ core and that only the organic part of the linker contributes to the mass loss was estimated and found to be 0.28 mmol/g of magnetite.

TGA data of Fe₃O₄@SiO₂ show that the quantity of silica shell grafted on Fe₃O₄ core was around 1.89 wt.%. MNPS-NH₂, MNPS-Alcy and MNPS-Trz exhibit the same thermal degradation profiles with weight losses of 6.01 wt.%, 7.60 wt.% and 10.53 wt.% respectively corresponding to the organic functional group decomposition (Fig. S2. supplementary file).

The morphology and the size of neat Fe₃O₄ and Fe₃O₄@SiO₂ NPs and their functionalized analogues were studied by TEM. Some typical TEM images are shown in Fig. 4a–c. All investigated particles have a spherical shape with a low particle size polydispersity. The mean particle size is almost the same for crude and functionalized magnetic particles, which is around 9.6 nm. This indicates that the used chemical reactions occur only at the surface of magnetic particles and have no drastic effect on the colloidal stability of the dispersion.

Close

Figure 4

TEM images of (a) neat Fe₃O_4, (b) Fe₃O₄@SiO_2, (c) MNP-NH₂, (d) MNP- Trz; SEM images of (e) neat Fe₃O_4, (f) MNP-Trz magnetic nanoparticles.

Export to PPT

Similar results, i.e. size and spherical shape of investigated MNP were obtained by SEM analyses (Fig. 4e–f). Particle sizes obtained by dynamic light scattering were however much larger than those deduced from TEM and SEM images. This can be explained by the possible low aggregation of particles after various magnetic separation and redispersion cycles.

The effect of pH on zeta potential (ζ) of MNP was also investigated and confirmed the immobilization of triazole units. As shown in Fig. 5, the magnetite particles are positively charged at lower pH and negatively charges at higher pH. The isoelectric point (IEP) was found to be around pH of about 6.45 consistent with the values reported in the literature (Badruddoza et al., 2011; Takafuji et al., 2004). After modification of the particles with APTES, the IEP shifted to 8.2 close to the pKa of primary amine group) revealing the presence of amine groups on MNP-NH₂ particles. The positive zeta potential below the IEP is attributed to the protonation of amine groups on the particle surface (Xu et al., 1997).

Close

Figure 5

Zeta potential of crude Fe₃O₄, MNP-NH₂and MNP-Trz magnetic nanoparticles as function of pH in 1 mM NaCl solution.

Export to PPT

Since tertiary amines gave rise to less protonation reactions then primary amines of APTES, the IEP of MNP-Trz shifted from 10 to 4.3 after cycloaddition reaction leading to the formation of triazole synthons and revealed that the amino-functionalized magnetic nanoparticles were positively charged at pH < 4.3 and negatively charged at pH > 4.3. These results confirmed the triazole functionalization of MNP.

Similar results were obtained for silica-coated magnetite (plot is shown in Supplementary file Fig. S3). The IEP of silica-coated magnetite nanoparticles Fe₃O₄@SiO₂ was at pH = 3, and is in good agreement with reported values (Karatapanis et al., 2011).

The magnetic properties of Fe₃O₄, MNP-NH₂, MNP-Trz and their silica coated homologous were investigated by vibrating sample magnetometer (VSM). Typical hysteresis curves are depicted in Fig. 6. All samples exhibit superparamagnetic behavior without remanence or coercivity. This feature indicates that these particles are responsive to an external magnetic field without residual magnetism after removal of the applied magnetic field. The saturation magnetization values (Ms) for Fe₃O₄, MNP-NH₂, and MNP-Trz nanoparticles are 59.07, 58.68 and 43.10 emu/g respectively. The relatively slighter Ms value of Fe₃O₄ MNP in comparison with bulk magnetite (92 emu/g) was already reported in previous study and explained by the rather small size of the Fe₃O₄ MNP prepared by co-precipitation method (Feng et al., 2008). The triazole-functionalized MNP shows lower magnetization value than crude magnetite particles probably due to the presence of non-magnetic materials on the surface of nanoparticles such as passivation non-magnetic layer.

