Recognition and selective adsorption of pesticides by superparamagnetic molecularly imprinted polymer nanospheres

Abstract

Magnetic molecularly imprinted polymers (MMIPs) were synthesized by cross-linking methyl methacrylate (MMA) and maleic anhydride (MA) copolymer, poly(MMA-co-MA), with triethylenetetramine (TETA) which has been used for the selective adsorption of pesticides phosalone, diazinon, and chlorpyrifos from aqueous solutions. To introduce superparamagnetic properties, amino-functionalized magnetic Fe₃O₄ nanoparticles (Fe₃O₄-NH₂) were prepared by a simple one-pot method. The MMIPs were characterized by FT-IR, XRD, VSM, TGA, TEM and SEM methods. After templates removal, the selectivity of the MMIPs was verified by direct adsorption of a single reference pesticide and mixed pesticides. The MMIPs indicated excellent recognition and binding affinity toward the tested pesticides, and the maximum adsorption capacity for phosalone, diazinon, and chlorpyrifos by Langmuir equation was 196.07, 192.30 and 172.41 mg g⁻¹, respectively. After four adsorption–desorption cycles, the MMIP maintained its adsorption capacity without significant loss. The prepared MMIPs possess the enhanced capacity and selectivity of the tested pesticides and have the value of practical applications.

---

Introduction

Nowadays, the presence of agrochemicals named as fertilizers, fungicides, and pesticides as residual compounds in every component of the environment, such as water, soil and also crops, vegetables, fruits and juices has been a great concern for health care.¹ Among various methods for monitoring of agrochemicals, molecular imprinting technology (MIT) based on a mechanism of molecular recognition is an interesting technique for the design of polymeric materials in order to protect people from the toxicity of such contaminations.^2,3 MIT, first constructed by Wulff⁴ and Mosbach,⁵ is a well-established and facile technique for synthesizing molecularly imprinted polymers (MIPs) with specific molecular recognition capacity. MIPs are a group of compounds that are usually prepared by polymerizing a mixture of a target molecule (template), functional and crosslinking monomers having vinyl or acrylic groups that can interact with functional groups of the template by covalent or non-covalent bonding.^6–8 The interaction of the binding site with the template should be very fast and reversible. Subsequent removal of the imprint molecule reveals binding sites that are complementary in size and shape to the imprinted analyte. The crosslinking agents control the morphology of the polymer matrix by providing mechanical stability and fix the guest-binding sites firmly in the desired structure. They also make the imprinted polymers insoluble in solvents and facilitate their practical applications. As a result, MIPs have binding sites consist of functional groups attached, capable of interacting with the template molecules and ideally the functional groups exist on the surface of the cavity left by the template, readily available for rebinding. Thus molecular imprinting is a technique for the synthesis of polymeric materials with predetermined ligand selectivity.^9–11 The imprinted polymers have been applied in an increasing number of applications where molecular binding events are of interest,^12–19 and because of low cost in industrial scale production and application.^20–22 Also, researchers have applied these techniques to hydrogel systems,²³ biologically significant target molecules,²⁴ controlled drug delivery systems,²⁵ and pesticides like 2,4-D atrazine,^26,27 bentazone,²⁸ and chlorotriazines.²⁹ Fe₃O₄ particles, as special biomolecule immobilizing carriers, have gained wide attraction. When Fe₃O₄ particles are encapsulated inside of MIPs, the resulting polymer material will have magnetically susceptible characteristics, and can be easily separated by external magnetic fields after they had finished their adsorption and recognition. The successful applications of magnetic MIPs in the recognition of biomolecules have also been reported by researchers.^30–46

The present work reports preparation of magnetic MIP (MMIP) to improve and simplify the determination of the remaining concentration of pesticides such as phosalone, chlorpyrifos, and diazinon in water which are used against various fungal pathogens in fruit trees and rice. For this purpose, a simple and efficient method was used for the preparation of hydrophilic MMIP, in which there is no contact between the vinyl monomers and template molecules. Because the selected templates bear functional groups that may inhibit or retard free radical polymerization or may not be stable at 80 °C, thus poly(MMA-co-MA) copolymer was prepared first in the absence of templates. Therefore, the synthesis process of poly(MMA-co-MA) should not affect the extraction selectivity of MMIP. In order to attach magnetic Fe₃O₄ nanoparticles chemically to the matrix of the copolymer, the Fe₃O₄ nanoparticles were functionalized with amine groups. The prepared copolymer, poly(MMA-co-MA), was cross-linked by TETA in the presence of Fe₃O₄-NH₂ and the template molecules to form nanospherical MMIP. After particle formation, the template molecules were removed by washing with methanol on a magnetic plate. The prepared MMIPs were characterized by FT-IR, X-ray diffraction (XRD), transmission electron microscopy (TEM), scanning electron microscopy (SEM), TGA, vibrating sample magnetometer (VSM), and its binding properties were investigated. The magnetic responsivity of MMIPs was strong enough to purify and separate MMIP by the external magnetic field. The adsorption kinetics, isotherms, selective recognition capacities, and regeneration of the MMIPs were investigated. Experimental results demonstrate that the MMIP are selective toward the target molecules and showed much better template affinity than the corresponding magnetic non-imprinted polymer (MNIP). Easy preparation, chemical stability, and recognition capacity of template molecule make this approach attractive and broadly applicable in separation and sensors.

