Preparation of magnetic Pb(II) and Cd(II) ion-imprinted microspheres and their application in determining the Pb(II) and Cd(II) contents of environmental and food samples

Abstract

A novel magnetic ion-imprinted polymer (MIIP) was prepared by inverse microemulsion polymerization using ethylene glycol dimethacrylate (EGDMA) as the cross-linker, and 2,2-azo-bis-isobutyronitrile (AIBN) as the initiator. Fe₃O₄ particles were incorporated into the imprinted polymer matrix containing 4-vinyl pyridine and acrylate-modified Spirulina platensis as the functional monomers, Pb(II) and Cd(II) as double-templates. The MIIP product was characterized by Fourier-transform infrared spectroscopy, thermal gravimetric analysis and scanning electron microscopy and then evaluated for the magnetic solid-phase extraction of Pb(II) and Cd(II) from environmental and food samples. Inductively coupled plasma-optical emission spectrometry detection was also performed. The obtained MIIP showed the magnetic property, agglomerates of globular particles and rough surface morphologies. The maximum static adsorption capacities of MIIP for Pb(II) and Cd(II) were 108 and 56 mg g⁻¹, respectively. The factors affecting the adsorption and separation of Pb(II) and Cd(II), namely, solution pH, adsorbent dosage, eluting solvent, and elution time, were investigated. Under the optimized conditions, the detection limits (S/N = 3) of the method were 0.1 and 0.08 μg kg⁻¹ for Pb(II) and Cd(II), respectively. The relative standard deviations (RSDs) for 10-replicate detections of a soil sample were 3.3% and 3.7% for Pb(II) and Cd(II), respectively. Standard solutions containing Pb(II) and Cd(II) in the concentration range of 0.8 to 5000 μg kg⁻¹ were analyzed by the proposed procedure and it was found that the calibration curve was linear in this range (R² ≥ 0.999). The proposed method was successfully used to determine the Pb(II) and Cd(II) contents of cinnamon, lotus root and sweet potato with recovery rates ranging from 87.5% to 106%.

---

1. Introduction

Cadmium and lead are two of the most toxic elements to humans and animals, even at low concentrations.¹ Lead causes chronic inflammation of the kidney and heart, inhibits brain development, and decreases nerve conduction velocity. Cadmium is carcinogenic to animals and humans, and chronic cadmium accumulation can lead to kidney damage, impaired regulation of calcium and phosphorus, bone demineralization, osteomalacia, and pathological fractures.^2–4 Monitoring the level of lead and cadmium ions in environmental and food samples is gaining increased attention because these elements are a severe threat to animals, plants, humans, and the surrounding environment. Considering the low concentration of lead and cadmium ions in environmental and food samples, separation and pre-concentration are often necessary prior to the determination itself.

Spirulina platensis (S. platensis) is a blue-green alga that is an alternative source of protein for human food and animal feeds.⁵ The surface of this alga has many –NH₂, –COOH, –OH, and other charged groups that can serve as chelation sites for metals, and the organism can potentially act as a biosorbent for heavy metal removal.^6–9 S. platensis has been used as well to remove Cd²⁺, Cr³⁺, Cu²⁺, Co²⁺, Zn²⁺, Pb²⁺, and Ni²⁺ ions from wastewater.^10–12 S. platensis also shows great sorption capacity toward these metals, especially Pb²⁺, in both one- and three-metal systems. However, the organism lacks selectivity and cannot be rapidly and effectively separated after wastewater treatment.

Magnetic nanoparticles are attracting worldwide attention because of their excellent magnetic response, high dispersibility, relatively large surface area, and easy surface modification. Magnetite nanoparticles can reportedly remove Cr(VI), Pb²⁺, Hg²⁺ from wastewater and food samples.^13–15 Li et al.¹⁶ reported the separation and accumulation of Cu(II), Zn(II), and Cr(VI) from aqueous solution using magnetic chitosan modified with diethylenetriamine. Sadeghi et al.¹⁷ reported the use of surface-modified magnetic Fe₃O₄ nanoparticles as a selective sorbent for the solid-phase extraction (SPE) of uranyl ions from water samples. However, all of these modified magnetic nanoparticles are applied in the non-specific absorption of metals and have low selectivity. Magnetic nanoparticles modified with ionic imprinted polymer have also been developed.^18–24 Metal ion-imprinted polymers are synthesized with metal ion templates in the presence of a suitable monomer (vinylated reagent) or a monomer and an auxiliary/complexing non-vinylated reagent that can selectively distinguish the template ion from related analogous compounds. Different approaches to the synthesis of metal ion-imprinted polymers were reviewed by Rao et al.²⁵ These conventional methods exhibit the drawbacks of low binding capacity and selectivity, slow mass-transfer rate, and few complexing ligands with vinyl groups. Latif et al.²⁶ reported the synthesis of Ni²⁺ and Cu²⁺ imprinted polymer by functionalized sol–gel materials containing amino group as ligand.

