Synthesis of surface ion-imprinted magnetic microspheres by locating polymerization for rapid and selective separation of uranium(VI)

Abstract

Uranium(VI) may pose a great threat to human health and the environment owing to its high chemical toxicity and radioactivity. A novel method is reported herein to synthesize surface ion-imprinted magnetic microspheres (SII-MM) for efficient removal of uranium(VI). Specifically, uranyl ion-imprinted polymer-functionalized Fe₃O₄@SiO₂ microspheres were prepared by surface-locating copolymerization of N-hydroxyethylacrylamide and 1-vinylimidazole. The effects of chemical composition, pH, adsorbent dose, competing ions and initial concentration on the adsorption of uranyl ions were evaluated. The exothermic spontaneous adsorption kinetically followed a pseudo-second-order model, and the process of SII-MM could reach equilibrium with a capacity of 146.41 mg of U per g within 1.0 min at pH 5.0 and 298.15 K. Compared with non-imprinted composites, SII-MM showed higher selectivity, faster kinetics, and larger capacity for uranyl adsorption. This work indicates that the SII-MM can be used as a promising adsorbent to effectively remove uranium(VI) from aqueous solutions.

---

1. Introduction

Uranium is a heavy metal associated with the nuclear industry and weapons that may cause severe adverse effects to human health because of its high chemical toxicity and radioactivity.^1,2 Therefore, the efficient removal of uranium from aqueous solutions has received great attention, especially in the fields of materials development and environmental science. In the past decades several methods based on extraction,³ precipitation,⁴ ion exchange,⁵ flotation⁶ and adsorption,^7–9 have been developed for removal of uranyl ions from aqueous solutions. Among them, adsorption is the most often utilized owing to its high efficiency, simple operation and the availability of different adsorbents. Silica-based materials are the most common adsorbents due to their low reactivity and thermal stability,¹⁰ and mesoporous silica^11–13 and silica nanotubes¹⁰ have been modified with various groups to interact with metallic cations. However, filtration or centrifugation is commonly required to separate the adsorbents from the aqueous media after adsorption, which is time-consuming.¹⁴

This problem can be solved when employing magnetic adsorbents, which can be easily separated from the system by an external magnetic field.^15–18 For example, Wang et al.¹⁹ reported bisphosphonate-modified magnetic nanoparticles to remove uranyl ion from water and blood; and Zhao et al.²⁰ synthesized amidoxime-functionalized magnetic microspheres to pre-concentrate and separate uranium(VI) from aqueous solution. Nevertheless, most traditional magnetic adsorbents lack of highly selective binding sites and fast adsorption kinetics. To overcome these drawbacks, ion-imprinting technique has been introduced into the adsorption field by copolymerization and cross-linking of functional monomers in the presence of template ions.^21,22 Selectivity is attributed to the affinity of the ligand towards uranyl ions and the size of the generated cavities.^22,23 However, ion-imprinted materials by bulk polymerization exhibit low rebinding capacity and poor binding site accessibility, because binding sites are almost deeply embedded into the polymer matrix.^23,24

Surface ion-imprinting technique has the advantages of high selective binding sites, fast adsorption kinetics and quick mass transfer.^23,25,26 However, there are only a few reports about preparing surface ion-imprinted magnetic composites for removal of uranium(VI) until now. For instance, Sadeghi et al.²⁷ reported ferromagnetic nanoparticles with an imprinted polymer coating for selective extraction of uranyl ions; and recently Liu et al.²⁸ prepared magnetic ion-imprinted composites with a core–shell structure by copolymerization of a ternary complex, which displayed high adsorption capacity and selective separation of uranium(VI). Obviously, surface ion-imprinted magnetic composites are promising adsorbent materials for efficient removal of uranium(VI). Therefore, there is still great significance in the development of new methods to fabricate novel surface ion-imprinted magnetic composites for efficient separation of uranium(VI).

Recently Shamsipur et al.²⁹ grafted ion-imprinted polymers on the surface of silica gel particles through covalently surface-bound initiators, and Kim et al.³⁰ fabricated TiO₂ nanoparticles with zwitterionic polymer brushes by surface-mediated seeded emulsion polymerization. Motivated by these findings, we report herein a novel method to synthesize surface ion-imprinted magnetic microspheres by locating polymerization for efficient removal of uranium(VI). Specifically, uranyl ion-imprinted polymer-functionalized Fe₃O₄@SiO₂ microspheres were prepared by surface-locating copolymerization of N-hydroxyethylacrylamide (HEMAA) and 1-vinylimidazole (VI) (Scheme 1). During polymerization, the hydrophobic initiator, 2,2′-azoisobutyronitrile (AIBN), dispersed onto the surface of Fe₃O₄@SiO₂–C=C microspheres due to hydrophobic interaction in the aqueous solution, which may control the polymerization initiation and propagation on the surface. Importantly, the polymers would be covalently bound onto microsphere surface because of polymerization participation of –C=C groups of Fe₃O₄@SiO₂–C=C microspheres. HEMAA and VI were selected as functional monomers, because carbonyl and amino groups coordinate well with uranyl ions.^31–33 It is expected that the obtained novel surface ion-imprinted magnetic microspheres (SII-MM) could achieve efficient removal of uranium(VI) from aqueous solution.