Experimental

Materials

The pesticides of phosalone (P), chlorpyrifos (C) and diazinon (D) (Scheme 1) were supplied by Sobhan Darou Pharmaceutical Company, Iran. Ferric chloride hexahydrate (FeCl₃·6H₂O, >99%) as a single iron source, 1,6-hexamethylenediamine (HMDA), benzoyl peroxide (BPO), methyl methacrylate (MMA), maleic anhydride (MA), triethylenetetramine (TETA), triethylamine (TEA), anhydrous sodium acetate, ethylene glycol, methanol and THF were purchased from Fluka (Germany). All the chemicals were used without further purification.

Scheme 1 Chemical structures of template pesticides.

Preparation of superparamagnetic Fe₃O₄-NH₂ nanoparticles

Fe₃O₄-NH₂ nanoparticles were prepared via the facile solvothermal method as reported in the literature.⁴⁷ Briefly, HMDA, anhydrous sodium acetate, and FeCl₃·6H₂O in a weight ratio of 3.0 : 1.0 : 0.5 were dissolved in 30 mL ethylene glycol at 50 °C to give a transparent solution. This solution was then heated to 200 °C in a two-necked flask equipped with a condenser for 6 h. The prepared black magnetite nanoparticles were separated by applying an external magnet, washed several times using deionized water and methanol to effectively remove the solvent and unreacted HMDA, and then vacuum dried at 50 °C for 24 h. Naked Fe₃O₄ nanoparticles (MNPs) were prepared in the same procedure without the addition of HMDA.

Synthesis of poly(MMA-co-MA)

Free radical copolymerization of MA and MMA was carried out according to our previous report.⁴⁸ Briefly, 2.33 mL (0.020 mol) MMA (M_w = 100 g mol⁻¹), 2.0 g (0.020 mol) MA (M_w = 98 g mol⁻¹) and 50 mL THF were placed in a 250 mL three-necked round bottom flask equipped with a magnet stir bar, a condenser and an inlet tube for the inert gas. After degassing the mixture for 30 min by nitrogen, BPO (2.0 × 10⁻⁵ mol) was added to the reaction mixture at 80 °C and refluxed under these conditions for 8 h. Finally, the copolymer precipitate was obtained by pouring the reaction mixture into the nonsolvent of methanol and water mixture (1 : 2, v/v), washed with water thoroughly, and then dried under vacuum at 80 °C. The M_n and M_w of poly(MMA-co-MA) were 5.09 × 10³ and 9.34 × 10³ g mol⁻¹, respectively, with molar mass distribution (MMD) of 1.83.

Preparation of MMIPs

MMIP was prepared by reaction between poly(MMA-co-MA) as the functional polymer to form cross-links with TETA as the cross-linker to enhance the adsorption capacity in the presence of Fe₃O₄-NH₂ nanoparticles and a template pesticide in THF solvent and methanol porogen. A weight ratio of 1 : 1 : 1.5 : 0.5 from Fe₃O₄-NH₂ : TETA : poly(MMA-co-MA) : template has been used in the preparation of MMIP. For the synthesis of non-imprinted polymer (MNIP), no pesticide was used to serve as a template during preparation. Briefly, a mixture of poly(MMA-co-MA) (0.5 g) and aminated-Fe₃O₄ nanoparticles (0.16 g) in 30 mL THF was ultrasonicated for 15 min. The mixture was poured into a three-necked round-bottomed flask equipped with a condenser, a gas inlet tube, and a magnet stir bar. After raising the temperature to 80 °C, TEA (0.5 mL, 0.004 mol) was added as a catalyst and refluxed for 15 min under N₂ atmosphere. Then, the template diazinon (0.08 g) solution in 10 mL methanol was added to the above solution. Finally, TETA cross-linker (0.16 g) was added and the temperature was maintained at 80 °C for a further 4 h. The prepared MMIPs were extracted for 24 h in a Soxhlet apparatus using THF to remove the residuals of the copolymer and then washed with methanol to remove the template, and dried in vacuum oven at 60 °C. The same procedure was used to synthesize phosalone-MMIP (P-MMIP) and chlorpyrifos-MMIP (C-MMIP).

Adsorption experiments

The adsorptions were carried out in batch experiments at room temperature by changing binding time (20–120 min) and initial pesticides concentrations (0.2–1.2 mg mL⁻¹). Each experimental variable was under verification while the other one was constant at the optimum value. A certain amount of MMIP nanospheres (0.05 g) was placed in a 25 mL stopper conical flask, mixed with different concentrations of pesticides (0.01–0.06 g in 50 mL methanol/distilled water (1 : 1 v/v), calculated as 0.2–1.2 mg mL⁻¹). 25 mL of the prepared pesticide solutions in methanol/water was used for adsorption with stirring at 25 °C. Samples were obtained at specific time intervals for pesticide concentration measurements. The concentration of each pesticide in the methanol/water (1 : 1 v/v), was determined by UV-Vis spectrophotometer at maximum adsorption wavelength of 265 nm, 254 nm and 230 nm for phosalone, diazinon and chlorpyrifos, respectively and by using calibration curves which were made at different concentrations for each pesticide. The adsorption capacity (Q_e, mg g⁻¹) of MMIP and MNIP for each pesticide was calculated from eqn (1).where C₀ (mg mL⁻¹) is the initial concentration of pesticide, C_e (mg mL⁻¹) is the pesticide concentration of the supernatant solution after adsorption, V (mL) is the volume of the initial pesticide solution and W (g) is the mass of MMIP or MNIP. All tests were conducted in triplicate, and binding amounts were calculated based on a standard curve. In the case of the reuse, all MMIPs and MNIPs in this study were washed after absorption by methanol three times until there was no pesticide detected by UV-Vis spectrophotometer.