The hierarchical imprinting method was introduced by Dai et al.²⁷ with surfactant micelles and metal ions as template. Hierarchically imprinted polymers can be precisely controlled at adsorption sites and pore structures. Dickert et al.²⁸ introduced the idea of “double molecular imprinting” using anthracene and chrysene as templates that were simultaneously imprinted in a crosslinked polyurethane thin film, which was able to recognize both the templates. Compared with single-template imprinted polymers, hierarchically imprinted polymers and double-template imprinted polymers demonstrate high binding capacity, high selectivity, and fast mass transfer.^29,30 The use of vinylated S. platensis instead of non-vinylated S. platensis to fix the complexing agent to the polymeric matrix also reportedly improves selectivity toward heavy metals.³¹

In the present study, a novel magnetic ion-imprinted polymer (MIIP) was synthesized by water-in-oil-in-water (W/O/W) emulsions polymerization using lead and cadmium ions as double-templates, 4-vinylpyridine (4-VP) and acrylate-modified S. platensis as functional monomer with Fe₃O₄ particles existing. W/O/W emulsions are three-phase systems in which small internal aqueous droplets, surrounded by a primary surfactant-stabilizing layer, are dispersed in the oil phase, which, in turn, is dispersed in the external aqueous phase and is also surrounded by a secondary surfactant layer. After preparing a stable W/O/W emulsion composed of water1, monomers, and water2, adding initiator, a radical polymerization was carried out selectively in the monomer oil phase. The resulting polymer (MIIP) was separated by a magnet and was washed several times with 0.1 mol L⁻¹ EDTA solution and with methanol several times to remove lead and cadmium ions, organic solvent, internal aqueous droplets and surfactant micelles etc., respectively. The removal of the metal ion from the polymer leaves cavities that exhibit ionic recognition. The removal of the organic solvent and internal aqueous droplets will probably result in the formation of rough and porous structures or multihollow structures which includes large surface areas and excellent metal ion transport kinetics.³² The MIIP was characterized by Fourier-transform infrared spectroscopy (FTIR), Thermal gravimetric analysis (TGA) and scanning electron microscopy (SEM). The adsorption performance, reusability, and application of MIIP for Pb(II) and Cd(II) were investigated in detail.

2. Materials and methods

2.1. Materials and instruments

Cd(NO₃)₂, Pb(NO₃)₂, Mg(NO₃)₂, Zn(NO₃)₂, and Cu(NO₃)₂ standards were supplied by CMP Sinoexpo Biological Technology Co., Ltd. (Shanghai, China). Tween 80, Span 80, ferrous chloride (FeCl₂·4H₂O), and ferric chloride (FeCl₃·6H₂O) were obtained from Tianjin Chemical Reagent Co. (Tianjin, China). Carboxymethyl cellulose (CMC) and cetyltrimethylammonium bromide (CTAB) were purchased from Pure Crystal Shanghai Reagents Co., Ltd. (Shanghai, China). Azodiisobutyronitrile (AIBN), 4-VP, and ethylene glycol dimethacrylate (EGDMA) were purchased from Aladdin Chemistry Co., Ltd. (Shanghai, China). S. platensis was provided by the Zhaoqing Bangjian Pharmacy Plant in China. Water was purified through a Milli-Q water system (Bedford, USA). All other chemicals were analytical-reagent grade.

UV-vis spectra and absorbance were measured using a double beam spectrophotometer (GBC 916 UV-vis model). FTIR spectra were recorded with a Shimadzu model FTIR-84003 spectrophotometer. The surface morphology and structure of the materials were identified using a SEM system (NGB4-DXS-10AC). The specific surface areas and average pore size of the polymers were measured from nitrogen adsorption data according to the Barrett–Joyner–Halenda (BJH) and Brunauer–Emmett–Teller (BET) methods, respectively, using an ASAP2020M + C system (Micromeritics, USA). A Thermo Electron inductively coupled plasma atomic emission spectrometer (ICAP-6300-DUO, Thermo Electron, USA) was used for multi-elemental determinations. pH was measured using a PHS-3C pH meter (Dapu Instrumentation Corp., Ltd. Shanghai, China). Thermal gravimetric analysis was carried out on a NETZSCH STA 449F3 instrument under N₂ atmosphere at the heating rate of 10 °C min⁻¹.

2.2. Preparation of Fe₃O₄ particles

Fe₃O₄ magnetic nanoparticles were prepared as previously reported³³ with modifications. FeCl₃·6H₂O (1.35 g) and FeCl₂·4H₂O (0.6 g) were dissolved in 16 mL of deionized water under magnetic stirring at 40 °C and nitrogen bubbling. Then, 1.6 mL of 25% ammonia solution was added. The resulting suspension was stirred and heated at 50 °C for 1 h. The suspension was decanted, and the black residue was washed several times with deionized water to neutralize the aqueous solution. Three grams of emulsifier (m_{Span 80} : m_{Tween 80} = 1 : 2) was added, and the mixture was stirred and heated at 80 °C for 1 h. The resulting material was washed four times with ethanol and dried in a vacuum.

2.3. Acrylate-modified S. platensis

A 250 mL three-necked flask equipped with a condenser and an anhydrous CaCl₂ drying tube, thermometer, and hydrochloric acid (HCl) tail gas absorption device was placed in a heat collection-type oil bath under magnetic stirring. Acrylic (1.2 mol) and SOCl₂ (1.0 mol) were added to the flask, stirred, and heated at 40 °C for 2 h. Acrylic acid chloride was collected after distilling the mixture to remove the solvent.

A 250 mL three-necked flask equipped with a dropping funnel and thermometer was placed in the heat collection-type oil bath under magnetic stirring. 4.0 g dry S. platensis powder, 5 mL of triethylamine, and 40 mL acetone were added to the flask under magnetic stirring and nitrogen bubbling, slowly dripping a mixture solution of the collected acryloyl chloride and 25.0 mL acetone. The resulting solution was stirred in an ice-cooled bath for 8 h and at room temperature for 12 h. The reaction solution was transferred to a rotary evaporator from which vacuum distillation was performed to remove the acetone. The resulting material was filtered and acrylate-modified S. platensis was obtained. A schematic diagram of this synthesis was presented in Fig. 1.

