Hollow imprinted polymer nanorods with a tunable shell using halloysite nanotubes as a sacrificial template for selective recognition and separation of chloramphenicol

Abstract

The wide use of antibiotics in human therapy and veterinary practice has resulted in the presence of residual antibiotic compounds in water environments, which are harmful to ecology and health. In this work, novel hollow molecularly imprinted nanorods (HMINs) with uniform and controllable thickness of the polymer shell were successfully prepared via a combination of in situ surface precipitation polymerization and halloysite nanotubes sacrificial template method, and were used as an advanced selective nanoadsorbent to remove chloramphenicol (CAP). The physicochemical properties of HMINs were well characterized by FE-SEM, TEM, FT-IR and TG/DTA. HMINs with a shell thickness of 62 nm (HMINs-2) displayed excellent adsorption capacity and fast kinetics. The experimental adsorption equilibrium and kinetic data were best described by the Freundlich isotherm model and the pseudo-second-order rate equation, respectively. Furthermore, HMINs-2 possessed highly specific recognition to CAP in aqueous solutions, as compared with other reference antibiotics. Meanwhile, HMINs-2 also had excellent dispersibility, regeneration properties and thermal stability for the promising potential application in wastewater treatment.

---

1 Introduction

Almost all antibiotics are poorly metabolized in organisms and as a result, a large proportion of them are excreted through urine and feces into the environment despite the maintenance of biological activity.¹ Antibiotic residues in water environment have attracted great attention because they have been proved to be a class of potent organic pollutants.² Chloramphenicol (CAP) has been used largely in aquaculture and animal husbandry, which has led to a seriously large amount of CAP residues in the aquatic environment.³ There is an increasing potential risk to human health which derives from the excess amounts of CAP residues in the food chain and drinking water. Thus, the development of efficient and reliable methods for the removal of CAP from aquatic environments is very necessary. Generally, the treatment techniques of antibiotics residues include degradation,⁴ bio-removal,⁵ adsorption,^6,7 and membrane filtration⁸ etc. Among of them, adsorption has become the most commonly method to remove antibiotics residue in environment due to the advantages of economic feasibility, easy handling and high efficiency. However, bad selective adsorption ability of the common adsorbents cannot meet the requirement of actual application.

Molecular imprinting technique (MIT) could endow molecularly imprinted polymers (MIPs) with the ability of specific recognition to the target species.⁹ The MIPs with tailor-made specific recognition sites display the ability of selective rebinding to the target species in preference to similar substances, which simulates the relationship between lock and key. To obtain the recognition sites, the template molecules are integrated with functional monomers via various interactions in a suitable solvent, and then co-polymerization of functional monomers and cross-linker monomer is carried out to produce a three-dimensional polymeric matrix. Finally, the specific binding cavities in MIPs matched with the template molecules in size, geometry structure and functional groups were achieved after the removal of template molecules from the polymeric matrix by extraction procedure or chemical reaction.¹⁰ MIPs have a lot of advantages, such as low cost, stability and simple preparation, which can promote MIPs widely be used to eliminate CAP residues in environment.¹¹ However, the MIPs prepared by the traditional MIT exhibited some defects including active recognition sites embed deeply, poor dynamic, weak mechanical and bad regeneration performance,^12,13 which greatly prevented their applications in practice. The surface molecular imprinting technique (SMIT) and hollow molecular imprinting technique have shown great superiority as both of these methods are capable of blotting polymer on the surface of the solid support with controllable thickness, which is beneficial to interaction of target molecule and binding site.

In addition, SMIT is able to coat MIPs onto various matrix materials,^14–16 which can overcome the defects of traditional MIT. As we all know, the silicon-based materials were commonly modified with MIPs.^11,15 Generally, the amino, carboxyl, vinyl and triggered groups could be modified onto the surface of the silicon-based materials via different methods. Then, MIPs thin layer was coated on the surface of silicon-based materials to form core–shell structure. The binding sites are exposed on the surface of the scaffold owing to the thin thickness of imprinted layer prepared by SMIT, which is conducive to the elution and reidentification of template molecules. Importantly, substrate materials (e.g. carbon nanomaterials,¹⁷ nanoparticles,¹⁸ polymer membrane¹⁹) with the different characteristics can be chosen to meet various needs. Hollow materials have special properties, such as lager interior space and excellent surface properties, and thus are widely used in many fields.^20,21 Synthesis of hollow polymer via a sacrificial template route allow the polymer shell to coat at the surface, and then templates are removed by physical or chemical method to get hollow structured polymer. Halloysite nanotubes (HNTs, Al₂Si₂O₅(OH)₄·2H₂O) is a kind of natural minerals in possession of superior property such as high chemical stability, large surface area and tunable surface property, which have been widely applied in various research fields, namely, drug delivery,²² separation,²³ nano-templates²⁴ and catalysis.²⁵ HNTs are rich in natural reserves with the advantage of cost saving compared with other expensive nanomaterials (e.g. carbon nanotubes). Notably, they also possess unique tubular structure. Taking the advantages of cost saving and unique structure into consideration, HNTs can be widely used in the MIT as a sacrificial hard template.²⁶