Characterization

UV-Vis spectra were recorded in solution using a Perkin Elmer, Lambda 800 spectrophotometer. FT-IR spectra were recorded on a Bruker Tensor 27 spectrometer using KBr pellets for solids. Particles surface morphology was examined using field emission scanning electron microscopy (FESEM) (Model: Hitachi S4160). Magnetic properties were measured at room temperature with a vibrating-sample magnetometer (VSM), model 155, PAR. Transmission electron microscopy (TEM) (Zeiss, EM10C, 80 kV) was used to obtain information on the particle size and morphology of template-MMIP. The gel permeation chromatography (GPC) measurements were conducted at 30 °C with a Perkin-Elmer instrument equipped with a differential refractometer detector. The columns used were packed with a polystyrene/divinylbenzene copolymer (PL gel MIXED-B from Polymer Laboratories) and THF was used as fluent at a flow rate of 1 mL min⁻¹. Calibration of the instrument was performed with monodisperse polystyrene standards.

Results and discussion

Synthesis and characterization

In this study, a copolymer instead of the monomer was used to prepare the magnetic-imprinted polymer through a single-step procedure. A schematic representation of the preparation procedure is shown in Scheme 2.

Scheme 2 Illustration of synthetic procedure for preparation of MMIP.

The prepared poly(MMA-co-MA) was imprinted by three pesticides as templates in the presence of Fe₃O₄-NH₂ nanoparticles and TETA as cross-linker. A weight ratio of 1 : 1 : 3 : 0.5 of Fe₃O₄-NH₂ : TETA : poly(MMA-co-MA) : template was selected for the preparation of MMIP. Assuming a weight ratio of 1 : 1 of MMA (100 g mol⁻¹) and MA (98 g mol⁻¹) in the poly(MMA-co-MA) chain, therefore, the weight ratio of MA/TETA is 1.5/1. The reactions between the amine groups of TETA and Fe₃O₄-NH₂ with the anhydride rings of the copolymer not only provide cross-links but also form carboxylic acid and amide groups which can immobilize pesticide molecules onto the MMIP beads via hydrogen bonding. Methanol was used as the porogenic solvent. The porogenic solvents should not only provide the environment for the cross-linking reaction but also produce a porous structure in the resulting polymers and influence the polymer's morphology and the strength of non-covalent interactions.^8,49 After preparation of MMIP, the template was removed by methanol and the eluted pesticide was detected by UV-Vis spectrophotometer at the corresponding wavelength. The concentration of each pesticide was determined by using maximum adsorption of each pesticide on the calibration curves which were prepared in methanol for different concentrations of the individual pesticide. The results validated that about 85% of the pesticides were successfully washed away from MMIPs, which confirm the efficient removal of the templates. After the template molecules were removed, the nanosphere particles were studied by SEM and TEM.

All materials synthesized in this study were characterized by FT-IR. Fig. 1 shows the FT-IR spectra of neat Fe₃O₄, Fe₃O₄-NH₂, poly(MMA-co-MA), and diazinon-imprinted polymer (D-MMIP). In FT-IR spectrum of the neat Fe₃O₄ in Fig. 1(a), the characteristic absorption band appeared strongly at 582 cm⁻¹ confirms the Fe–O vibration, and in Fig. 1(b) the absorption bands at 2927, 1625, 1465, and 875 cm⁻¹ are ascribed to the N–H stretching vibration, aliphatic C–H stretching, N–H scissoring and aliphatic C–N stretching, respectively which matched well with that of 1,6-hexanediamine. Comparison of FT-IR spectra of poly(MMA-co-MA) and D-MMIP indicates that characteristic absorption bands of the anhydride ring of poly(MMA-co-MA), in Fig. 1(c), at 1728, 1784 and 1854 cm⁻¹ disappeared after reaction with the amine groups, and the observed adsorption bands for D-MMIP, in Fig. 1(d), at ∼1703 cm⁻¹ and ∼3428 cm⁻¹ are related to the amide and carboxylic groups, respectively.

Fig. 1 FT-IR spectra of (a) neat Fe₃O₄, (b) Fe₃O₄-NH₂, (c) poly(MMA-co-MA), and (d) D-MMIP.

XRD analysis

For comparison, the crystal size of Fe₃O₄-NH₂ and D-MMIP nanoparticles were determined from the XRD pattern by using Scherrer's equation and the results were 9.7 and 19 nm, respectively. The structure of neat Fe₃O₄ and D-MMIP nanoparticles was characterized by XRD and the results, in Fig. 2(a), reveal the polycrystalline nature of Fe₃O₄ and D-MMIP. The characteristic peaks for Fe₃O₄ are observed at 2θ = 30.14, 35.48, 43.15, 53.4, 57.03 and 62.69°. Comparing the XRD patterns of neat Fe₃O₄ and D-MMIP, the information indicates that peak positions remained unchanged and particles crystalline structure has not been altered by surface modification.