Fig. 1 A schematic model for synthesis of magnetic ion-imprinted polymer.

2.4. Preparation of lead and cadmium ion-imprinted microspheres

The imprinted polymer, lead, and cadmium ion-imprinted microspheres were synthesized by emulsion polymerization. In a typical procedure, 3.0 g of emulsifier (Span 80 : Tween 80 = 1 : 3) was dissolved in a mixture of 20 mmol EGDMA and 10 mL toluene under magnetic stirring. A 20 mL aqueous solution containing 3.0 g of acrylate-modified S. platensis, 3.0 mmol 4-VP, 1.0 mmol Pb(NO₃)₂, 1.0 mmol Cd(NO₃)₂, 1.5 g Fe₃O₄ particles, and 0.5 g CMC was added to the above solution. The mixture was sonicated for 20 min to yield W/O emulsions. The emulsions were placed in a 150 mL aqueous solution containing 0.15 g CTAB, which was stirred at rate of 400 r per min to form a W/O/W emulsion. The oxygen in the emulsion was removed by bubbling nitrogen for 10 min. After the addition of 0.2 g AIBN, the mixture was polymerized at 70 °C for 12 h under nitrogen flow and stirred at a rate of 400 r per min. The resulting polymer (MIIP) was placed in a 150 mL Erlenmeyer flask and separated by a magnet, then the polymer was washed with 0.1 mol L⁻¹ EDTA solution several times (5 min each times) under ultrasound until no lead and cadmium ions were detectable in the effluent, after which the polymer was washed with methanol several times to remove surfactant micelle etc. and dried under vacuum.

The single magnetic ion-imprinted polymer (MIIP/Pb and MIIP/Cd) was prepared by following the procedure mentioned above using Pb(NO₃)₂ and Cd(NO₃)₂ as a template, respectively. Magnetic non-imprinted polymer (MNIP) was also prepared by the same procedure but without the addition of Pb(NO₃)₂ and Cd(NO₃)₂ (Fig. 1).

2.5. Adsorption experiments

Adsorption experiments were performed by batch method in aqueous solution. MIIP or MNIP of 30.0 mg weight was added to 10 mL sodium acetate–acetic acid solutions containing Pb(II) and Cd(II) with initial concentrations of 0.20 mg mL⁻¹ to 1.20 mg mL⁻¹ (ref. 19) in 150 mL conical flasks, then shaken at 150 rpm for 2 h at room temperature. MIIP was collected with the help of an external magnetic force. The supernatant was analyzed by inductively coupled plasma-optical emission spectrometry (ICP-OES), the adsorption amount of Pb(II) and Cd(II) was calculated using eqn (1), and the binding isotherm was plotted.where Q is the amount of Pb²⁺ or Cd²⁺ adsorbed (mmol g⁻¹); C₀ and C are the initial and equilibrium concentrations (mg L⁻¹) of Pb²⁺ or Cd²⁺, respectively; V is the volume of solution (L); M is the relative molecular mass of Pb²⁺ or Cd²⁺; and W is the weight of dry adsorbent (g).

The binding isotherm of MNIP for Pb²⁺ or Cd²⁺ ions was measured using similar procedures as with MIIP for Pb²⁺ or Cd²⁺ ions.

All the experiments were performed in duplicates and the average values were recorded.

2.6. Selectivity experiments

The competitive absorption experiments were conducted with Cu²⁺ and Zn²⁺ ions as the structural analogues of the template Cd²⁺ and Pb²⁺ ions, and with Mg²⁺ as the reference metal ion. The experiment was carried out by adding 30.0 mg of sorbents (MIIP, MNIP, MIIP/Pb and MIIP/Cd) in an Erlenmeyer flask with 25.0 mL standard solution of each metal ion at an initial concentration of 1.0 × 10⁻³ mol L⁻¹ and pH 6. The solution was shaken and adsorbed for 30 min at room temperature, and the supernatant was analyzed by ICP-OES. The distribution coefficient and selection coefficient were calculated by the following formula:where K_d is the distribution coefficient; C_e is the equilibrium concentration of metal ions; Q is the binding amount; k is the selectivity coefficient; and X²⁺ represents Pb²⁺, Cd²⁺, Cu²⁺, or Zn²⁺.

The relative selectivity coefficient k′ used to estimate the effect of imprinting on selectivity is defined by eqn (4):³⁴

where k_MIIP and k_MNIP are the selectivity coefficients of the prepared magnetic imprinted polymers and MNIPs, respectively. The larger the k′, the better the adsorption affinity and selectivity of the magnetic imprinting material for the template as compared with the non-imprinting material.

2.7. Sample preparation

A soil sample was collected in a plastic bag from an industrial zone in Dinghu Zhaoqing, China, transported to the laboratory as fast as possible, air-dried, pulverized, and sieved through a 0.1 mm mesh. Cinnamon, lotus root, and sweet potato samples were acquired from local supermarkets, transported to the laboratory, dried, and ground. The dried soil, cinnamon, lotus root, or sweet potato sample (2.0000 g) was weighed into a Teflon flask and decomposed with 10.0 mL of concentrated HNO₃ and 1.0 mL of 30% (v/v) H₂O₂. The mixture was heated at 100 °C and evaporated nearly to dryness. After cooling, the sample was diluted to 50 mL with a sodium acetate–acetic acid buffer solution of pH 6.0.