Scheme 1 Schematic depiction of the synthesis of surface ion-imprinted magnetic microspheres (SII-MM).

2. Experimental

2.1 Materials and reagents

N-Hydroxyethylacrylamide (HEMAA, TCI Development Co., Ltd., CP) was purified by dissolving with water then passing through a column with aluminum oxide base and stored at −20 °C prior use. 1-Vinylimidazole (VI, Sinopharm Chemical Reagent Co., Ltd., CP) was distilled under reduced pressure to eliminate the inhibitor and stored at −20 °C prior use. 2,2′-Azoisobutyronitrile (AIBN, Sinopharm Chemical Reagent Co., Ltd., CP) was re-crystallized from methanol and dried in a vacuum oven at room temperature. The microsphere Fe₃O₄@SiO₂–C=C was prepared according to the modified literature method^34–36 (Fig. S1, ESI†). UO₂(NO₃)₂·6H₂O (AR) was purchased from Fluka Co., Ltd. N,N′-Methylenebisacrylamide, EDTA, hydrochloric acid, FeCl₃·6H₂O, CH₃COONa, polyethylene glycol, sodium dodecyl sulfonate, ethylene glycol, ethanol, NH₃·H₂O, tetraorthosilicate, 3-(methacryloxy)propyltrimethoxysilane, ZnCl₂, Ni(NO₃)₂, CoCl₂, Pb(NO₃)₂, NaOH, NaCl, NaNO₃, Na₂SO₄ and Na₃PO₄ were purchased from Sinopharm Chemical Reagent Co., Ltd. and were of analytical reagent grade.

2.2 Characterization methods

Fourier transform infrared (FT-IR) spectra were recorded on a Varian-1000 spectrometer. Field-emitting scanning electron microscopy (SEM) observations were performed by using a HITACHI S-570 microscope operated at an accelerating voltage of 15 kV. Transmission electron microscopy (TEM) images were taken with a FEI Tecnai G20 electron microscope (accelerating voltage, 200 kV). Elemental mapping was taken with a FEI Tecnai G² F20 electron microscope. Energy-dispersive X-ray (EDX) analysis was carried out by a Hitachi S570 scanning electron microscope with an EDAX-PV 9100 energy dispersion X-ray fluorescence analyzer. Magnetization curves were performed on a Lakeshore 7407 vibrating sample magnetometer (VSM) with an applied magnetic field of 20 kOe. The concentration of uranyl ions was determined by using Thermo 6300 Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES). X-ray photoelectron spectroscopy (XPS) was carried out by ESCALAB 250Xi XPS. Surface Areas were determined by ASAP 2020 V3.01 H analyzer.

2.3 Preparation of SII-MM

A typical procedure is described below: Fe₃O₄@SiO₂–C=C particles (0.100 g), HEMAA (0.410 g, 3.56 mmol), VI (0.200 g, 2.12 mmol), N,N′-methylenebisacrylamide (MBA, 0.077 g, 0.499 mmol) and AIBN (0.020 g, 0.122 mmol) were added to uranyl solution (10⁻⁴ mol L⁻¹, 20 mL) with continuous stirring at room temperature for 4 h. The mixture was then heated to 343.15 K under argon atmosphere for 24 h. Afterwards, the uranyl ions were removed by washing with 150 mL of EDTA solution (0.1 mol L⁻¹). Then, the product was washed three times with 50 mL of distilled water and lyophilized for 12 h. The recipe of the copolymerization was changed to obtain the SII-MM with different composition (Table 1). In order to prepare surface non-imprinted magnetic microspheres (SNI-MM), a similar experiment was performed but in the absence of uranyl ions.

Table 1 Synthesis of surface ion-imprinted and non-imprinted polymer functionalized magnetic microspheres

Sample^a	UO2^2+ (mol L^−1)	HEMAA (g)	VI (g)	MBA (g)	AIBN (g)	Fe3O4@SiO2–C=C (g)
a Surface ion-imprinted magnetic microspheres with different content of 1-vinylimidazole: 0.2 g, 0.1 g and 0.05 g for SII-MM1, SII-MM2, and SII-MM3, respectively; SNI-MM is surface non-imprinted microspheres with the same recipe of SII-MM1.
SII-MM1	10^−4	0.41	0.20	0.077	0.02	0.10
SII-MM2	10^−4	0.41	0.10	0.077	0.02	0.10
SII-MM3	10^−4	0.41	0.05	0.077	0.02	0.10
SNI-MM	0	0.41	0.20	0.077	0.02	0.10

---

2.4 Uranium adsorption experiments

All adsorption experiments were carried out in polyethylene tubes. A typical procedure is described below: the adsorbents (2.5 mg) were added in 5 mL of uranyl ion solution (10⁻⁴ mol L⁻¹). After being shaken until the achievement of adsorption equilibrium, the solid phase was separated from the suspension by using a permanent magnet. The concentration of uranyl ions in the supernatant was determined by ICP-AES. The amounts of uranyl ions onto the composites (q_e) and adsorption efficiency (AE) of uranyl ions were calculated according to eqn (1) and (2), respectively:where C₀ and C_e (mg L⁻¹) are the initial and residual concentration of uranium(VI) in solution, respectively. V (L) is the volume of aqueous solution, and M (g) is the adsorbent weight.