Precipitation polymerization was extensively employed in preparation of MIPs. Commonly, monomers, template molecules, initiator and cross-linker were dissolved in an appropriate solvent for obtaining precipitation of uniform polymer microspheres. Particularly, the size of polymer microspheres could be adjusted by changing the solvent species or solvent dosage.²⁷ Importantly, the reaction process do not add any extra dispersant or surfactant, resulting in none impurities residue and simple post-processing, which guarantees the wide application in MIPs preparation. Su et al. prepared the MIPs by precipitation polymerization, showing the good selectivity for vanillin in the specific adsorption experiments.²⁸ However, the imprinted polymer microspheres generally showed slow kinetics ability. Therefore, the development of facile and general methods for the fabrication of nanoadsorbents with fast kinetics and high adsorption capacity is an urgent subject.

In the present work, we first prepared hollow molecularly imprinted polymer nanorods by the combination of sacrificial template and precipitation polymerization methods. Firstly, the HNTs were modified with vinyl through a simple silanization reaction to form polymerizable vinyl layers. Then, surface molecular imprinting nanorods (SMINs) were obtained by the precipitation polymerization on surface of modified HNTs with CAP as template molecule. Finally, for the purpose of obtaining hollow molecular imprinting nanorods (HMINs), the CAP template and HNTs were removed by soxhlet extraction and HF etching, respectively. The schematic specific diagram for technological route HMINs was shown in Fig. 1. The morphology and structure of HMINs were analyzed by different characterizations. The effect of MIPs layer thickness on adsorption was investigated. The impact of non-modified HNTs (n-HNTs) for MIPs shell structure was studied as well. Adsorption isotherms, kinetics, competitive adsorption and regeneration experiments in details were studied to evaluate performance of HMINs.

Fig. 1 Schematic representation of synthesis route for HMINs.

2 Materials and methods

2.1 Materials

HNTs were purchased from Henan, China and used with purification procedure by grinding, sieving, sedimentation and pickling. 3-(Trimethoxysilyl) propyl methacrylate (KH570), ethylene glycol dimethacrylate (EGDMA, 98%), methacrylic acid (MAA, 98%), chloramphenicol (CAP, 98%), chlortetracycline (CTC, 98%), cefalexin (CFX, 99%), ciprofloxacin (CIP, 98%) and tetracycline (TC) obtained from Aladdin reagent co., Ltd. (Shanghai, China). Azobis(2-methyl-propionitrile) (AIBN, 99%), anhydrous toluene (99.8%), methanol, acetone, ethanol, acetonitrile, acetic acid, nitric acid (65%), hydrofluoric acid (HF, 40%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All reagents were of analytical grade and used directly without any further purification.

2.2 Instruments

The morphology was characterized by field emission scanning electron microscope (FE-SEM, S-4800, Japan) and transmission electron microscopy (TEM; IEM-200CX, JEOL, Japan), Fourier transform infrared spectra were implemented on a Tensor FT-IR spectrometer (FT-IR; Nicolet Nexus 470, USA) with KBr pellets. Thermal stability was studied using a Diamond TG/DTA instruments (STA 449C Jupiter, Netzsch, Germany) from 25 to 800 °C with a ramp of 10 °C min⁻¹ under N₂. The concentration of the CAP remaining in the solution was measured by UV-vis spectrophotometer (Agilent Cary 8454).