3. Results and discussion

3.1. Characterization of MIIP and MNIP

The FTIR spectra of Fe₃O₄, acrylate-modified S. platensis, MIIP and MNIP were shown in Fig. 2. A characteristic Fe–O band at about 580 cm⁻¹ was found in MIIP and MNIP spectra, which indicated that Fe₃O₄ was embedded in polymers. The characteristic peaks of –NH or –OH, –C=O, –C=C, –CO–NH and –C–O groups were at 3445 cm⁻¹, 1680 cm⁻¹, 1600 cm⁻¹, 1545 1260 cm⁻¹ and 1094 cm⁻¹ in acrylate-modified S. platensis, respectively. In MIIP and MNIP spectra, the wide and strong adsorption bands at about 3566 and 3450 cm⁻¹ were due to the stretching vibrations of –NH and –OH groups from S. platensis, respectively. The stretching bands and bending vibration bands of CH₃ were observed at about 2950 cm⁻¹, 1450 cm⁻¹ and 1390 cm⁻¹, respectively. The strong absorption bands at 1730 cm⁻¹ indicated the existence of a large number of carbonyl functional groups (–COO) derived from EGDMA and S. platensis. Compared to the peak at 1600 cm⁻¹ of acrylate-modified S. platensis, MIIP and MNIP have very weak C=C vibrations at 1620 cm⁻¹, suggesting that the C=C double bond is broken up after polymerization. The band at 1450 cm⁻¹ was attributed to the bending vibrations of C–N from 4-VP¹⁹ and S. platensis. The strong adsorption bands at 1260 and 1165 cm⁻¹ indicated the existence of C–O stretching vibration in polymers. In comparison with the infrared data of acrylate-modified S. platensis and Fe₃O₄, the existence of these characteristic peaks in MIIP and MNIP showed the successful coating of the magnetic core by a polymeric layer, which was synthesized using acrylate-modified S. platensis and 4-VP.

Fig. 2 Infrared spectra for Fe₃O₄ (a), S. platensis with ethylene (b), MNIP (c) and MIIP (d).

For the determination of the amount of coated polymer and Fe₃O₄ encapsulated in MIIP, the TG analysis was performed on the MIIP. As shown in Fig. 3, as the temperature grew to 143 °C, the mass of MIIP lost 3.32%, which probably caused by the evaporation of water molecules. From 143 to 309 °C, the thermal degradation of MIIP lost 18.12% of mass, which might be due to the decomposition free small molecular in MIIP. When the temperature rose to 572 °C, a significant mass loss of 60.36% was observed, mainly caused by the degradation of polymer. The final residue (16.23%) at 700 °C was found to be consistent with calculated value for 1.5 g of Fe₃O₄ particles in the MIIP (calc. 16.65%). Thus, this magnetic MIP was stable up to 300 °C and about 62% and 16% of this composite was polymer and Fe₃O₄, respectively.

Fig. 3 Thermogravimetric curve of MIIP.

SEM images of the morphological features of Fe₃O₄ (M), MIIP (A) and MNIP (B) are shown in Fig. 4. The particle size of Fe₃O₄, MIIP and MNIP is about 60 nm, 500 nm and 450 nm, respectively. The MIIP and MNIP particles exhibited similar agglomerates of globular particles and rough surface. The specific surface area, average pore diameter and pore volume of Fe₃O₄, MIIP and MNIP were summarized in Table 1. As shown in Table 1, the surface areas, average pore diameter of MIIP and MNIP were higher than the Fe₃O₄. There were no significant differences in specific surface areas, average pore diameter, and pore volume between MIIP and MNIP. Therefore, the different adsorption capacity between the MIIP and MNIP in the study could be due to imprinting effect and not due to the morphological difference.

Fig. 4 SEM image of MIIP (A) and MNIP (B).

Table 1 Characteristics of the porous structure of the Fe₃O₄, MIP and NIP

Sample	The specific surface area (m^2 g^−1)	Average pore diameter (nm)	Pore volumes (m^3 g^−1)
Fe3O4	113	1.8	0.211
MIP	261.5	3.4	0.293
NIP	284.6	3.6	0.276

---

3.2. Adsorption behavior of magnetic imprinted polymers

3.2.1. Effects of pH. The effect of pH on the adsorption was investigated within pH 3 to 7. The results show that the adsorption capacity of MIIP for Pb²⁺ and Cd²⁺ was highly dependent on the pH of the solution. The adsorption capacity of the Pb²⁺ and Cd²⁺ ions significantly increased with increased pH from 3 to 6 and then gradually decreased with further increased pH. Thus, the optimum pH was determined to be pH 6.0. This sorption behavior can be attributed to the competitive binding of H⁺ with Pb²⁺ or Cd²⁺ ions for –NH₂ at a low pH. Beyond pH 6, small amounts of Pb²⁺ or Cd²⁺ ions are bound due to the formation of insoluble Pb(OH)₂ and Cd(OH)₂.³⁵ Thus, pH 6 was chosen as the optimal pH in the following experiments.

3.2.2. Adsorption isotherm. Fig. 5A shows that the adsorption capacity of Pb²⁺ or Cd²⁺ ions rapidly increased with increased initial concentration but reached a plateau at 1.0 mg mL⁻¹ for MIIP and at 0.6 mg mL⁻¹ for MNIP. This finding was likely due to incompletely saturated specific binding sites at low concentrations. The MIIP maximum adsorption capacity estimated from the experimental data was 108 mg g⁻¹ for Pb²⁺ and 56 mg g⁻¹ for Cd²⁺, which was about two times that of MNIP, indicating that MIIP has a higher affinity for Pb²⁺ or Cd²⁺ ions than MNIP. As FTIR spectra showed, the observed high selectivity in MIIP was due to the presence of a large number of –NH₂, –OH, C–N, C–O and –COOH functional groups on the MIIP surface.