Kinetic studies were conducted using 10⁻⁴ mol L⁻¹ of uranium(VI) at pH 5.0 and 298.15 K. The effects of chemical composition, solution pH, adsorbent dose, competing ions, salts, contact time (0–40 min) and temperature on uranium(VI) adsorption were studied. Isotherm experiments were conducted by using different initial concentrations of uranium(VI) at 298.15 K and pH 5.0.

In addition, to explore adsorbent's application to real samples, lake water (Dushu Lake, Suzhou) and tap water were also chosen to prepare uranyl ion solutions (23.8 ppm) for the adsorption.

2.5 Desorption and regeneration studies

To evaluate the reusability of the SII-MM, adsorption of uranium(VI) and regeneration of SII-MM–UO₂²⁺ complex were performed in five consecutive cycles. After saturated adsorption, SII-MM–UO₂²⁺ complex was washed with 0.1 mol L⁻¹ EDTA solution until uranyl ions were almost removed, and then washed with deionized water three times and lyophilized for reuse.

In order to investigate preconcentration factor, 0.5 mg UO₂(NO₃)₂·6H₂O were dissolved in different volumes (50–400 mL) of deionized water, and then 10 mg SII-MM1 were added. After being shaken until the achievement of adsorption equilibrium, the solid phase was separated from suspension and washed with deionized water, and was finally eluted with 2.5 mL of 0.1 mol L⁻¹ EDTA solution. The concentrations of uranyl ions in the supernatant and EDTA solution were determined by ICP-AES, respectively. The preconcentration limit is the concentration of original uranyl aqueous solution when the recovery is just less than 95%, and the preconcentration factor is the ratio of the volume of uranyl aqueous solution to EDTA volume.

3. Results and discussion

3.1 Characterization of SII-MM

The novel SII-MM was prepared by locating copolymerization of HEMAA and VI (Scheme 1). The hydrophobic initiator AIBN locates the polymerization initiation and propagation on the surface to form the ion-imprinted materials. The ion-imprinted polymer is covalently bound onto the surface of magnetic microspheres due to the polymerization participation of –C=C groups of Fe₃O₄@SiO₂–C=C microspheres (Fig. S1, ESI†). The morphology was characterized by SEM and TEM. The microspheres had a clear core–shell structure with a diameter of ∼300 nm (Fig. 1A). Similar results were obtained for the non-imprinted magnetic composites SNI-MM (Fig. 1B).

Fig. 1 SEM images of (A) SII-MM1 and (B) SNI-MM. Scale bar: 500 nm. Individual microspheres at higher magnification are shown as the inset TEM images in (A) and (B) (scale bar: 100 nm). (C) FTIR spectrum of: (a) Fe₃O₄@SiO₂–C=C and (b) SII-MM1; and (D) EDX spectrum of SII-MM1.

The surface ion-imprinted magnetic microspheres were also characterized by FT-IR, XPS spectrum and EDX spectra. Compared with Fe₃O₄@SiO₂–C=C (Fig. 1C, trace (a)), the band occurred for SII-MM at 3422 cm⁻¹ corresponding to the stretching of amino group, while vibration bands of C=N and C–N appeared at 1526 cm⁻¹ and 1221 cm⁻¹, respectively (Fig. 1C, trace (b)). XPS spectrum of C 1s of SII-MM1 can be curve-fitted with seven peak components attributed to C=C, C–C, C–N, C=N, C–O, N–C=O and C=O species (Fig. S2, ESI†), which indicated that HEMAA and VI were successfully coated onto the microspheres by the locating polymerization. EDX spectrum further demonstrated this point. The appearance of N element indicated that HEMAA and VI were copolymerized on the surface of Fe₃O₄@SiO₂ microspheres (Fig. 1D). Meanwhile, the absence of U element suggested that the imprinted magnetic composites have been washed completely by EDTA solution.

In addition, the special saturation magnetization loops are depicted in Fig. S3 (ESI†). In comparison with Fe₃O₄@SiO₂–C=C microspheres, there was a marked decrease of saturation magnetization for SII-MM, which might be attributed to the existence of the polymer on the surface. However, SII-MM could still be easily separated from the aqueous solution by placing a magnet near the vessel. This property may enable SII-MM to efficiently separate uranyl ions from aqueous solution.

3.2 Effect of polymer chemical composition on the adsorption of uranyl ions

The selectivity of ion-imprinted materials is associated with the affinity of the ligand with the imprinting metal ion.^18,19 In order to investigate this effect, SII-MM with different surface chemical composition was used in the adsorption experiments (Fig. S4, ESI†). 5 mL of uranyl ion solution (10⁻⁴ mol L⁻¹) was mixed with 2.5 mg of SII-MM until achieving adsorption equilibrium at 298.15 K. Results are listed in Table 2 and Fig. S5 (ESI†). The surface nitrogen content of SII-MM played an important role in the adsorption: the more nitrogen content, the shorter equilibrium time and higher adsorption efficiency (Table 2). It should be noted that adsorption could reach the equilibrium within 1.0 min for SII-MM1 with 91.6% of adsorption efficiency. Hence, SII-MM1 was selected for the further studies.

Table 2 Effect of surface chemical composition of SII-MM on the adsorption of uranyl ions. (Experiment condition: 5 mL solution, 10⁻⁴ mol L⁻¹ uranyl ion, 298.15 K, 0.5 g L⁻¹ SII-MM.)