2.3 Surface functionalization of HNTs

In detail, nitric acid (5.0 mL) and distilled water (30 mL) were mixed, and then the HNTs (1.0 g) were added to the acid solution with ultrasonic treatment for 30 min. The mixture was heated to 80 °C with reflux condensation under vigorous stirring for 12 h. The pickling of HNTs were obtained by the suction filter, washed with distilled water to neutral and dried under vacuum at 60 °C for 12 h. Then, treated HNTs (1.0 g), toluene (100 mL) and KH570 (3.0 mL) were mixed under ultrasound for 30 min. The mixture was heated to 90 °C with reflux condensation under vigorous stirring for 24 h. After the reaction, the resultants were washed repeatedly with ethanol and dried under vacuum at 60 °C for 12 h to obtain the vinyl functionalized HNTs (HNTs@KH570).

2.4 The preparation of HMINs

Briefly, CAP (template molecule), MAA (functional monomer) and EGDMA (cross-linker monomer) were dissolved in acetonitrile (40 mL) to self-assemble in the dark for 12 h at room temperature. 80 mg of HNTs@KH570 and 10 mg of AIBN (polymerization initiator) were added into the above solution by ultrasonication and poured with nitrogen for 30 min to remove oxygen. Then the reaction was performed in water bath oscillator at 60 °C for 24 h with the rate of 180 rpm. The product was collected by centrifugation and washed with ethanol several times. The CAP molecule was washed by soxhlet extraction using a mixture of methanol/acetic acid mixture (9 : 1, v/v) as eluent for 24 h. The surface molecular imprinted nanorod (SMINs) based onto HNTs were obtained by repeatedly washed with distilled water to neutral and dried to constant weight. SMINs were added to HF (15 wt%) aqueous solution to remove HNTs at room temperature for 24 h. The finally products were washed with distilled water to neutral and dried under vacuum to hollow molecularly imprinted nanorods (HMINs). In the preparation process, the mole ratio of CAP : MAA : EGDMA was kept at 1 : 4 : 20. The different thickness of polymer shell for SMINs was controlled by adjusting the addition amount of CAP, MAA and EGDMA. When the addition amount of CAP were 0.2, 0.3 and 0.4 mmol, the corresponding surface molecular imprinted nanorods was named SMINs-1, SMINs-2, SMINs-3, respectively. After HF etching, the products derived from SMINs-1, SMINs-2, SMINs-3 were named HMINs-1, HMINs-2, HMINs-3, respectively.

SNIN-2 and HNINs-2 was also obtained under the same condition as SMINs-2 and SMINs-2, respectively, without the addition of template CAP molecule. Meanwhile, the effect of the added HNTs without KH570 modification (n-HNTs) on shell structure of SMINs was investigated. The n-HNTs surface molecular imprinted nanorods (n-SMIN-2) were obtained with the same parameters of SMINs-2 but using n-HNTs as template.

2.5 Batch adsorption experiments

All adsorption experiments were conducted on centrifuge tubes containing 5.0 mg of nanoadsorbents and 10 mL of CAP solution under dark condition. The supernatant were obtained by high-speed centrifugation for detecting CAP concentration via UV-vis spectrophotometer at the maximum absorption wavelength of 278 nm. Notably, all the batch adsorption experiments were carried out in triplicate.

The adsorption isotherms of HMINs-1, HMINs-2, HMINs-3 and HNINs-2 were implemented within the initial CAP concentration range of 10–350 μmol L⁻¹ for 12 h to reach equilibrium at 25 °C, respectively. The equilibrium adsorption capacity Q_e (μmol g⁻¹) was calculated by the following equation:

The kinetic property of HMINs-2 and HNINs-2 were studied at appropriate time intervals ranging from 5 to 120 min using 100 μmol L⁻¹ CAP solutions at 25 °C, respectively. The Q_t (μmol g⁻¹) of adsorption amount at t time (min) was obtained according to the following equation:

Herein, C₀ and C_t (μmol L⁻¹) are initial CAP concentration and residual CAP concentration at t time, respectively. C_e (μmol L⁻¹) is CAP concentration at equilibrium. V (L) and M (g) are solution volume and nanoadsorbents mass, respectively.

CTC, CIP, CFX and TC were chose as reference antibiotics to research selective ability of HMINs-2 toward CAP. Selective experiments were carried out by adding 5.0 mg of HMINs-2 and HNINs-2 into 10 mL of CTC, CIP, CFX, TC and CAP solutions with an initial concentration of 100 μmol L⁻¹ at 25 °C, respectively. Supernatant was obtained through the high-speed centrifuge after adsorption and tested, and then the adsorption amounts were calculated. The distribution coefficient (K_d), selective coefficient (k) and relative selective coefficient (k′) were obtained by the following equation:

Where K_d(CAP) and K_d(x) (L g⁻¹) are the distribution coefficient of CAP and reference antibiotics, respectively. The k_M and k_N are the selective coefficient of HMINs and HNINs, respectively.