Fig. 5 Adsorption isotherms (A) and Langmuir isotherms (B) of Pb²⁺ or Cd²⁺ ions on MIIP and MNIP. Polymer weight, 30 mg; pH 6.0; volume and concentration, 10 mL of 0.2 mg mL⁻¹ to 1.2 mg mL⁻¹; binding time, 2 h.

The Langmuir equation was also used to evaluate the behavior of adsorbent:³⁰

where C_e (mg mL⁻¹) is the equilibrium concentration of Pb²⁺ or Cd²⁺ ions, Q (mg g⁻¹) is the adsorption capacity at equilibrium, K_d (mL mg⁻¹) is the adsorption–desorption equilibrium constant related to the binding energy, and Q_max (mg g⁻¹) is the maximum adsorption amount. The plot of C_e/Q versus C_e showed a good linear relationship (Fig. 6B) and a high correlation coefficient (r > 0.99), indicating that the experimental results of the Pb²⁺ or Cd²⁺ ion adsorption on sorbents fit well with the Langmuir adsorption model. The MIIP maximum adsorption capacity was calculated as 136.6 mg g⁻¹ for Pb²⁺ and 66.7 mg g⁻¹ for Cd²⁺. The experimental Q_max values for Pb(II) and Cd(II) were slightly different from the calculated ones. Thus, the adsorption behavior of Pb(II) and Cd(II) onto MIIP can be considered a monolayer adsorption.

Fig. 6 Effect of concentrations of EDTA (A), amount of MIIP (B), and elution time (C) on the recovery rate of the proposed method.

3.2.3. Adsorption selectivity. To investigate the selectivity of MIIP, MNIP, MIIP/Pb and MIIP/Cd, the adsorption capacities of MIIP, MNIP, MIIP/Pb and MIIP/Cd in a mixed solution containing Pb²⁺, Cd²⁺, Cu²⁺, Zn²⁺, Mg²⁺ was examined. Table 2 shows the data of the adsorption capacity Q, the distribution coefficient K_d, the selectivity coefficient k. It can be found that the doubly imprinted polymers (MIIP) had higher adsorption capacity (Q) for Pb(II) and Cd(II) than the singly imprinted polymers (MIIP/Pb and MIIP/Cd) as well as non-imprinted polymers (MNIP). The Pb(II) and Cd(II) adsorption capacity of MIIP was much higher than that of other ions due to imprinting. It can be concluded that the MIIP showed the following metal ion affinity order under competitive conditions: Pb(II) > Cd(II) > Cu(II) > Zn(II) > Mg(II). A comparison of the k values for the MIIP with the MNIP, MIIP/Pb and MIIP/Cd shows a significant increase in k values for Pb²⁺ and Cd²⁺ by double template imprinting approach, with the largest k values for Pb(II) (2.94), Cd(II) (2.66). As shown in Table 2, MIIP clearly demonstrates a strong ability to selectively adsorb Pb(II) and Cd(II) ions from mixed metal ions in an aqueous solution.

Table 2 Adsorption selectivity of MIIP and MNIP

Metal ion	Q (mg g^−1)		Kd (L g^−1)		k		k′
	MIIP	MNIP	MIIP	MNIP	MIIP	MNIP
Pb^2+	108	55.9	1.38	0.40	2.94	1.11	2.65
Cd^2+	56.2	31.5	1.25	0.42	2.66	1.17	2.27
Cu^2+	26.0	17.1	0.81	0.40	1.72	1.11	1.55
Zn^2+	24.9	17.0	0.70	0.38	1.49	1.06	1.41
Mg^2+	7.29	6.10	0.47	0.36

---

3.3. Desorption and repeated use

To investigate the reusability of the MIIP sorbents, the adsorption–desorption cycle was checked 10 times using the same MIIP sorbents and with 0.1 mol L⁻¹ EDTA as the eluting solvent. The results shows that the adsorption capacity of MIIP for Pb(II) and Cd(II) ions slowly decreased with increased cycle number. After seven cycles, the adsorption capacity of MIIP for Pb(II) and Cd(II) ions decreased by 9.6% and 11.9%, respectively, indicating that MIIP had good reusability and stability for Pb(II) and Cd(II) adsorption.

3.4. Optimization of magnetic SPE (M-SPE) conditions

The type and concentration of eluting agent, MIIPs amount, extraction time, elution time and sample volume affect the recovery of the extraction. These parameters were investigated by analyzing spiked samples (1.0 μg mL⁻¹) in order to get the optimum extracting conditions.

Elution experiments were performed with 0.1 mol L⁻¹ HCl, 0.1 mol L⁻¹ HNO₃, and 0.1 mol L⁻¹ EDTA. Among these solutions, 0.1 mol L⁻¹ of EDTA solution was found to be most effective for desorbing Pb(II) and Cd(II) ions from the loaded adsorbents. The reason is that the complexing action between the EDTA with Pb(II) and Cd(II) is in favor of the desorption of Pb(II) and Cd(II) ions from the MIIP. To achieve maximum recovery of Pb(II) and Cd(II) ions, the effect of different concentrations of EDTA, ranging from 0.01 mol L⁻¹ to 0.2 mol L⁻¹, on Pb(II) and Cd(II) recovery was investigated by ultrasound-aided elution. The results in Fig. 6A show that the recovery increased dramatically with increased EDTA concentration up to 0.10 mol L⁻¹, beyond which the recovery remained constant. Thus, an EDTA concentration of 0.10 mol L⁻¹ was selected for the succeeding experiments.