Sample^a	Surface nitrogen content^b (%)	Equilibrium time (min)	Adsorption efficiency (%)
a Surface ion-imprinted magnetic microspheres with different content of 1-vinylimidazole: 0.2 g, 0.1 g and 0.05 g for SII-MM1, SII-MM2, and SII-MM3, respectively.b Determined by EDX spectra (Fig. S4, ESI†).
SII-MM1	9.80	1.0	91.6
SII-MM2	6.92	5.0	78.6
SII-MM3	1.85	25.0	33.1

---

3.3 Effects of pH, adsorbent dose, competing ions and salts on the adsorption of uranyl ions

In order to find the optimal conditions, the effects of pH, adsorbent dose and competing ions were investigated. The influence of pH was first investigated on the adsorption of uranyl ions by SII-MM1. The result is shown in Fig. 2A. The adsorption efficiency increased with increasing pH until pH 5.0 and then slightly decreased, which may be associated with the new speciation of uranyl ions and imidazolyl groups at different pH: at low pH, imidazolyl groups should be protonated, which may be detrimental to the adsorption; when pH > 5, the new complex species of uranyl ions would be formed with –OH or CO₃²⁻ anions,^33,37 which may lead to the low binding capability between UO₂²⁺ and SII-MM1. Therefore, pH 5.0 was selected for further assays. In addition, the effect of adsorbent dose was depicted in Fig. 2B, as can be seen the adsorption efficiency reached the equilibrium at 0.5 g L⁻¹ of adsorbent dose, which could be selected as the optimum dose.

Fig. 2 Effect of (A) pH, (B) adsorbent dose on adsorption efficiency (AE) of uranyl ions by SII-MM. (Experimental conditions: 5 mL solution, 10⁻⁴ mol L⁻¹ uranyl ion, (A) 298.15 K, 0.5 g L⁻¹ SII-MM1; (B) 298.15 K, pH 5.0.)

To test the selectivity of SII-MM, competitive experiments were conducted with the other ions (such as Zn²⁺, Ni²⁺, Co²⁺, and Pb²⁺) at the same concentration of uranyl ions. Distribution ratios (K_d) of imprinted and non-imprinted composites were calculated by eqn (3):

Selectivity coefficient (β) was determined by eqn (4):

where K_UO2²⁺ and K_Mⁿ⁺ represent the distribution ratio of uranyl ions and other metal ions in aqueous solution, respectively. The larger β suggests a stronger selectivity for uranyl ions.

To study the effect of imprinting on the adsorption, relative selectivity coefficient (β_r) was calculated by eqn (5):

where β_imprinted and β_{non-imprinted} are the selectivity coefficient of SII-MM and SNI-MM, respectively. β_r indicates adsorption affinity of recognition sites of SII-MM to the imprinting ions.

The selective adsorption parameters for SII-MM and SNI-MM are summarized in Table 3. The distribution ratio (K_d) of SII-MM for uranyl ions was 411 time higher than that of SNI-MM, and the selectivity coefficients (β) of SII-MM were far larger than that of SNI-MM, suggesting a much stronger binding ability and selectivity of SII-MM for uranyl ions. Furthermore, the relative selectivity coefficient (β_r) is far greater than 1, which indicates higher affinity of imprinted composites to uranyl ions than that of non-imprinted materials. All the observations should be ascribed to the specific recognition cavities for uranyl ions created by surface ion-imprinted technique. Thus, results showed that SII-MM had a strong selective ability to adsorb uranyl ions from aqueous solution in comparison with the other metal ions.

Table 3 Selective adsorption of uranium(VI) on SII-MM and SNI-MM. (Experiment condition: 5 mL solution, 10⁻⁴ mol L⁻¹ uranyl ion, 10⁻⁴ mol L⁻¹ each ion, 0.50 g L⁻¹ adsorbent dose, pH 5.0, 298.15 K.)

Metal ions	Distribution ratio (Kd, L g^−1)		Selectivity coefficient (β)		Relative selectivity coefficient (βr)
	SII-MM	SNI-MM	SII-MM	SNI-MM
UO2^2+	591.610	1.438	—	—	—
Zn^2+	16.289	0.436	36.320	3.300	11.01
Ni^2+	24.220	0.882	24.427	1.630	14.99
Co^2+	11.520	0.346	51.355	4.156	12.36
Pb^2+	3.838	0.604	154.145	2.380	64.77

---

The effects of salts (such as NaCl, NaNO₃, Na₂SO₄ and Na₃PO₄) on the adsorption of uranyl ions were investigated. The experiments were conducted with the other ions at 500–1500 times concentration of uranyl ions. The tolerance limits was defined as the largest amount of interfering ions causing a change of recovery of uranyl ions less than 5% compared to the one without any interference. Results in Table 4 show that the tolerance limits of electrolytes are very high (>500), which indicates that salts commonly present in water samples have little effect on the adsorption of uranyl ions.