Close

Figure 6

Magnetization curves of MNP and MNPS nanoparticles at room temperature.

Export to PPT

No specific effect on superparamagnetic properties was observed for silica coated nanoparticles in comparison with Fe₃O₄ NPs (Ms_(MNP-Trz) = 43.1 emu/g, Ms_(MNPS-Trz) = 47.66 emu/g) supporting the idea that both particle series could be detected easily and can be used for separation of metal-adsorbed MNP. The lower plateau values obtained for silica series except for MNPS-Trz are due to the magnetic silica shell surrounding the magnetite core. For this latter material, the slightly lower functionalization percentage detected by TGA analysis (10.53 wt.% for MNPS-Trz and 11.94 wt.% for MNP-Trz) should be at the origin of the slightly higher magnetization value.

The surface area calculated by BET analysis of neat magnetite Fe₃O₄, Fe₃O₄@SiO₂ and the triazole-functionalized NPs MNP-Trz was determined by nitrogen sorption measurements and was found to be 90.65, 94.96 and 119.35 m² g⁻¹ respectively (Figs. S4–S5, Supplementary file). The surface areas of crude and silica-coated Fe₃O₄ are substantially similar suggesting similar morphological properties of these materials.

3.2 Triazole-based magnetic nanoparticles for heavy metal adsorption

3.2.1

3.2.1 Effect of initial pH

The pH of the aqueous solution is a critical controlling parameter in the adsorption process. Thus, it is important to determine the optimum pH value for better metal ion adsorption on synthesized magnetic nanoadsorbents.

The effect of pH on the adsorption of Cu²⁺, Zn²⁺ and Pb²⁺ ions by synthesized MNP-Trz nanoparticles was investigated at 25 °C and at initial ion concentration of 400 mg L⁻¹. The investigated pH range was 1–6 since at higher pH values copper, zinc and lead ions precipitated as Cu(OH)₂, Zn(OH)₂ and Pb(OH)₂, respectively.

The evolution of adsorption ability of magnetic particles as a function of pH v is illustrated in Fig. 7. It can be seen that the adsorption ability is strongly pH dependent and increased along with increasing pH value in the range of 2–5 but changed slightly above pH 5.

Close

Figure 7

Effect of pH on the adsorption capacity of Cu (II), Pb (II) and Zn(II) metal ions by MNP-Trz in aqueous solution. Initial concentration of metal ions 400 mg L⁻¹; reaction time 240 min; mixed 250 rpm at T = 25 °C.

Export to PPT

To explain this feature, two aspects should be considered. First, the surface composition of metal oxide adsorbent may undergo protonation/deprotonation reaction depending on pH solution as expressed in Scheme 2, and secondly the ion species present at investigated pH media.

Close

Scheme 2

Protonation/deprotonation reactions of MNP in aqueous solution.

Export to PPT

Thereby, in high acidic solution, the occurrence of high proton concentrations induces competitive adsorption reactions of proton ions for existing adsorption sites resulting in lower metal ion adsorption. Furthermore, the positively charged surface of MNP as illustrated in schema 2 induces electrostatic repulsions of Cu²⁺, Zn²⁺ and Pb²⁺ ions’ disadvantaging adsorption process.

Another reaction that may take place at the surface of triazole functionalized MNP surface in strong acid media is related to the protonation of the triazole heterocycles to produce a protonated 1,2,3-triazolium salts (Scheme 2) which can also be involved in electrostatic repulsive interactions responsible of low affinity between magnetic nanoparticles and metal ions in acidic (Yacob and Liebscher, 2015).

As pH values increased, the involved repulsive electrostatic interactions become weaker and the negatively charged MNP, as demonstrated by zeta potential values for pH > pH_IEP = 4.3, attracts the positively charged metal ions. In this condition, the nitrogen atoms of 1,2,3-triazole moiety may act as N donor ligands and the unpaired electron of nitrogen atoms could create coordination bonds with M²⁺ ions leading to mono, di or trinucleating ligand architectures (Badruddoza et al., 2011). Meantime, the competition of proton adsorption sites becomes less significant. Both effects result in enhancement of sorption capacity.