Different mounts of MIIP sorbent ranging from 20 mg to 70 mg were applied to 30 mL spiked samples. The results presented in Fig. 6B demonstrate that a high recovery ranging from 102% to 97% was obtained at an MIIP amount range of 30 mg to 50 mg. Further increase in the amount of sorbent led to a decrease in recovery. Thus, 50 mg of the sorbent was chosen as the optimum amount.

The extraction and elution time is a key factor in the M-SPE procedure. The extraction and elution time ranging from 0 min to 30 min was studied. The results showed that 2 min for a completing the extraction was achieved. For the elution time, as shown in Fig. 6C, the recovery increased as the elution time increased from 0 min to 10 min, then slightly declined from 10 min to 30 min, possibly because the Pb(II) and Cd(II) ions in the solution rebound to the sorbent. Thus, the optimum extraction and elution time was set at 2 min and 10 min, respectively.

The optimization experiment of sample volume was performed with 10–200 mL spiked samples. It was found that high recoveries were obtained when sample volume was equal or less than 100 mL. Therefore, 100 mL was selected as the break-through volume.

3.5. Effect of coexisting ions

The effect of various possible coexisting ions, such as SO₄²⁻, CO₃²⁻, NO₃⁻, PO₄³⁻, NH₄⁺, K⁺, Na⁺, Ca²⁺, Fe³⁺, Co²⁺, Mn²⁺, Ni²⁺, and Cr³⁺ was investigated. The tolerance limit of the coexisting ions was defined as the maximum concentration of foreign substances that can cause approximately ±5% relative error in determination. The following ion concentrations were found not to affect the absorption capacity: 1000-fold NO₃⁻, NH₄⁺, K⁺, and Na⁺; 500-fold Ca²⁺ and Fe³⁺; 200-fold Mn²⁺, Cr³⁺, SO₄²⁻, CO₃²⁻, and PO₄³⁻; and 100-fold Co²⁺ and Ni²⁺ ions, which imply that MIIP exhibits high selectivity for Pb(II) and Cd(II) in the presence of other metal ions. The main possible reasons for the high selectivity are that the recognition sites of the double template Pb(II) and Cd(II) imprinted sorbent for the chelating amine group and hydroxyl group are arranged in a suitable way for coordination-geometry selectivity, and that the hole size selectivity is dominated by Pb(II) and Cd(II) ions.^30,36 Anion, which strongly bind with Pb(II) and Cd(II) ions, can influence the adsorption of Pb(II) and Cd(II) ions. The order of effect is PO₄³⁻ > CO₃²⁻ > SO₄²⁻. Therefore, the sample can't use sulfuric acid and phosphoric acid to dissolve.

3.6. Validation and applications

A method based on M-SPE coupled to ICP-OES was established. Under the optimum conditions, an external aqueous calibrations in 0.1% nitric acid and standard addition calibrations in blank samples extract were constructed containing Pb(II) and Cd(II) ions between 0.8 to 5000 μg kg⁻¹ in order to study possible matrix effects. The calibration equation in aqueous was y = 0.00387 + 0.9863 c_Pb²⁺ (R² = 0.9998) and y = 0.00187 + 0.9983 c_Cd²⁺ (R² = 1) for Pb(II) and Cd(II), respectively. The calibration equation in matrix was y = 0.00202 + 0.9605 c_Pb²⁺ (R² = 0.999) and y = 0.00096 + 0.9718 c_Cd²⁺ (R² = 0.9992) for Pb(II) and Cd(II), respectively. There are no significant differences of the calibration curve slopes for Pb or Cd obtained both for in aqueous and for in matrix. The results showed that the matrix could be efficiently removed during the M-SPE pre-concentration stage. Therefore, the external aqueous calibration in 0.1% nitric acid can be used to determine Pb and Cd in practical samples after M-SPE pre-concentration procedure.

Limit of detection (LOD) and limit of quantification (LOQ) were calculated according to LOD = (3SD)/m and LOQ = (10SD)/m, where SD is the standard deviation of 10-replicate measurements of a procedural blank (acidified Milli-Q water treated as a sample) and m is the slope of the external aqueous calibration curve. The LODs were found to be 0.10 and 0.08 μg kg⁻¹ for the Pb(II) and Cd(II) ions, respectively. Similarly, the LOQs were found to be 0.34 and 0.27 μg kg⁻¹ for the Pb(II) and Cd(II) ions, respectively. These LODs and LOQs values are similar to those reported by other authors when using M-SPE combined with ICP-OES (0.18 μg L⁻¹ for Pb³⁷) or FAAS (0.11 μg L⁻¹ for Cd³⁸) detection. However, that ICP-MS will give a higher sensitivity by a factor of 100–1000 (0.004 μg L⁻¹ for Pb using ionic imprinted polymer based solid phase extraction).³⁷

The precision of the overall procedure was assessed by analyzing a contaminative soil (Section 2.7) 10 replicate. The determined mean values were 3.38 ± 0.11 and 1.08 ± 0.04 mg kg⁻¹ for Pb and Cd, respectively. The relative standard deviation (RSD) was 3.3% and 3.7% for Pb and Cd, respectively, indicating that the method had good precision. The accuracy of the method was evaluated by recovery tests at three spiked levels (2.0, 5.0, and 10.0 mg kg⁻¹) and by analysis of certified reference materials. The recoveries obtained from soil were 96.5 ± 3.3%, 95.4 ± 3.3% and 92.5 ± 3.3% for 2.0, 5.0, and 10.0 mg kg⁻¹ of Pb, and 98.8 ± 3.7%, 97.1 ± 3.7%, 94.8 ± 3.7% for 2.0, 5.0, and 10.0 mg kg⁻¹ of Cd, respectively, indicating that the developed method is reliable for determining Pb and Cd in environmental samples. Furthermore, GBW07402 soil certified reference materials were determined by duplicate. The determined mean values were 19.2 ± 0.7 and 0.068 ± 0.004 mg kg⁻¹ for Pb and Cd, respectively, which were coincident with the corresponding certified values 20.0 ± 3.0 and 0.071 ± 0.014 mg kg⁻¹ for Pb and Cd, respectively.