Additionally, the cyclic adsorption–desorption experiments were conducted to investigate the regeneration and stability of HMINs-2. The saturated HMINs-2 was recovered by high-speed centrifugation and desorbed via a methanol/acetic acid mixture (9 : 1, v/v) to reuse. The desorbed HMINs-2 (5.0 mg) was added into 10 mL of fresh CAP solution with the initial concentration of 100 μmol L⁻¹, and the adsorption reached equilibrium at 25 °C. The adsorption amount was then calculated. The cyclic adsorption–desorption experiments were continually implemented 6 times under the same condition.

3 Results and discussion

3.1 Characterization

The chemical groups of HNTs, HNTs@KH570, SMINs-2 and HMINs-2 were characterized by FT-IR spectroscopy. Fig. 2a shows FT-IR spectrum of HNTs and HNTs@KH570. The peaks at 2981 cm⁻¹ and 2940 cm⁻¹ were attributed to the –CH₃ and –CH₂ stretching vibration, respectively. The peak at 1719 cm⁻¹ was ascribed to C=O stretching vibrations, which demonstrated the KH570 molecule was successfully modified on the surface of HNTs.²⁹ The FT-IR spectra of SMINs-2 and HNINs-2 are shown in Fig. 2b. The broad peak at 3450 cm⁻¹ was attributed to –OH stretching of MAA molecule. The significant peaks located 1733 (C=O stretching), 1258 and 1156 cm⁻¹ (C–O–C stretching) indicated that the EGDMA (cross-linker agent) participated in the imprinted polymerization.³⁰ After HF etching of HNTs, the peaks at 3697 cm⁻¹, 3626 cm⁻¹, 1087 cm⁻¹, 1033 cm⁻¹, 1912 cm⁻¹ and 535 cm⁻¹ in the spectrum of HMINs-2 were disappeared, indicating the successful removal of HNTs template.

Fig. 2 FT-IR spectra of (a) HNTs and HNTs@KH570 (b) SMINs and HMINs.

SEM images of HNTs, HNTs@KH570, SMIN-2 and HMINs-2 are given in Fig. 3. As shown in Fig. 3a, HNTs presented long-fibrous morphology with a length of 1.0–3.0 μm and outer diameter of 50–180 nm. The morphology of HNTs@KH570 remained almost no change, owing to the KH570 modification was at the molecule level. In Fig. 3c, after surface coating of imprinted polymer, SMINs-2 showed the different characters as compared with HNTs and HNTs@KH570. SMINs-2 was still a long-range rod-shape with the diameter increasing to about 170–300 nm. Moreover, the surface of SMINs-2 was smooth and dispersibility was greatly improved, indicating imprinted polymer nanoshell was successfully coated onto the surface of HNTs. As shown in Fig. 3d, HMINs-2 still kept a long-range cylindrical nanorod and good dispersibility, suggesting that HF etching process did not affect the morphology of HMINs-2. Interestingly, the n-SMIN-2 displayed different structure from SMINs-2. As shown in Fig. S1,† the n-SMIN-2 showed a candied gourd-like structure with rough and nonuniform surface, illustrating the vinyl functionalization of HNTs were crucial for uniform structure of polymer nanolayer.

Fig. 3 SEM images of HNTs (a), HNTs@KH570 (b), SMINs-2 (c) and HMINs-2 (d).

The morphological features of samples were further studied by TEM. Fig. 4 presents TEM images of HNTs, SMINs-1, SMINs-2 and SMINs-3. As shown Fig. 4a and b, we could see HNTs had a hollow tubular structure with some structure damage. The three SMINs samples exhibited the obvious core–shell nanorod-like structures with HNTs as core and imprinted polymer as shell. Imprinted polymer nanoshell was uniform and grown along the longitudinal direction. With the increase in the addition amount of polymerization precursor, the thickness of imprinted polymer increased gradually. The average thickness of polymer nanoshell for SMINs-1, SMINs-2 and SMINs-3 was about 42, 58 and 86 nm, respectively. However, Fig. S2† exhibited that n-SMIN-2 had an uneven polymer shell, which was matched well with SEM images. It also can find that there was some agglomeration in TEM image of n-SMIN-2. Vinyl functionalized HNTs@KH570 contained a lot of vinyl groups, which could co-polymerize with functional monomer (MAA) and cross-linker (EGDMA) in the initiation of AIBN, which was favor for the formation of uniform polymer shell.¹⁰ The surface of HNTs mainly exist the Al–OH and Si–OH groups, which only could form hydrogen-interaction with CAP and MAA in the solution. Thus, the vinyl modification had a significant influence on polymer morphology.