As shown in Fig. 7, it is interesting to notice that the adsorption capacity for Pb²⁺ ions is higher than that for Cu²⁺ and Zn²⁺ irrespective of pH, suggesting the possible selectivity of magnetic nanoadsorbents for Pb²⁺.

3.2.2

3.2.2 Adsorption isotherm

Adsorption isotherm describes how adsorbate (pollutant) interacts with adsorbent. To understand the adsorption mechanism for a better effective use of the nanoadsorbent, it is important to correlate the experimental equilibrium data with the theoretical or empirical models. In the present work, the equilibrium data were analyzed according to Langmuir and Freundlich models expressed by Eqs. (4) and (5), respectively.

(4)

$$\frac{C_e}{q_e}=\frac{C_e}{q_m}+\frac{1}{K_L{q}_m}$$

(5)

$${\mathit{lnq}}_e={\mathit{lnK}}_F+\frac{1}{n}{\mathit{lnC}}_e$$ where q_e (mg g⁻¹) represents the amount of metal ions adsorbed on the absorbent at equilibrium; C_e (mg L⁻¹), the equilibrium ion concentration in solution; q_m (mg g⁻¹), the maximum adsorption capacity of the adsorbent corresponding to complete monolayer coverage; K_L (L mg⁻¹), the Langmuir adsorption equilibrium constant; K_F (L mg⁻¹), the Freundlich constant and n is the heterogeneity factor that lies for favorable adsorption process in the range of 1–10.

The Langmuir model assumes that adsorption takes place at specific homogeneous sites within the adsorbent and is usually applied for many monolayer adsorption processes while the Freundlich model describes heterogeneous adoption systems.

The equilibrium isotherms for the adsorption of Cu²⁺, Zn²⁺ and Pb²⁺ ions by MNP-Trz at pH 5.5 and 25 °C are shown in Fig. 8. Initial metal ion concentrations varied from 50 to 600 mg L⁻¹. It can be seen that the equilibrium adsorption capacity (q_e) of the metal ions increased with increasing the equilibrium ion concentrations in aqueous solution (C_e) until reaching equilibrium.

Close

Figure 8

Equilibrium isotherms of Cu(II), Pb(II) and Zn(II) metal ions by MNP-Trz in aqueous solution. Initial pH = 5.5; reaction time 240 min; mixed 250 rpm at T = 25 °C.

Export to PPT

The Langmuir and Freundlich parameters are determined and summarized in Table 1 and Fig. 9. The values of q_m and K_L were determined from the slope and intercept of the linear form of Langmuir equation C_e/q_e versus C_e and K_F and 1/n parameters from the intercept and slope of the linear form of the Freundlich equation ln q_e versus ln C_e. It can be seen from the best fitting parameter R² values, that Langmuir model fits the experimental data better than Freundlich model suggesting a monolayer homogeneous adsorption.

Table 1 Parameter values of Langmuir and Freundlich isotherms for adsorption capacity of Cu (II), Pb (II) and Zn(II) metal ions onto MNP-Trz at 25 °C.

Metal ions	Langmuir				Freundlich
	qm (mg g−1)	KL (L g−1)	RL range	R2	n	Kf (L g−1)	R2
Cu2+	87.87	0.0176	0.0835–0.5738	0.9965	2.59	7.97	0.9523
Pb2+	167.78	0.0065	0.1966–0.7508	0.9982	1.79	4.45	0.9708
Zn2+	51.20	0.0632	0.0258–0.2137	0.9996	7.21	21.72	0.9099

Close

Figure 9

The Langmuir (a) and Freundlich (b) isotherm plots of Cu(II), Pb(II) and Zn(II) metal ions by MNP-Trz

Export to PPT

The dimensionless constant separation factor R_L, which is considered as the more reliable indicator for the adsorption capacity, was evaluated according to the following equation:

(6)

$$R_L=\frac{1}{1+K_L{C}_0}$$ where K_L (L g⁻¹) is the Langmuir constant and C₀ is the initial concentration (mg L⁻¹). The calculated R_L values are gathered in table 1 and are between 0 and 1 indicating that the adsorptions of metal ions are favorable.