Fig. 2 (a) XRD patterns of neat Fe₃O₄ and D-MMIP nanoparticles, (b) TGA curves of neat Fe₃O₄, Fe₃O₄-NH₂, and MMIP, (c) VSM curves of neat Fe₃O₄, Fe₃O₄-NH₂, MMIP and MNIP, and (d) dispersion of MMIP in water and magnetic separation.

Thermal analysis

Thermogravimetric analysis (TGA) was performed to further estimate the thermal stability of the prepared MMIP. Fig. 2(b) shows the weight of Fe₃O₄, Fe₃O₄-NH₂ and D-MMIP with the temperature changing from 30 °C to 600 °C. To estimate the amount of amine attached to the surface of Fe₃O₄ nanoparticles, the TGA analysis of the naked Fe₃O₄ and Fe₃O₄-NH₂ nanoparticles was conducted. As shown in Fig. 2(b), the weight lose for the naked Fe₃O₄ nanoparticles over the temperature range from 30 °C to 600 °C is only about 5%, which can be due to the loss of absorbed moisture in the sample. On the other hand, the TGA curve for Fe₃O₄-NH₂ shows two weight loss steps. Below 180 °C, the weight loss is quite small due to the removal of absorbed moisture. After that, there is a significant weight loss from 200 °C to 300 °C. Comparing the TGA curve of Fe₃O₄-NH₂ with that of Fe₃O₄ nanoparticles, the HMDA could be responsible for Fe₃O₄-NH₂ weight loss. When the temperature is higher than 330 °C, there is no distinct weight loss, implying that only iron oxide is present in the temperature range of 300–600 °C. The amount of HMDA bound on Fe₃O₄ nanoparticles is estimated from the percentage weight loss in the TGA curve, the average mass content of HMDA in Fe₃O₄-NH₂ is about 62%. In the case of D-MMIP, there is a small weight loss of about 5% at 80–130 °C due to absorbed moisture. After that, there is a steadily and significant weight loss from 150 °C up to 600 °C which can be due to decomposition and loss of organic fractions including the template.

Magnetic analysis

It is essential that magnetic sorbent should possess sufficient magnetic properties for easy separation during application. Therefore, to examine their magnetic properties, the magnetic nanoparticles were studied by a vibrating sample magnetometer (VSM) at room temperature, and the magnetic hysteresis loops are shown in Fig. 2(c). As can be seen in M–H curves, neat Fe₃O₄, Fe₃O₄-NH₂, MMIP and MNIP nanoparticles are superparamagnetic and have a magnetization saturation value of 70 emu g⁻¹, 56 emu g⁻¹, 35 emu g⁻¹, and 22 emu g⁻¹, respectively. In the case of MMIP, the decrease of the saturation value is most likely attributed to the existence of polymer layer on the surface of the magnetic core, confirming the core–shell morphology for the present MMIP. In comparison with the MNIP, MMIP has a higher magnetic strength which can be due to the existence of empty spaces in the MMIP network after removal of template molecules. Fig. 2(d) shows the facile separation of the MMIP nanospheres in aqueous solution by external magnetic field. The dispersed MMIP nanospheres in water can be easily collected and then again is readily re-dispersed with slightly shaking. The results reveal that the particles exhibit well magnetic responsible and re-disperse property, suggesting them a potential application of adsorbents. During the procedure of four cycles of adsorption–desorption, the magnetization saturation of MMIP is 34.6 emu g⁻¹, and still so strong that it could be separated easily from liquids by a magnet with a recovery of about 98.8%.

Particle size and morphology of MMIP and MNIP

Fig. 3 presents the FESEM images of Fe₃O₄-NH₂ (a), MNIP (b), and D-MMIP nanospheres after template removal ((c) and (d) at two different scales), and TEM image of D-MMIP nanospheres (e). As the SEM images show, the surface morphology of the prepared materials was different from each other. As shown in Fig. 3(a), the Fe₃O₄-NH₂ is mono-dispersed nanoparticles of spherical shape with narrow size distribution. It is indicated that the mean diameter of the prepared Fe₃O₄-NH₂ particles is about 27 nm. This confirms that coating of Fe₃O₄ nanoparticles with HMDA using one-pot method has not increased the particles size which is in close agreement with magnetic properties obtained by XRD. Moreover, amine modification has formed a coating on the surface of Fe₃O₄ spheres which is not only more favorable to react with active functional groups of polymers compared to neat Fe₃O₄, but also can act as a protective layer keeping interior magnetic nanoparticles from decay during application in water treatment. As can be seen in Fig. 3(b), the MNIPs present relatively larger particles with rougher surface compared to the surface presented by Fe₃O₄-NH₂. This can probably confirm that Fe₃O₄-NH₂ nanoparticles are captured into the cross-linked network of the composite. As shown in Fig. 3(c) and (d), the MMIP has also a rough surface with many pores. This can be mainly attributed to the cavities left after leaching of the imprinting molecules out of the cross-linked network, resulting in considerable recognition sites for the target pesticide forming on the surface of the MMIP. What's more, the rough surface structure is in favor of mass transfer and the formation of three-dimensional recognition sites. The TEM micrograph in Fig. 3(e) provides information on the particle size and morphology of D-MMIP. As can be seen, these particles have an average diameter of 20 nm with an approximately spherical shape.