Fig. 4 TEM images of HNTs (a and b), SMINs-1 (c and d), SMINs-2 (e and f) and SMINs-3 (g and h).

Fig. 5 shows TEM images of HMINs-1, HMINs-2 and HMINs-3. All the three samples displayed a hollow nanorod-like structure with uniform polymer shell. With the increase of the monomer, the thickness of imprinted polymer shell increases gradually. Notably, the rod-like structure of polymer was still intact after HF etching. Meanwhile, the thickness of polymer shell increased slightly. The average thickness of imprinted polymer shell of HMINs-1, HMINs-2 and HMINs-3 was about 44, 62 and 90 nm, respectively. This phenomenon might be caused by the difference of electron microscope selection and the inward contraction of polymer shell after the removal of bearing HNTs via HF etching.

Fig. 5 TEM images of HMINs-1 (a), HMINs-2 (b) and HMINs-3 (c).

TGA was used to quantify the content of imprinted polymers coated on the surface of HNTs. TGA curves of HNTs, HNTs@KH570, SMINs-1, SMINs-2 and SMINs-3 are shown in Fig. 6. Low percentage of weight loss was mainly caused by desorption of physically absorbed water loss for all the samples at the temperature below 250 °C. The mass loss of HNTs@KH570 was higher than blank HNTs, showed vinyl functional group was successfully introduced on the surface of HNTs. The rapid rate of mass loss for SMINs-1, SMINs-2 and SMINs-3 when temperature was higher than 250 °C, owing to the thermal decomposition of organic polymers. The loss ratios of SMINs-1, SMINs-2 and SMINs-3 increased with the increase of polymer shell from 59.0, 63.3–71.2%, respectively. The weight essentially unchanged owing to the remaining substances (mainly included stable inorganic and small amounts of carbon residue) at the temperature above 600 °C. The results indicated the as-prepared SMINs possessed pretty thermal stability for practical applications.

Fig. 6 TGA curves of the HNTs (a), HNTs@KH570 (b), SMINs-1 (c), SMINs-2 (d) and SMINs-3 (e).

3.2 Adsorption isotherms

The adsorption capacity of HMINs-1, HMINs-2, HMINs-3 and HNINs-2 for CAP were investigated by adsorption isotherm experiment. As shown in Fig. 7, adsorption capacity of HMINs-1, HMINs-2, HMINs-3 and HNIN-2 for CAP was linearly ascended with the increase of initial concentration. The experimental values (Q_e,exp) of HMINs-1, HMINs-2, HMINs-3 and HNIN-2 are 192.8, 167.8, 139.3 and 114.9 μmol g⁻¹, respectively. Obviously, the adsorption capacity of imprinted nanoadsorbents (HMINs-1, HMINs-2 and HMINs-3) for CAP was higher than the non-imprinted nanoadsorbents (HMINs-2). The results demonstrated abundant efficient recognition sites located at the surface of HMINs. To the best of our knowledge, the imprinted bulk materials is greatly limited by the small total amount of binding sites due to the low surface area of substrate. Accordingly, the imprinted bulk materials display always a low rebinding capacity. However, imprinted nanolayer has a small dimension with higher surface-to-volume ratio, and thus most of template molecules are located at the surface, providing a complete removal of templates and an excellent accessibility to target species. From Fig. 7b, the HMINs-2 exhibited highest adsorption capacity for CAP, which indicated that the optimized thickness of imprinting layer was about 62 nm. As for HMINs-1 with an imprinting layer thickness of 44 nm, it had thin imprinting shell with little binding sites. But for HMINs-3, the thickness increased to 90 nm, with the incomplete removal of CAP template molecules and some inefficient binding sites, which were located in the deeply inside the hollow rod with the closed port structure. Thus, the adsorption amount per unit mass of was HMINs-3 lower than HMINs-2, which had more effective recognition sites. The thickness of imprinted polymer shell was important to adsorption performance.

Fig. 7 Freundlich isotherm linear fitting for HMINs-1, HMINs-2, HMINs-3 and HNINs-2 toward CAP (a); Langmuir and Freundlich non-linear fitting of HMINs-1, HMINs-2, HMINs-3 and HNINs-2 toward CAP (b).