Table 4 Tolerance limits of different electrolytes

Ion	Electrolyte	[Ion]/[UO2^2+]
Cl^−	NaCl	800
NO3^−	NaNO3	500
SO4^2−	Na2SO4	1000
PO4^3−	Na3PO4	1200

---

3.4 Effects of temperature and adsorption thermodynamics

In practice, Gibbs free energy and entropy factors are always used to determine which process will occur spontaneously. Experiments were carried out in the temperature range of 288.15–338.15 K with 10⁻⁴ mol L⁻¹ of the initial uranyl concentration, and the adsorption time was long enough to achieve equilibrium. Results are depicted in Fig. 3. The removal amount (q_e) of uranyl ions decreased with temperature increasing for SII-MM and SNI-MM, and SNI-MM was more sensitive to temperature for SNI-MM (Fig. 3A). The enthalpy change (ΔH), entropy change (ΔS) and Gibbs free energy change (ΔG) of imprinted and non-imprinted composites were estimated using the equilibrium constant K_c (q_e/C_e), which depends on temperature:³⁸

ΔG = −RT ln K_c (7)

where R and T are the ideal gas constant (8.314 J mol⁻¹ K⁻¹) and the absolute temperature, respectively. ΔH and ΔS can be obtained from the slope and intercept of the plot of ln K_c versus 1/T (Fig. 3B). Thermodynamic parameters are summarized in Table 5.

Fig. 3 (A) Effect of temperature on the adsorption of uranyl ions by SII-MM1 (trace (a)) and SNI-MM (trace (b)); and (B) plots of ln K_c versus 1/T for SII-MM1 (trace (a)) and SNI-MM (trace (b)). (Experimental conditions: 5 mL solution, 10⁻⁴ mol L⁻¹ uranyl ions, 0.5 g L⁻¹ adsorbents, pH 5.0.)

Table 5 Thermodynamic data for uranyl ion adsorption by SII-MM1 and SNI-MM

	ΔH (kJ mol^−1)	ΔS (J mol^−1 K^−1)	ΔG (kJ mol^−1)
			288.15 K	298.15 K	308.15 K	318.15 K	328.15 K	338.15 K
SII-MM1	−18.63	−20.88	−12.61	−12.40	−12.19	−11.98	−11.77	−11.57
SNI-MM	−35.80	−93.60	−8.83	−7.89	−6.96	−6.02	−5.09	−4.15

---

From Table 5, the negative ΔH indicated that adsorption was exothermic, and the negative ΔS reflected a decrease in the randomness during the adsorption process, which might be ascribed to the stable structure formed by the combination of uranyl ions and binding sites of magnetic composites. The values of ΔG of both SII-MM1 and SNI-MM were negative, which suggested that the process was spontaneous within the temperature range. The more negative ΔG for SII-MM1 indicated a more accessibility of adsorption.

3.5 Effects of contact time and adsorption kinetics

To understand the effect of contact time on uranyl adsorption, the adsorption kinetics experiments were conducted at pH 5.0. The relationship between adsorption amount (q_t) and contact time (t) is depicted in Fig. 4. The adsorption of imprinted composites could reach the equilibrium within 60 seconds (Fig. 4A), while 30 min were needed for the non-imprinted materials (Fig. 4B). The adsorption rate was analyzed by two common semi-empirical kinetic models: the pseudo-first-order equation proposed by Lagergren,³⁹ and pseudo-second-order equation by Ho and Mckay.^40,41

Fig. 4 Effect of contact time on the adsorption of uranyl ions by (A) SII-MM and (B) SNI-MM. (Experimental condition: 5 mL solution, 10⁻⁴ mol L⁻¹ uranyl ions, 0.5 g L⁻¹ adsorbents, pH 5.0, 298.15 K).

Pseudo-first-order kinetic equation describes the relationship between the adsorption rate and adsorption amount q_t at time t:

where k₁ (h⁻¹) is the pseudo first order kinetic constant. k₁ and q_e can be determined from the slope and intercept of the plot of log(q_e − q_t) versus t, respectively (Fig. S6A and C†).

Pseudo-second-order model is expressed as the following eqn (9):

where k₂ (g mg⁻¹ h⁻¹) is the second order rate constant. If the plot of t/q_t versus t shows a linear relationship, k₂ and q_e can be determined from the slope and intercept of the line (Fig. S6B and D†).

The kinetics parameters, such as kinetic rate constants, correlation coefficients, and q_e, are shown in Table 6. The pseudo-second-order model is more suitable to describe the kinetic profiles of SII-MM1 and SNI-MM, because the calculated q_e from this model agreed with the experimental value. In addition, it was noticed that k₂ of SII-MM1 was much larger than that of SNI-MM, suggesting that surface ion-imprinted composites had a faster kinetics and rapid adsorption, which might be attributed to the strong affinity of the imprinted surfaces. To clarify this point, the nitrogen sorption experiments were performed, and the surface area was determined by applying BET equation. The results show that the surface area of the samples could be increased by surface ion-imprinted technique (Table S1, ESI†), which may result in faster kinetics and larger capacity.