Fig. 3 FESEM images of (a) Fe₃O₄-NH₂ nanoparticles, (b) MNIP, (c) and (d) D-MMIP after template removal at different scales, and (e) TEM image of D-MMIP.

Adsorption isotherms

The binding capacity for the tested templates of initial concentrations in the range of 0.2–1.2 mg mL⁻¹ was investigated for both MMIP and MNIP nanospheres. The adsorption isotherms in Fig. 4(a) show how the template was distributed on the MMIP or MNIP nanospheres. They show that the equilibrium adsorption capacities of both the MMIP and MNIP increased with increasing the initial pesticides concentration. The MMIP demonstrated much higher adsorptions than the MNIP within all the tested solution concentrations. This suggested that molecular recognition sites were successfully formed on the MMIP by the template molecules through network formation process and, therefore, MMIP nanospheres have better steric matching with the template molecules. However, the adsorption of all three templates by MMIP has reached the equilibrium level above 0.8 mg mL⁻¹. Four isotherm models were used to analyze the equilibrium values, the Langmuir (eqn (2)), Freundlich (eqn (3)), Dubinin–Radushkevich (eqn (4)) and Temkin (eqn (5)) models:

ln Q_e = ln Q_D − 2B_DRT ln(1 + 1/C_e) (4)

Q_e = B ln(AC_e) (5)

where C_e, Q_e, and Q_m are the equilibrium concentration (mg mL⁻¹), the equilibrium adsorption capacity (mg g⁻¹) and the maximum adsorption capacity of the templates (mg g⁻¹), respectively. K_L is the Langmuir constant (g mL⁻¹) and is obtained from the intercept of the plot between C_e/Q_e and C_e, n and K_F are the Freundlich constants (g mL⁻¹) and are obtained from the slope and intercept of the plot between ln Q_e and ln C_e, respectively. Q_D is the theoretical isotherm saturation capacity (mg g⁻¹). Dubinin–Radushkevich isotherm constant (B_D, mol² kJ⁻²) is predicted from the plot between ln Q_e and ln(1 + 1/C_e) (eqn (4)). B is the constant related to heat of sorption (J mol⁻¹), A is the Temkin isotherm equilibrium binding constant (L g⁻¹) and R is the universal gas constant (8.314 J mol⁻¹ K⁻¹) (eqn (5)). Parameters of the isotherm models and coefficients of determination (R²) were calculated to identify the conformity of the models with the experimental data; the results are presented in Table 1. The R² values indicate that the Temkin equation fits better with the experimental data for all the synthesized MMIP nanospheres. This result suggests that the heat of adsorption decreases linearly with the coverage because of adsorbent–adsorbate interactions and that the adsorption is characterized by a uniform distribution of binding energies, up to some maximum binding energy,⁵⁰ while MNIP nanoparticles obey with Dubinin–Radushkevich model.

Fig. 4 (a) Effect of initial concentration of template on the adsorption capacity (at 25 °C, after 60 min, 0.05 g adsorbent), and (b) effect of contact time on the adsorption capacity (at 25 °C, C₀ = 0.8 mg mL⁻¹).

Table 1 The isotherm model constants and correlation coefficients^a

Isotherm model	Parameters	P-MMIP	D-MMIP	C-MMIP	P-MNIP	D-MNIP	C-MNIP
a Temperature = 25 °C, contact time = 60 min, C0 = 0.2–1.1 mg mL^−1, adsorbent dosage = 0.05 g.
Langmuir	Qm (mg g^−1)	196.07	192.31	172.41	84.75	102.04	95.24
	KL (L mg^−1)	0.9272	0.9038	1	1.3563	0.9158	1.0500
	R^2	0.8622	0.9102	0.9000	0.9159	0.9183	0.9314
Freundlich	KF [(mg g^−1)(L mg^−1)^(1/n)]	93.1387	105.9174	90.0881	50.14	50.5161	50.3850
	n	0.8175	1.4170	1.4029	1.6260	1.2943	1.4484
	R^2	0.9480	0.9509	0.9599	0.9557	0.9665	0.9679
Dubinin–Radushkevich	BD	2.22 × 10^−4	3.91 × 10^−4	2.04 × 10^−4	1.84 × 10^−4	2.19 × 10^−4	2.08 × 10^−4
	QD	208.8127	379.6045	176.7909	92.8684	50.5161	100.6319
	R^2	0.9638	0.9661	0.9270	0.9634	0.9877	0.9852
Temkin	A (L g^−1)	10.3888	11.3117	10.5842	12.2432	9.9985	10.5774
	B	40.216	41.339	36.477	19.5460	21.066	20.6210
	R^2	0.9781	0.9729	0.9628	0.9405	0.9831	0.9745