The static adsorption data were analyzed by Langmuir and Freundlich adsorption isotherm models.^31,32 Fig. 7 showed the fitting curve of the adsorption isotherms and corresponding parameters are listed in Table 1. The Langmuir isotherm had a low correlation coefficients (R² < 0.98), implied the Langmuir model could not well describe the adsorption isotherm data. However, the Freundlich model displayed a well linear correlation (R² > 0.99), and the theoretical value (Q_m) by calculation was close to the experimental value (Q_e,exp), suggested the heterogeneous surface and multi-molecular layer adsorption.

Table 1 Langmuir and Freundlich isotherm model parameters of HMINs and HNINs-2 for CAP

Adsorbents	Qm,exp (μmol g^−1)	Langmuir isotherm model			Freundlich isotherm model
		Qm,c (μmol g^−1)	KL (L μmol^−1)	R^2	KF (μmol g^−1) (L μmol^−1/n)	1/n	R^2
HMINs-1	192.8	352.1	0.00249	0.9419	1.160	0.8758	0.9914
HMINs-2	167.8	571.4	0.00211	0.9686	1.575	0.8887	0.9971
HMINs-3	139.3	621.1	0.00157	0.6999	1.096	0.9336	0.9912
HNINs-2	114.9	303.9	0.00204	0.9747	0.854	0.8755	0.9965

---

3.3 Adsorption kinetics

The pseudo-first-order and pseudo-second-order kinetic models were used to investigate the adsorption process.³³ The adsorption kinetics curves of HMINs-2 and HNINs-2 was shown in Fig. 8. As shown Fig. 8b, the adsorption capacity increased with the increase of contact time. The adsorption amount increased rapidly at the start of the adsorption reaction because of the polymer had many binding sites, but adsorption capacity decreased and reached the equilibrium gradually due to available sites had been occupied. HMINs-2 and HNINs-2 could reach equilibrium within about 50 min displaying a fast dynamics performance. Pseudo-second-order model fittings of HMINs-2 and HNINs-2 for CAP are shown in Fig. 8 and the fitting parameters are listed in Table 2. The correlation coefficient R² of HMINs-2 and HNINs-2 were closed to 1 showed a good linear correlation with the experiment data, and Fig. 8a exhibited a visual proof of this conclusion. The values of Q_e calculated through pseudo-second-order kinetic model were more closed to the experiment value as shown in Fig. 8b. Thus, the HMINs-2 and HMINs-2 could be well described by the pseudo-second-order kinetic model, indicated that the chemical adsorption process was mainly rate-controlled step.

Fig. 8 Pseudo-second-order model linear fitting of HMINs-2 and HNINs-2 toward CAP (a); kinetics data using pseudo-second-order kinetic model of HMINs-2 and HNINs-2 (b).

Table 2 Kinetics parameters of HMINs-2 and HNINs-2 for CAP

Adsorbents	Qe,exp (μmol g^−1)	Pseudo-first-order			Pseudo-second-order
		Qe,c (μmol g^−1)	k1 (min^−1)	R^2	Qe,c (μmol g^−1)	k2 (g μmol^−1 min^−1)	R^2
HMINs-2	67.86	24.92	0.0279	0.8249	70.32	0.0023	0.9991
HNINs-2	41.88	8.143	0.0244	0.5820	42.36	0.0076	0.9990

---

3.4 Selectivity adsorption performance

The selectivity performance tests are vital for the study of specific recognition mechanism of the as-prepared HMINs-2 to template molecules. Thus, CTC, CIP, CFX and TC were chose as reference antibiotics to research selective ability of HMINs-2 for CAP. As can be seen from Fig. 9, HNINs-2 had a high adsorption capacity for CAP, which might be caused by the small steric hindrance of CAP and the formation of hydrogen bonds between CAP and carboxyl groups in the polymer shell. Adsorption amount of HMINs-2 was higher than that of HNINs-2, suggesting the existence of imprinted cavities. Also, it can be readily found the HMINs-2 displayed a much higher binding capacity of CAP than that of other reference antibiotics, indicating the regeneration of abundant specific recognition sites during the imprinted procedure, and the different imprinted interaction may be based on the structure, functional groups and distinct sizes of CAP template molecules. During the imprinted polymerization, the binding cavities predetermined by size and structure of CAP molecules were obtained after the removal of CAP in the fixation of the linking groups. Importantly, the presence of the imprinted sites between CAP and functional monomer in the obtained binding cavities was significantly crucial for rebinding of CAP molecules. Thereby, the obtained imprinted cavities in the HMINs could recognize CAP molecules instead of other reference antibiotics. Accordingly, the HMINs-2 undoubtedly exhibits a highest adsorption capacity for CAP, indicating our HMINs-2 could be used as an excellent molecularly imprinted nanoadsorbent for selective separation of CAP.