Table 6 Kinetic parameters for the adsorption of U(VI) by SII-MM1 and SNI-MM

Sample	Pseudo-first order				Pseudo-second order
	qe,exp (mg g^−1)	k1 (h^−1)	qe,cal (mg g^−1)	R^2	k2 (g mg^−1 h^−1)	qe,cal (mg g^−1)	R^2
SII-MM1	14.48	219.60	13.67	0.9815	54.01	14.89	0.9957
SNI-MM	7.50	9.02	8.58	0.9790	1.77	8.26	0.9844

---

3.6 Adsorption isotherm

To understand the capacity of adsorbent, experiments of adsorption isotherm were conducted at 298.15 K and pH 5.0 with uranyl ions concentration over the range of 5 × 10⁻⁴–5 × 10⁻⁶ mol L⁻¹. The relationship between q_e and C_e is shown in Fig. S7 (ESI†). The q_e increased with the increasing of C_e. The equilibrium data were applied to Langmuir and Freundlich isotherm models. Langmuir model is given as eqn (10), which describes monolayer adsorption based on the assumption that all the adsorption sites have equal affinity, and that desorption at one site doesn't affect an adjacent site.²⁷where q_max (mg g⁻¹) is the maximum adsorption amount (i.e. adsorption capacity), and b (L mg⁻¹) is the Langmuir constant. They can be calculated from the linear plot of C_e/q_e against C_e (Fig. 5A).

Fig. 5 (A) Langmuir isotherm (B) Freundlich isotherm plots for the adsorption of uranium(VI) on SII-MM (a) and SNI-MM (b).

The Freundlich model can be applied for multilayer adsorption and the adsorption on heterogeneous surfaces, which can be expressed as eqn (11):

where K_F (L g⁻¹) and n are Freundlich constants related to adsorption capacity and adsorption intensity, respectively, which can be calculated from the linear plot of log q_e versus log C_e (Fig. 5B).

The isotherm parameters and the correlation coefficients of Langmuir and Freundlich models are summarized in Table 7. The relatively larger correlation coefficients indicated that the adsorption processes could be better described by Langmuir model. Importantly, q_max of imprinted composites could reach 146.41 mg of U per g, which is far larger than that of non-imprinted materials (19.7 mg g⁻¹). In comparison with the reference results (Table 8), there is a markedly higher adsorption rate and relatively larger capacity for the imprinted magnetic composite.

Table 7 Langmuir and Freundlich parameters for uranyl ions adsorption U(VI) by SII-MM and SNI-MM

Adsorbent	Langmuir			Freundlich
	qmax (mg g^−1)	b (L mg^−1)	R^2	KF (L g^−1)	N	R^2
SII-MM	146.41	0.87	0.997	65.39	1.56	0.977
SNI-MM	19.7	0.47	0.983	6.41	1.133	0.967

---

Table 8 Comparison of U(VI) adsorption capacities of the imprinted magnetic composites with other magnetic adsorbents

Matrix	pH	Capacity (mg g^−1)	Equilibrium time (min)	Reference
Magnetic Schiff base	6.0	94.30	360	42
Amidoxime modified Fe3O4@SiO2	5.0	105.0	120	20
Fe3O4@SiO2 magnetic composites	6.0	52.36	180	14
Quercetin modified magnetic nanoparticles	3.7	12.33	30	43
Functionalized Fe3O4@SiO2	4.0	7.15	30	27
Fe3O4@APTMS	6.0	151.80	60	17
Magnetic IIP	4.0	1.1	45	44
Ethylenediamine-modified magnetic chitosan	3.0	82.83	30	45
Nano-iron oxide modified resin	4.0	47.2	8.0	46
β-Cyclodextrin conjugated magnetic HNT/iron oxide	5.5	107.58	300	47
Imprinted magnetic composites	5.0	146.41	1.0	This work

---

3.7 Desorption and regeneration studies

It is important to conduct desorption studies for regenerating the spent adsorbent. In this study, 0.1 mol L⁻¹ of EDTA solution was used as the desorbing agent, and the reusability of the composites was investigated in five adsorption/desorption cycles. Results are depicted in Fig. 6. After five cycles, the adsorption efficiency remained almost constant (Fig. 6A). The SII-MM–uranyl ion complex was characterized by EDX spectra and SEM. The appearance of U element in EDX spectrum verified the adsorption of uranium (Fig. 6B). The elemental mapping results indicate that Si and O are distributed on the core of microsphere, while U is distributed on the surface, which further demonstrated the adsorption of uranium (Fig. S8, ESI†). And no obvious damage was observed for the microspheres (Fig. 6C), which might be attributed to the covalent combination of the ion-imprinted polymer onto the surface. Results suggested that SII-MM could be efficiently regenerated by 0.1mol L⁻¹ EDTA solution and reused with high adsorption efficiency after five cycles.

Fig. 6 (A) Recycling of SII-MM in the adsorption of uranium(VI), (B) EDX spectrum and (C) SEM image of SII-MM–uranyl ions complex after recycling. Scale bar: 300 nm. (Experimental conditions: 5 mL solution, 10⁻⁴ mol L⁻¹ uranyl ions; 0.5 g L⁻¹ adsorbent; 298.15 K; pH 5.0; contact time = 0.5 h.)

In order to investigate preconcentration factor, 0.5 mg UO₂(NO₃)₂·6H₂O were dissolved in different volumes (50–400 mL) of deionized water for the adsorption, and 2.5 mL of 0.1 mol L⁻¹ EDTA solution was used for elution. It could be found that the adsorbed uranyl ions were eluted with 2.5 mL EDTA solution with 95.2% recovery when uranyl aqueous solution (300 mL, 3.32 × 10⁻⁶ mol L⁻¹) was used for adsorption, and the recovery reduced (≤95%) in volume above 300 mL of uranyl aqueous solution. Therefore, the preconcentration limit is 3.32 × 10⁻⁶ mol L⁻¹ of uranyl aqueous solution, and the preconcentration factor is approximately 120.