---

Adsorption kinetics

In order to validate the ability of the MMIP to recognize the template, adsorption dynamic experiments of pesticides on MMIP and MNIP were carried out. Binding experiments as a function of time (20–120 min) were carried out to estimate how long it takes to reach the binding equilibrium for both MMIP and MNIP. The MMIP or MNIP was incubated with pesticide solution (0.8 mg mL⁻¹) for a certain time at room temperature and the adsorption capacity of MMIP and MNIP for templates was plotted as a function of time and the results are shown in Fig. 4(b). As can be seen, the adsorption capacity increased considerably with increasing contact time and reached the maximum after 60 minutes. Obviously, pesticides adsorption on the surface of MMIP was fast at the beginning, but after initial adsorption penetration of pesticides molecules into the MMIP nanospheres becomes difficult. According to the results shown in Fig. 4(b), the adsorption capacity of MMIP nanospheres at the equilibrium (Q_e) reached 82.8, 80.5 and 72.8 mg g⁻¹ for templates phosalone, diazinon, and chlorpyrifos, respectively, which is much higher than the equilibrium adsorption capacity value of 41.0 mg g⁻¹ for MNIP nanospheres. The adsorption trend of the MNIP is similar to the MMIP, but the adsorption capacity of the MNIP against the target pesticide molecule was relatively small. This can be explained by the facile connection of templates to the imprinting cavities of the MMIP. However, the MNIP nanospheres have no recognition site for the template and the adsorption capacity was mainly from non-specific adsorption. In order to validate the ability of the MMIP to recognize the pesticide template, the experimental data were tested using pseudo-first-order and pseudo-second-order kinetic equations, respectively, as follows:where Q_t and Q_e (mg g⁻¹) are the adsorption capacities at time t and equilibrium, respectively, K_L (min⁻¹) and n (g mg⁻¹ min⁻¹) are the equations rate constants and determined via the slopes and intercepts of the linear plots of log(Q_e − Q_t) versus t (eqn (6)), and t/Q_t versus t (eqn (7)). Parameters of the kinetic models and coefficients of determination (R²) were calculated to identify the conformity of the models with the experimental data and the results are tabulated in Table 2. According to the results, it is clear that the values of correlation coefficients (R²) for the first-order kinetic model are lower than the R² value of the second-order kinetic model. Therefore, the pseudo-second-order model is more suitable to describe the sorption process of phosalone, diazinon and chlorpyrifos onto the related MMIPs and the adsorption is in agreement with the chemical adsorption which is the rate-controlling step.^50–52

Table 2 Kinetic models and their statistical parameters^a

Kinetic model	Parameters	P-MMIP	D-MMIP	C-MMIP	P-MNIP	D-MNIP	C-MNIP
a Temperature = 25 °C, contact time = 0–120 min, C0 = 0.8 mg mL^−1, adsorbent dosage = 0.05 g.
Pseudo-first-order	Qe,exp. (mg g^−1)	78.9814	76.0752	72.9141	42.0253	40.2253	41.1253
	K1 (min^−1)	0.00967	0.01197	0.01296	0.02210	0.01566	0.03063
	Qe,cal. (mg g^−1)	37.0169	37.4541	40.0682	43.2016	40.8056	42.1793
	R^2	0.9454	0.8668	0.8563	0.9999	0.9985	0.9992
Pseudo-second-order	n (g mg^−1 min^−1)	5.42888	5.64334	4.88281	1.09481	1.01553	1.17507
	Qe,cal. (mg g^−1)	96.1538	90.9090	85.4500	76.3358	78.1250	75.18797
	R^2	0.9956	0.9917	0.9892	0.8930	0.8658	0.9135

---

Selectivity of MMIP

In order to investigate the selectivity of the prepared MMIP towards pesticides, batch selective adsorption experiments were carried out. Because of the similarity of molecular weights and volumes of phosalone, chlorpyrifos and diazinon, a test concentration of 0.8 mg mL⁻¹ was selected as a reference concentration to study the selectivity of the P-MMIP, C-MMIP, and D-MMIP towards phosalone, chlorpyrifos, and diazinon, respectively. To test the selective recognition of MMIP nanospheres, the same experiments were carried out with MNIP nanospheres. The adsorption capacity results of the mentioned MMIP nanospheres for different pesticides are illustrated in Fig. 5. It can be seen from the diagram that the adsorption capacity of each MMIP towards the target molecules was obviously higher than adsorption toward the other two pesticides, as was the imprinting factor and specific re-binding capacity. As expected, the P-MMIP, C-MMIP, and D-MMIP nanospheres displayed the highest adsorption capacity for phosalone, chlorpyrifos, and diazinon, respectively, among the other test pesticides, while the MNIP nanospheres had almost the same adsorption capacity for all the test pesticides. It shows the MMIP can selectively recognize the pesticide because of the existence of the imprinted sites, which only match the template molecule. The imprinting factor (α), defined as the ratio of adsorption quantities of imprinted (Q_MMIP) to non-imprinted polymers (Q_MNIP), and selectivity factor (β), defined as the ratio of adsorption quantities of each template (Q_P, Q_C or Q_D) to that of the competitor reference pesticide molecules (QRP), were defined by following equations, respectively:

Fig. 5 Adsorption selectivity of P-MMIP, D-MMIP and C-MMIP and MNIP for phosalone, diazinon, and chlorpyrifos (at 25 °C and after 60 min, C₀ = 0.8 mg mL⁻¹ and 0.05 g adsorbent).