Fig. 9 Chemical structure of reference antibiotics and adsorption amount of HMINs-2 and HNINs-2 towards antibiotics.

From Table 3, we can find the adsorption amount of HMINs-2 for CTC, CIP, CFX, TC and CAP are 26.97, 7.87, 3.53, 20.69 and 67.86 μmol g⁻¹, respectively, indicating the HMINs-2 has best adsorption capacity for CAP. The distribution coefficient K_d of CAP in HMINs-2 was higher than other antibiotics, which indicated that the distribution of CAP in HMINs-2 was the highest. The k values of HMINs-2 for CTC, CIP, CFX and TC were 3.295, 4.284, 28.63 and 12.54, respectively, indicated the order of selectivity was CAP > CTC > TC > CIP > CFX. k′ represented the selectivity of HMINs-2 and HNINs-2 on the CAP. All of the k′ values were more than 1.417 (Table 3), showing HMINs-2 has a high specificity for the CAP template molecules. The above results showed a successful imprinted synthesis process of HMINs-2 with specificity identification performance for the CAP.

Table 3 Adsorption selectivity of HMINs-2 and HNINs-2 for CAP

Antibiotics	HMINs-2				HNINs-2
	Ce (μmol L^−1)	Qe (μmol g^−1)	Kd (L g^−1)	k	Ce (μmol L^−1)	Qe (μmol g^−1)	Kd (L g^−1)	k	k′
CTC	86.52	26.97	0.3117	3.295	89.77	20.45	0.2278	2.325	1.417
CIP	96.07	7.87	0.0819	12.54	96.55	6.89	0.0714	7.421	1.689
CFX	98.24	3.53	0.0359	28.63	98.41	3.18	0.0323	16.43	1.743
TC	86.29	20.69	0.2398	4.284	90.43	19.13	0.2116	2.503	1.711
CAP	66.07	67.86	1.0270		79.06	41.88	0.5297

---

3.5 Regeneration performance

The cyclic adsorption–desorption experiments were conducted to investigate the regeneration and stability. As shown in Fig. 10, after 6 times of adsorption and regeneration, the adsorption capacity of HMINs-2 for CAP was not significantly decreased (only down 9.31%) suggested HMINs with a high stability and superior persistence for adsorbing CAP, which give promise of enormous potential for practical applications in waste water treatment.

Fig. 10 Reusability of HMINs-2.

4 Conclusions

In summary, we successfully prepared hollow molecular imprinting nanorods with controllable thickness by combining sacrificial template (HNTs@KH570) and in situ surface precipitation polymerization method for selective recognition and rapid separation of CAP from water environment. The HMINs presented hollow rod-like structure and the imprinted shell thickness could be controlled by changing the monomer content. The adsorption experiments showed that the HMINs-2 with a shell thickness of 62 nm had the highest adsorption capacity for CAP. The Freundlich model could well fit the adsorption equilibrium data indicated the multi-molecular layer adsorption characteristics. Meanwhile, adsorption of HMINs-2 for CAP rapidly reached the equilibrium within 50 min showed a fast kinetics property. Also, the HMINs possessed highly specific recognition to CAP template from water solutions as compared with other reference antibiotics. Moreover, the HMINs-2 showed excellent thermal stability and regeneration properties for the potential practical application in water treatments. The facile and general fabrication of novel hollow molecular imprinting nanorods nanoadsorbents is promising for application in selective remove various organic or inorganic contaminations from water environment.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 21277063, 21407058, 21446015, 21546013 and U1510126), the National Basic Research Program of China (973 Program, 2012CB821500), Natural Science Foundation of Jiangsu Province (BK20140534), and Research Fund for the Doctoral Program of Higher Education of China (20133227110022 and 20133227110010).