To further demonstrate the application of the method, cinnamon, lotus root and sweet potato samples (Section 2.7) were detected. The results are listed in Table 3, which shows that the recovery rates of Pb and Cd were in the range of 87.5% to 103% and 89.0% to 106%, respectively, indicating that the M-SPE procedure has a high selectivity for extracting Pb(II) and Cd(II) from food samples.

Table 3 Determined Pb(II) and Cd(II) values in the lotus root, sweet potato, cinnamon, and soil samples (mean ± SD, n = 3)

Sample	Analyte	Found (mg kg^−1)	Spiked (ng g^−1)	Sum	Recovery (%)
Lotus root	Pb	2.28 ± 0.08	10.0	11.40 ± 0.55	91.2
			25.0	23.58 ± 0.50	85.2
			50.0	47.98 ± 0.58	91.4
	Cd	0.18 ± 0.05	8.0	8.18 ± 0.42	100
			20.0	18.22 ± 0.47	90.2
			40.0	36.10 ± 1.12	89.8
Sweet potato	Pb	0.74 ± 0.02	10.0	10.84 ± 0.16	101
			25.0	23.04 ± 1.13	89.2
			50.0	47.09 ± 1.04	92.7
	Cd	0.06 ± 0.00	8.0	7.87 ± 0.27	97.6
			20.0	17.88 ± 0.45	89.1
			40.0	35.18 ± 1.47	87.8
Cinnamon (2-3899)	Pb	7.07 ± 0.24	10.0	15.96 ± 0.56	88.9
			25.0	29.70 ± 1.19	90.5
			50.0	49.27 ± 1.13	84.4
	Cd	0.60 ± 0.02	8.0	8.60 ± 0.15	100
			20.0	21.80 ± 1.30	106
			40.0	38.88 ± 0.82	95.7
Commercial cinnamon	Pb	12.3 ± 0.40	10.0	21.53 ± 0.50	92.3
			25.0	33.70 ± 1.15	85.6
			50.0	56.05 ± 1.50	87.5
	Cd	0.24 ± 0.00	8.0	8.24 ± 0.27	100
			20.0	21.24 ± 0.34	105
			40.0	39.16 ± 0.58	97.3
Contaminated soil	Pb	3.38 ± 0.11	10.0	12.03 ± 0.32	86.5
			25.0	26.23 ± 1.42	91.4
			50.0	48.63 ± 1.02	90.5
	Cd	1.08 ± 0.03	8.0	9.08 ± 0.20	100
			20.0	20.10 ± 0.63	95.1
			40.0	38.20 ± 1.25	92.8

---

The determined Pb(II) values in the lotus root, sweet potato, cinnamon (2-3899) and commercial cinnamon samples, which were 2.28 ± 0.08, 0.74 ± 0.03, 7.07 ± 0.25 and 12.30 ± 0.42 mg kg⁻¹, respectively, were much higher than Chinese standard (Pb ≤ 0.2 mg kg⁻¹ for potato crops and cinnamon). The determined Cd(II) values in the lotus root and sweet potato samples, which were 0.18 ± 0.01, and 0.06 ± 0.00 mg kg⁻¹, respectively, were lower than the Chinese standard (Cd ≤ 0.2 mg kg⁻¹ for potato crops). The determined Cd(II) values in the cinnamon (2-3899) and commercial cinnamon samples, which were 0.60 ± 0.02 and 0.24 ± 0.01 mg kg⁻¹, respectively, were higher than the Chinese standard (Cd ≤ 0.2 mg kg⁻¹ for cinnamon).

4. Conclusions

In this paper, a novel Pb–Cd–MIIP with double template and two functional monomers (4-VP and acrylate-modified S. platensis) were successfully synthesized by inverse microemulsion polymerization method. The sorbent exhibited high selectivity, large adsorption capacity, fast adsorption rate, and good reusability and stability for Pb(II) and Cd(II) ions. Pb–Cd–MIIP was successfully used with M-SPE materials for the rapid separation and enrichment of Pb(II) and Cd(II) in food and environmental samples, avoiding the time-consuming column passing and filtration steps in traditional SPE. The precision and accuracy of the presented method were satisfactory. The proposed method provided a rapid and effective tool for the extraction of trace amounts of Pb(II) and Cd(II) in all kinds of samples.

Acknowledgements

This work is financially supported by the Science and Technology Innovation Project of Guangdong provincial education department (no. 2012kjcx0104), the Science and Technology Planning Project of Guangdong province (no. 2011B020314011) and the Natural Science Foundation of Guangdong Province (no. S2011010004004; no. S2012040007710).