3.8 Application to real samples

Due to the high capacity and fast kinetics, it is worth to explore adsorbent's application to real samples. In this study, deionized water, lake water and tap water were chosen to be solvents for preparing uranyl ion solutions (23.8 ppm). 2.5 mg adsorbents were added in 5 mL sample solutions, and the adsorption results are listed in Table 9. The high adsorption efficiencies suggest that SII-MM is a potential promising adsorbent to efficiently remove uranium(VI) from aqueous solution.

Table 9 Comparison of U(VI) adsorption efficiency in different real samples

Samples	Cadded (ppm)	Adsorption efficiency (%)	RSD (%)
Deionized water	23.8	94.3	1.08
Lake water (Dushu Lake, Suzhou)	23.8	91.2	2.16
Tap water	23.8	92.1	1.20

---

4. Conclusion

In this work, we demonstrate a new method to prepare surface ion-imprinted magnetic microspheres by locating copolymerization for rapid and selective extraction of uranium(VI). The initiation and propagation processes were controlled on the surface of the microspheres due to the hydrophobic initiator, and ion-imprinted polymer was covalently bound onto the surface of magnetic microspheres. The effects of chemical composition, pH, adsorbent dose and competing ions on uranyl ion adsorption were investigated. The more surface nitrogen content of SII-MM, the shorter equilibrium time and higher adsorption efficiency. The investigation of adsorption thermodynamics and kinetics showed an exothermic spontaneous adsorption process that kinetically followed pseudo-second-order model. Compared with non-imprinted composites, SII-MM showed higher selectivity, faster kinetics, and larger capacity for uranyl ions adsorption. The adsorption could reach equilibrium with a capacity (q_max) of 146.41 mg g⁻¹ within 60 seconds at pH 5.0 and 298.15 K, and SII-MM could be efficiently regenerated and reused with high adsorption efficiency after five cycles.

In comparison with previous literature results, the imprinted magnetic composite has a higher adsorption capacity with faster kinetics. This work indicates that SII-MM could be a promising adsorbent for rapid and selective removal of uranium(VI) from aqueous solution. In this study, the similar concept of locating polymerization should be applicable to the other initiators and monomers for preparation of surface ion-imprinted magnetic composites. This work may provide a new approach to prepare novel adsorbents for highly efficient removal of uranium(VI) from aqueous solution.

Acknowledgements

This work is supported by Natural Science Foundation of China (91326202, 21174100), a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), Qing-Lan Project of Jiangsu Province, and Jiangsu Key Laboratory of Radiation Medicine and Protection. We thank Prof. Zhifang Chai and Prof. Shuao Wang for helpful constructive suggestion. Jun Qian and Shuang Zhang contributed equally to this work.