The maximum adsorption values obtained by Langmuir isotherm were 87.87, 167.78 and 51.20 mg g⁻¹ for Cu²⁺, Pb²⁺ and Zn²⁺, respectively and are in accordance with the obtained experimental values. We can deduce the following capacity order: Pb²⁺ > Cu²⁺ > Zn²⁺.

This selectivity series could be the result of various factors related to the electronic properties of different atoms present in the ligand as well as metal ion properties which govern the complexation process. It is known that the higher the metal ionic potential (which is related to metal ion charge/radius ratio), the higher will be its tendency to attract electrons of the ligand, and hence the complex formed will be the most stable (Gurdeep, 2010). For cations having the same charge, the stability of complexes decreased with increasing of the ionic radii; smaller is the ion, easier it reacts with the ligand. The hydrated ion sizes of the metal increased in the order Pb²⁺ (4.01 A°) < Cu²⁺ (4.19 A°) < Zn²⁺ (4.30 A°) (Calvo et al., 2009). This classification is consistent with the obtained capacity order.

Zhu and al. reported the application of xanthate-modified magnetic chitosan for the removal of multi-metal Pb(II), Cu(II) and Zn(II) solutions. Based on the covalent index of metal ion $X_m^{2}r$ where $X_m$ is electronegativity and r is ionic radius, they obtained the same order of preference of metal ions onto magnetic chitosan particles. The higher covalent index of Pb(II) ions indicated its stronger attraction to the lone pair of electrons in sulfur and nitrogen atoms of xanthate-modified chitosan to form complexes (Zhu et al., 2012).

The maximum uptakes of Cu²⁺, Zn²⁺ and Pb²⁺ by MNP-Trz particles obtained on this work are compared with results reported in literature for other magnetic adsorbents as shown in Table S1 - Supplementary file. In general, the MNP-Trz particles prepared and used in this work have higher adsorption capacities and then reported adsorbents, except for (Acrylic acid-crotonic acid) copolymer-modified magnetic nanoparticles (Ge et al., 2012). For this latter material, the content of polymer was estimated to 36.7% in weight bases on the total weight of modified MNP. The lower density of triazole synthons in our material (11.94 wt.%) should be at the origin of the lower adsorption capacity.

3.2.3

3.2.3 Adsorption kinetics

The adsorption kinetic of metal ions was investigated for initial metal ion concentration of 400 mg L⁻¹ at pH = 5.5 and at 25 °C. The obtained results are illustrated in Fig. 10.

Close

Figure 10

Effect of contact time on adsorption capacity of Cu (II), Pb (II) and Zn(II) metal ions onto MNP-Trz at 25 °C.

Export to PPT

The equilibrium was reached for copper and Zinc ions at ∼ 175 min and at ∼ 240 min for lead ion. These results are in good agreement with data reported for adsorption of different metal ions onto Fe₃O₄ based nanoadsorbents for which similar equilibrium times were described (Zhu et al., 2012).

The uptake–time curves show that the maximum uptake follows the order Pb²⁺ > Cu²⁺ > Zn²⁺.

A stirring time of 4 h was used in all further adsorption experiments to ensure equilibrium.

In order to examine the controlling mechanism of adsorption, several kinetic models are used to fit experimental data. We used in this work the pseudo first-order and second-order models expressed by the following equations:

Pseudo-first order kinetic:

(7)

$$\mathit{ln}(q_e-q_t)={\mathit{lnq}}_e-k_{1}t$$ Second order kinetic:

(8)

$$\frac{t}{q_t}=\frac{1}{k_2{q}_e^2}+\frac{1}{q_e}t$$ where q_t (mg g⁻¹) is the adsorption at time t (min); q_e (mg g⁻¹) is the adsorption capacity at equilibrium, and k₁ (min⁻¹) and k₂ (g mg⁻¹ min⁻¹) are the kinetic rate constants for the pseudo-first order and second order models respectively.