References

1. G. F. Nordberg, B. A. Fowler, M. Nordberg and L. T. Friberg, Handbook on the Toxicology of Metals, Academic Press Inc., 2007.
2. H. Needleman, Lead Poisoning, Annu. Rev. Med., 2004, 55, 209.
3. B. Hultberg, A. Andersson and A. Isaksson, Toxicology, 1998, 126, 203.
4. M. P. Waalkes and S. Rehm, Fundam. Appl. Toxicol., 1992, 19, 512.
5. O. Ciferri, Spirulina, Microbiol. Rev., 1983, 47, 551.
6. J. Wang and C. Chen, Biotechnol. Adv., 2009, 27, 195.
7. V. K. Gupta, A. Rastogi and A. Nayak, J. Colloid Interface Sci., 2010, 342, 533.
8. A. Çelekli, M. Yavuzatmaca and H. Bozkurt, J. Hazard. Mater., 2010, 173, 123.
9. L. Fang, C. Zhou, P. Cai, W. Chen, X. Rong, K. Dai, W. Liang, J. Gu and Q. Huang, J. Hazard. Mater., 2011, 190, 810.
10. A. Seker, T. Shahwan, A. E. Eroglu, S. Yılmaz, Z. Demirel and M. C. Dalay, J. Hazard. Mater., 2008, 154, 973.
11. N. Rangsayatorn, P. Pokethitiyook, E. S. Upatham and G. R. Lanza, Environ. Int., 2004, 30, 57.
12. S. V. Gokhale, K. K. Jyoti and S. S. Lele, Bioresour. Technol., 2008, 99, 3600.
13. J. Hu, M. C. Lo and G. H. Chen, Water Sci. Technol., 2004, 50, 139.
14. K. Cheng, Y. M. Zhou, Z. Y. Sun, H. B. Hu, H. Zhong, X. K. Konga and Q. W. Chen, Dalton Trans., 2012, 41, 5854.
15. M. K. Rofouei, A. Rezaei, M. Masteri-Farahania and H. Khanib, Anal. Methods, 2012, 4, 959.
16. H. Li, S. Bi, L. Liu, W. Dong and X. Wang, Desalination, 2011, 278, 397.
17. S. Sadeghi, H. Azhdari, H. Arabi and A. Z. Moghaddam, J. Hazard. Mater., 2012, 215–216, 208.
18. L. Chen, X. Zhang, L. Sun, Y. Xu, Q. Zeng, H. Wang, H. Xu, A. Yu, H. Zhang and L. Ding, J. Agric. Food Chem., 2009, 57, 10073.
19. Y. Ji, J. Yin, Z. Xu, C. Zhao, H. Huang, H. Zhang and C. Wang, Anal. Bioanal. Chem., 2009, 395, 1125.
20. M. Zhang, Z. Zhang, Y. Liu, X. Yang, L. Luo, J. Chen and S. Yao, Chem. Eng. J., 2011, 178, 443.
21. Y. Ren, M. Zhang and D. Zhao, Desalination, 2008, 228, 135.
22. X. Luo, S. Luo, Y. Zhan, H. Shu, Y. Huang and X. Tu, J. Hazard. Mater., 2011, 192, 949.
23. B. Guo, F. Deng, Y. Zhao, X. Luo, S. Luo and C. Au, Appl. Surf. Sci., 2014, 292, 438.
24. F. Aboufazeli, H. R. L. Z. Zhad, O. Sadeghi, M. Karimi and E. Najafi, Food Chem., 2013, 141, 3459.
25. T. P. Rao, R. Kala and S. Daniel, Anal. Chim. Acta, 2006, 578, 105.
26. U. Latif, A. Mujahid, A. Afzal, R. Sikorski, P. A. Lieberzeit and F. L. Dickert, Anal. Bioanal. Chem., 2011, 400, 2507.
27. S. Dai, M. C. Burleigh, Y. H. Ju, H. J. Gao, J. S. Lin, S. J. Pennycook, C. E. Barnes and Z. L. Xue, J. Am. Chem. Soc., 2000, 122, 992.
28. F. Dickert, P. Achatz and K. Halikias, Fresenius' J. Anal. Chem., 2001, 371, 11.
29. F. Xie, H. Xuan, Y. Ge, Y. Wang, T. Cao and K. Zhang, Chin. J. Anal. Chem., 2011, 39, 77.
30. F. Xie, G. Liu, F. Wu, G. Guo and G. Li, Chem. Eng. J., 2012, 183, 372.
31. J. Otero-Romaní, A. Moreda-Piñeiro, P. Bermejo-Barrera and A. M. Esteban, Microchem. J., 2009, 93, 225.
32. J. W. Kim, J. Y. Ko, J. B. Jun, I. S. Chang, H. H. Kang and K. D. Suh, Colloid Polym. Sci., 2003, 281, 157.
33. H. B. Hu, Z. H. Wang and L. Pan, J. Alloys Compd., 2010, 492, 656.
34. E. Birlik, A. Ersoz, E. Acykkalp, A. Denizli and R. Say, J. Hazard. Mater., 2007, 140, 110.
35. H. T. Fan, J. Li, Z. C. Li and T. Sun, Appl. Surf. Sci., 2012, 258, 3815.
36. G. Z. Fang, J. Tan and X. P. Yan, Anal. Chem., 2005, 77, 1734.
37. A. Otero-Romaní, A. Moreda-Piñeiro, P. Bermejo-Barrera and A. Martin-Esteban, Talanta, 2009, 79, 723.
38. M. G. Segatelli, V. S. Santos, A. B. T. Presotto, I. V. P. Yoshida and C. R. T. Tarley, React. Funct. Polym., 2010, 70, 325.

---