Notes and references

1. A. Gilman, D. Villeuve, V. A. Yagminas, B. Tracy, J. Quinn, V. Valli, R. Willes and M. Moss, Toxicol. Sci., 1998, 41, 117.
2. S. C. Sheppard and W. G. Evenden, Arch. Environ. Contam. Toxicol., 1992, 23, 117.
3. A. C. Sather, O. B. Berryman and J. Rebek Jr, J. Am. Chem. Soc., 2010, 132, 13572.
4. A. Mellah, S. Chegrouche and M. Barkat, Hydrometallurgy, 2007, 85, 163.
5. K. Vaaramaa, S. Pulli and J. Lehto, Radiochim. Acta, 2000, 88, 845.
6. T. P. Rao, P. Metilda and J. M. Gladis, Talanta, 2006, 68, 1047.
7. M. O. Barnett, P. M. Jardine and S. C. Brooks, Environ. Sci. Technol., 2002, 36, 937.
8. S. Aytas, M. Yurtlu and R. Donat, J. Hazard. Mater., 2009, 172, 667.
9. H. Yan, J. Bai, X. Chen, J. Wang, H. Zhang, Q. Liu, M. Zhang and L. Liu, RSC Adv., 2013, 3, 23278.
10. X. Yang, H. Tang, K. Cao, H. Song, W. Sheng and Q. Wu, J. Mater. Chem., 2011, 21, 6122.
11. A. Sayari, S. Hamoudi and Y. Yang, Chem. Mater., 2005, 17, 212.
12. J. Aguado, J. M. Arsuaga, A. Arencibia, M. Lindo and V. Gascón, J. Hazard. Mater., 2009, 163, 213.
13. W. Yantasee, Y. Lin, G. E. Fryxell, B. J. Busche and J. C. Birnbaum, Sep. Sci. Technol., 2003, 38, 3809.
14. F. L. Fan, Z. Qin, J. Bai, W. D. Rong, F. Y. Fan, W. Tian, X. L. Wu, Y. Wang and L. Zhao, J. Environ. Radioact., 2012, 106, 40.
15. I. Ali, Chem. Rev., 2012, 112, 5073.
16. J. H. Jung, J. H. Lee and S. Shinkai, Chem. Soc. Rev., 2011, 40, 4464.
17. X. Zhang, J. Wang, R. Li, Q. Dai and L. Liu, New J. Chem., 2013, 37, 3914.
18. S. Sayin and M. Yilmaz, Desalination, 2011, 276, 328.
19. L. Wang, Z. Yang, J. Gao, K. Xu, H. Gu, B. Zhang, X. Zhang and B. Xu, J. Am. Chem. Soc., 2006, 128, 13358.
20. Y. Zhao, J. Li, L. Zhao, S. Zhang, Y. Huang, X. Wu and X. Wang, Chem. Eng. J., 2014, 235, 275.
21. D. Qi, H. Zhang, J. Tang, C. Deng and X. Zhang, J. Phys. Chem. C, 2010, 114, 9221.
22. C. Branger, W. Meouche and A. Margaillan, React. Funct. Polym., 2013, 73, 859.
23. M. Andac, E. Ozyapi, S. Senel, R. Say and A. Denizli, Ind. Eng. Chem. Res., 2006, 45, 1780.
24. Y. Zhan, X. Luo, S. Nie, Y. Huang, X. Tu and S. Luo, Ind. Eng. Chem. Res., 2011, 50, 6355.
25. M. Yoshida, K. Uezu, M. Goto and S. Furusaki, J. Appl. Polym. Sci., 1999, 73, 1223.
26. B. Guo, F. Deng, Y. Zhao, X. Luo, S. Luo and C. Au, Appl. Surf. Sci., 2014, 292, 438.
27. S. Sadeghi and E. Aboobakri, Microchim. Acta, 2012, 178, 89.
28. M. Liu, C. Chen, T. Wen and X. Wang, Dalton Trans., 2014, 43, 7050.
29. M. Shamsipur, J. Fasihi and K. Ashtari, Anal. Chem., 2007, 79, 7116.
30. Y. Kim, S. An, S. Kim, J. H. Park, H. A. Son, H. T. Kim, K.-D. Suh and J. W. Kim, Polymer, 2013, 54, 5609.
31. A. C. Sather, O. B. Berryman and J. Rebek, J. Am. Chem. Soc., 2010, 132, 13572.
32. M. M. Aly and M. F. Hamza, J. Dispersion Sci. Technol., 2013, 34, 182.
33. M. B. Andrews and C. L. Cahill, Chem. Rev., 2013, 113, 1121.
34. H. Deng, X. L. Li, Q. Peng, X. Wang, J. P. Chen and Y. D. Li, Angew. Chem., Int. Ed., 2005, 44, 2782.
35. D. B. Hua, J. Tang, L. L. Dai, Y. W. Pu, X. B. Cao and X. L. Zhu, J. Nanosci. Nanotechnol., 2009, 9, 6681.
36. W. Liu, F. Zhou, X. Y. Zhang, Y. Li, X. Y. Wang, X. M. Xu and Y. W. Zhang, J. Nanosci. Nanotechnol., 2014, 14, 3068.
37. A. Barkleit, H. Moll and G. Bernhard, Dalton Trans., 2009, 5379.
38. A. Sarı, M. Tuzen and M. Soylak, J. Hazard. Mater., 2007, 144, 41.
39. S. Lagergren, K. Sven. Vetenskapsakad. Handl., Band 24, 1898, 4, 1.
40. Y. S. Ho and G. McKay, Chem. Eng. J., 1998, 70, 115.
41. Y. S. Ho, J. Hazard. Mater., 2006, 136, 681.
42. X. Zhang, C. Jiao, J. Wang, Q. Liu, R. Li, P. Yang and M. Zhang, Chem. Eng. J., 2012, 198, 412.
43. S. Sadeghi, H. Azhdari, H. Arabi and A. Z. Moghaddam, J. Hazard. Mater., 2012, 215, 208.
44. N. T. Tavengwa, E. Cukrowska and L. Chimuka, J. Hazard. Mater., 2014, 267, 221.
45. J. S. Wang, R. T. Peng, J. H. Yang, Y. C. Liu and X. J. Hu, Carbohydr. Polym., 2011, 84, 1169.
46. O. A. Elhefnawy, W. I. Zidan, M. M. Abo-Aly, E. M. Bakier and G. A. Elsayed, J. Radioanal. Nucl. Chem., 2014, 299, 1821.
47. S. Yang, P. Zong, J. Hu, G. Sheng, Q. Wang and X. Wang, Chem. Eng. J., 2013, 214, 376.

---

Footnotes

† Electronic supplementary information (ESI) available: The synthesis of Fe₃O₄@SiO₂–C=C microspheres; SEM images of Fe₃O₄ and Fe₃O₄@SiO₂; FTIR spectra of Fe₃O₄, Fe₃O₄@SiO₂ and Fe₃O₄@SiO₂–C=C; magnetization curves of Fe₃O₄@SiO₂–C=C and SII-MM1; EDX spectra of SII-MM1, SII-MM2 and SII-MM3; XPS spectrum of C 1s of SII-MM1; the effect of chemical composition of SII-MM on the adsorption of uranyl ions; pseudo-first-order kinetics and pseudo-second-order kinetics of uranium(VI) adsorption by SII-MM and SNI-MM; surface area of different samples; adsorption isotherms of uranium(VI) on SII-MM1; TEM image of the core–shell structure of SII-MM1 and elemental mappings. See DOI: 10.1039/c4ra10037a

‡ J. Qian and S. Zhang contributed equally.

---