The α and β factors were calculated from adsorption experiments and tabulated in Table 3. As illustrated in Fig. 5, the adsorption quantity for selected template was much higher relative to other two templates as reference pesticides. This shows that the MMIPs named as P-MMIP, C-MMIP, and D-MMIP possessed pronounced adsorption selectivity for the template phosalone, diazinon, and chlorpyrifos, respectively.

Table 3 Pesticides bound percentages and adsorption selectivity by MMIPs and MNIPs^a

Substrate	QP-MMIP	QD-MMIP	QC-MMIP	QMNIP	α	β
a Temperature = 25 °C, contact time = 60 min, C0 = 0.8 mg mL^−1, adsorbent dosage = 0.05 g.
Phosalone	90.7523	48.342	51.4350	48.8700	1.8570	QP/QD	1.9535
						QP/QC	2.0310
Diazinon	46.4543	94.8275	49.6700	45.1348	2.1000	QD/QP	1.9615
						QD/QC	2.1722
Chlorpyrifos	44.6754	43.6540	83.9905	46.2300	1.8167	QC/QD	1.6909
						QC/QP	1.6329

---

Regeneration feature

Fewer attempts have been made in the regeneration process of the MIPs, probably due to the relatively harsh template removal conditions. In this work, methanol was used to remove the template pesticide from each of the templates loaded MMIP (D-MMIP, C-MMIP, P-MMIP). Regeneration of the MMIP nanospheres after re-binding was performed in the same way. The experimental results are shown in Fig. 6. As can be seen, after four adsorption–desorption cycle, the adsorption capacity of the MMIPs for the related pesticide has only decreased by 5%, and the absorption capacity of the MMIPs was maintained around 85–90% in the fourth cycle.

Fig. 6 Influence of regeneration cycle on the adsorption capacity of the MMIPs.

Conclusions

In this study, a novel magnetic molecularly imprinted polymer (MMIP) was prepared in a simple method to recognize three template pesticides, namely: phosalone, diazinon, and chlorpyrifos. In preparation of MMIP, copolymer poly(MMA-co-MA) with anhydride functional groups for cross-linking and super-paramagnetic Fe₃O₄-NH₂ nanoparticles prepared by simple solvothermal method were used. The copolymer was cross-linked with TETA in the presence of Fe₃O₄-NH₂ and a template pesticide. The selectivity of the MMIP was verified by direct adsorption of a single reference pesticide and mixed pesticides. The MMIPs exhibited much higher adsorption capacity for the tested pesticides than the MNIP, and after removal of the templates by methanol; the MMIPs recognized the template pesticide molecules successfully. The adsorption rates followed the pseudo-second-order kinetics model and binding of templates to MMIP agreed well to the Temkin adsorption model. This work provides a simple and new approach to the combination of magnetic separation and molecularly imprinting technique. The prepared MMIP could be a promising prospect for the practical application in the separation of various hazardous pesticides.