The kinetic parameters obtained from these models are tabulated in Table 2 and the plots for both models are given in Fig. 11. It is found that for all investigated metal ions, the correlation coefficients (R²) for the second order kinetic are higher than those of pseudo-first order. Furthermore, the adsorption capacity values at equilibrium q_e obtained from the slope of the plot of the second order equation (t/q_t against t) are more consistent with the experimental q_e values than those obtained from the slope of the pseudo-first order plot. It is clear that the adsorption kinetic of metal ions on magnetic nanoparticles material followed this model, revealing that chemisorption might be the rate-limiting step that controls the adsorption process. Usually, for magnetic nanoparticles bearing surface chelating functional groups, the main mechanism controlling the adsorption process is a chelating exchange process following a second-order chemical reaction (Fu and Wang, 2011).

Table 2 Kinetic parameters for Cu²⁺, Pb²⁺ and Zn²⁺ ion adsorption by MNP-Trz.

Metal ions	qe (exp) (mg g−1)	Pseudo-first order			Second order
		k1 (min−1)	qe (mg g−1)	R2	k2 (g mg−1 min−1)	qe (mg g−1)	R2
Cu2+	77.3	0.01581	31.07	0.8681	6.93 10−4	82.37	0.9973
Pb2+	136.3	0.01376	128.68	0.9166	1.04 10−4	162.34	0.9970
Zn2+	48.4	0.01248	17.81	0.8444	9.62 10−4	51.73	0.9962

Close

Figure 11

(a) Pseudo first-order kinetics and (b) second-order kinetics for metal ion adsorption onto MNP-Trz at 25 °C.

Export to PPT

3.2.4

3.2.4 Comparison of adsorption behavior of MNP-Trz and MNPS-Trz

The effect of silica shell on adsorption behavior of triazole functionalized particles was investigated for Pb²⁺ ions in standard conditions (pH = 5.5, T = 25 °C, 250 rpm). We noticed the same kinetic profile (Fig. 12) with a slightly higher metal uptake for MNPS-Trz material (q_e (MNP-Trz) = 136.3 mg g⁻¹, q_e (MNPS-Trz) = 145.1 mg g⁻¹). This feature should be the result of involvement of silanol group on complexation behavior according to the following equation (Hu et al., 2010).

(9)

$$\text{Si—OH}+{\text{Pb}}^{2+}+{\text{H}}_2\text{O}\iff \text{Si—}{\text{O}}^{-}{(\text{PbOH})}^{+}+2{\text{H}}^{+}$$

Close

Figure 12

Effect of contact time on adsorption capacity of Pb (II) metal ion onto silica coated (MNPS-Trz) and uncoated (MNP-Trz) magnetic particles at 25 °C.

Export to PPT

The silica coated particles could be used as an efficient adsorbent with reliable chemical stability and good dispersion in aqueous media.

3.2.5

3.2.5 Desorption study

Desorption study was conducted for Pb²⁺ pre-adsorbed on MNP-Trz to explore the possibility of recyclability of the magnetic nano-adsorbent. Referring to the pH effect on adsorption efficiency of MNP-Trz, it was expected that acid media with pH below 2 would be effective for desorption. Ge et al. reported that the desorption ratio of metal ions increased as H⁺ concentration increased and reached an optimal value for ∼0.1 mol L⁻¹ beyond which the magnetic adsorbent was corroded by acid media (the adsorbent is completely destroyed at 2 mol L⁻¹ after 3 h) (Ge et al., 2012).

Thus, Pb²⁺ preadsorbed on MNP-Trz dealt with weakly acidic HNO₃ solution in concentration range of 0.1–0.2 mol L⁻¹. As shown in Fig. 13, desorption ratio of adsorbed ions increased as a function of time and HNO₃ concentration. A constant desorption efficiencies of 88% were reached for 0.2 mol L⁻¹ HNO₃ within 30 min indicating a fast desorption process. The obtained results demonstrated that this magnetic solid support dispersion can be used efficiently for the removal and recovery of metal ions from wastewaters.

Close

Figure 13

Effect of contact time in desorption efficiency of Pb(II) metal ion from the MNP-Trz adsorbent with different HNO₃ concentrations at T 25 °C.

Export to PPT