References

1. J. L. Tadeo, C. Sánchez-Brunete and L. González, Analysis of Pesticides in Food and Environmental Samples, CRC Press, Taylor & Francis Group LLC, Boca Raton, 2008, pp. 1–35.
2. R. Garcia and A. M. C. Freitas, Am. J. Anal. Chem., 2011, 2, 16–25.
3. C. Lok and R. Son, Int. Food Res. J., 2009, 16, 127–140.
4. G. Wulff, A. Sarhan and K. Zabrocki, Tetrahedron Lett., 1973, 14, 4329–4332.
5. G. Vlatakis, L. I. Andersson, R. Müller and K. Mosbach, Nature, 1993, 362, 645–647.
6. M. J. Whitcombe, M. E. Rodriguez, P. Villar and E. N. Vulfson, J. Am. Chem. Soc., 1995, 117, 7105–7111.
7. B. Sellergren, Trends Anal. Chem., 1997, 16, 310–320.
8. P. A. Cormack and A. Z. Elorza, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2004, 804, 173–182.
9. G. Wulff, Chem. Rev., 2002, 102, 1–28.
10. I. A. Nicholls and J. P. Rosengren, Bioseparation, 2001, 10, 301–305.
11. K. Haupt and K. Mosbach, Chem. Rev., 2000, 100, 2495–2504.
12. P. Lulinski, Acta Poloniae Pharmaceutica–Drug Research, 2013, 70, 601–609.
13. R. Suedee, Pharm. Anal. Acta, 2013, 4, 264–287.
14. G. Vasapollo, R. D. Sole, L. Mergola, M. R. Lazzoi, A. Scardino, S. Scorrano and G. Mele, Int. J. Mol. Sci., 2011, 12, 5908–5945.
15. K. S. Lee, D. S. Kim and B. S. Kim, Biotechnol. Bioprocess Eng., 2007, 12, 152–156.
16. H. Yan and K. H. Row, Int. J. Mol. Sci., 2006, 7, 155–178.
17. G. Wulff, Angew. Chem., Int. Ed. Engl., 1995, 34, 1812–1832.
18. H. Zeng, Y. Wang, C. Nie, J. Kong and X. Liu, Analyst, 2012, 137, 2503–2512.
19. Q.-Q. Gai, F. Qu, Z.-J. Liu, R.-J. Dai and Y.-K. Zhang, J. Chromatogr. A, 2010, 1217, 5035–5042.
20. K. Haupt and K. Mosbach, Trends Biotechnol., 1998, 16, 468–475.
21. K. Haupt, Chem. Commun., 2003, 171–178.
22. V. B. Kandimalla and H. Ju, Anal. Bioanal. Chem., 2004, 380, 587–605.
23. M. E. Byrne, E. Oral, J. Zachary Hilt and N. A. Peppas, Polym. Adv. Technol., 2002, 13, 798–816.
24. J. Z. Hilt and M. E. Byrne, Adv. Drug Delivery Rev., 2004, 56, 1599–1620.
25. D. Cunliffe, A. Kirby and C. Alexander, Adv. Drug Delivery Rev., 2005, 57, 1836–1853.
26. M. Jakusch, M. Janotta, B. Mizaikoff, K. Mosbach and K. Haupt, Anal. Chem., 1999, 71, 4786–4791.
27. T. A. Sergeyeva, S. A. Piletsky, A. A. Brovko, E. A. Slinchenko, L. M. Sergeeva and A. V. El'Skaya, Analyst, 1999, 124, 331–334.
28. C. Baggiani, G. Giraudi, C. Giovannoli, A. Vanni and F. Trotta, Anal. Commun., 1999, 36, 263–266.
29. I. Ferrer, F. Lanza, A. Tolokan, V. Horvath, B. Sellergren, G. Horvai and D. Barceló, Anal. Chem., 2000, 72, 3934–3941.
30. C. J. Tan and Y. W. Tong, Anal. Chem., 2007, 79, 299–306.
31. Y. Li, X. F. Yin, F. R. Chen, H. H. Yang, Z. X. Zhuang and X. R. Wang, Macromolecules, 2006, 39, 4497–4499.
32. X. Kan, Q. Zhao, D. Shao, Z. Geng, Z. Wang and J.-J. Zhu, J. Phys. Chem. B, 2010, 114, 3999–4004.
33. X. Luo, Y. Zhan, Y. Huang, L. Yang, X. Tu and S. Luo, J. Hazard. Mater., 2011, 187, 274–282.
34. L. Li, X. He, L. Chen and Y. Zhang, Chem.–Asian J., 2009, 4, 286–293.
35. X. Wang, L. Wang, X. He, Y. Zhang and L. Chen, Talanta, 2009, 78, 327–332.
36. X. Kan, Z. Geng, Y. Zhao, Z. Wang and J.-J. Zhu, Nanotechnology, 2009, 20, 165601.
37. M. H. Lee, J. L. Thomas, M. H. Ho, C. Yuan and H. Y. Lin, ACS Appl. Mater. Interfaces, 2010, 2, 1729–1736.
38. M. Bouri, M. J. Lerma-García, R. Salghi, M. Zougagh and A. Ríos, Talanta, 2012, 99, 897–903.
39. H. B. Zheng, J. Z. Mo, Y. Zhang, Q. Gao, J. Ding, Q. W. Yu and Y. Q. Feng, J. Chromatogr. A, 2014, 1329, 17–23.
40. R. J. Ansell and K. Mosbach, Analyst, 1998, 123, 1611–1616.
41. Y. Zhang, R. Liu, Y. Hu and G. Li, Anal. Chem., 2009, 81, 967–976.
42. M. Ding, X. Wu, L. Yuan, S. Wang, Y. Li, R. Wang, T. Wen, S. Du and X. Zhou, J. Hazard. Mater., 2011, 191, 177–183.
43. Y. Ji, J. Yin, Z. Xu, C. Zhao, H. Huang, H. Zhang and C. Wang, Anal. Bioanal. Chem., 2009, 395, 1125–1133.
44. L. Xu, J. Pan, J. Dai, X. Li, H. Hang, Z. Cao and Y. Yan, J. Hazard. Mater., 2012, 233, 48–56.
45. F. F. Chen, X. Y. Xie and Y. P. Shi, J. Chromatogr. A, 2013, 1300, 112–118.
46. G. H. Yao, R. P. Liang, C. F. Huang, Y. Wang and J. D. Qiu, Anal. Chem., 2013, 85, 11944–11951.
47. F. Zhang, J. Jin, X. Zhong, S. Li, J. Niu, R. Li and J. Ma, Green Chem., 2011, 13, 1238–1243.
48. A. Masoumi and M. Ghaemy, eXPRESS Polym. Lett., 2013, 8, 187–196.
49. M. Lasáková and P. Jandera, J. Sep. Sci., 2009, 32, 799–812.
50. W. Shen, S. Chen, S. Shi, X. Li, X. Zhang, W. Hu and H. Wang, Carbohydr. Polym., 2009, 75, 110–114.
51. R. J. Umpleby, G. T. Rushton, R. N. Shah, A. M. Rampey, J. C. Bradshaw, J. K. Berch and K. D. Shimizu, Macromolecules, 2001, 34, 8446–8452.
52. C. Zhang, Y. Wang, Y. Zhou, J. Guo and Y. Liu, Anal. Methods, 2014, 6, 8584–8591.

---