References

1. H. Zhao, X. Liu, Z. Cao, Y. Zhan, X. Shi, Y. Yang, J. Zhou and J. Xu, Adsorption behavior and mechanism of chloramphenicols, sulfonamides, and non-antibiotic pharmaceuticals on multi-walled carbon nanotubes, J. Hazard. Mater., 2016, 310, 235–245.
2. M. B. Ahmed, J. L. Zhou, H. H. Ngo and W. Guo, Adsorptive removal of antibiotics from water and wastewater: progress and challenges, Sci. Total Environ., 2015, 532, 112–126.
3. A. T. M. Din, M. A. Ahmad and B. H. Hameed, Ordered mesoporous carbons originated from non-edible polyethylene glycol 400 (PEG-400) for chloramphenicol antibiotic recovery from liquid phase, Chem. Eng. J., 2015, 260, 730–739.
4. R. Ao, C. Liu, Y. Hong, W. Shi, L. Shuang and L. Ping, Enhanced visible-light-driven photocatalytic activity for antibiotic degradation using magnetic NiFe₂O₄/Bi₂O₃ heterostructures, Chem. Eng. J., 2014, 258, 301–308.
5. A. A. S. Al-Gheethi and N. Ismail, Biodegradation of pharmaceutical wastes in treated sewage effluents by bacillus subtilis 1556WTNC, Environmental Processes, 2014, 1, 459–481.
6. J. He, J. Dai, T. Zhang, J. Sun, A. Xie, S. Tian, Y. Yan and P. Huo, Preparation of highly porous carbon from sustainable α-cellulose for superior removal performance of tetracycline and sulfamethazine from water, RSC Adv., 2016, 6, 28023–28033.
7. A. Xie, J. Dai, J. He, J. Sun, Z. Chang, Y. Yan and C. Li, Converting obsolete copy paper to porous carbon materials with preeminent adsorption performance of tetracycline antibiotic, RSC Adv., 2016, 6, 13312–13322.
8. A. Garcia-Rodríguez, V. Matamoros, S. D. Kolev and C. Fontàs, Development of a polymer inclusion membrane (PIM) for the preconcentration of antibiotics in environmental water samples, J. Membr. Sci., 2015, 492, 32–39.
9. R. Ahmad, N. Griffete, A. Lamouri, N. Felidj, M. M. Chehimi and C. Mangeney, Nanocomposites of gold nanoparticles@molecularly imprinted polymers: chemistry, processing, and applications in sensors, Chem. Mater., 2015, 27, 5464–5478.
10. D. Gao, Z. Zhang, M. Wu, C. Xie, G. Guan and D. Wang, A surface functional monomer-directing strategy for highly dense imprinting of TNT at surface of silica nanoparticles, J. Am. Chem. Soc., 2007, 129, 7859–7866.
11. J. Dai, J. He, A. Xie, L. Gao, J. Pan, X. Chen, Z. Zhou, X. Wei and Y. Yan, Novel pitaya-inspired well-defined core-shell nanospheres with ultrathin surface imprinted nanofilm from magnetic mesoporous nanosilica for highly efficient chloramphenicol removal, Chem. Eng. J., 2016, 284, 812–822.
12. D. H. Carrales-Alvarado, R. Ocampo-Pérez, R. Leyva-Ramos and J. Rivera-Utrilla, Removal of the antibiotic metronidazole by adsorption on various carbon materials from aqueous phase, J. Colloid Interface Sci., 2014, 436, 276–285.
13. K. Flavin and M. Resmini, Imprinted nanomaterials: a new class of synthetic receptors, Anal. Bioanal. Chem., 2009, 393, 437–444.
14. R. Jalili, Surface molecular imprinting on silane-functionalized carbon dots for selective recognition of nifedipine, RSC Adv., 2015, 5, 74084–74090.
15. C. Qiu, Y. Xing, W. Yang, Z. Zhou, Y. Wang, H. Liu and W. Xu, Surface molecular imprinting on hybrid SiO₂-coated CdTe nanocrystals for selective optosensing of bisphenol A and its optimal design, Appl. Surf. Sci., 2015, 345, 405–417.
16. C. Shaona, L. Xin, Z. Yuanyuan, C. Limin and Q. Jingyao, High performance surface-enhanced Raman scattering via dummy molecular imprinting onto silver microspheres, Chem. Commun., 2014, 50, 14331–14333.
17. Y. Hao, R. Gao, D. Liu, Y. Tang and Z. Guo, Selective extraction of gallic acid in pomegranate rind using surface imprinting polymers over magnetic carbon nanotubes, Anal. Bioanal. Chem., 2015, 407, 1–10.
18. S. Bhakta, S. Seraji, S. L. Suib and J. F. Rusling, Antibody-like biorecognition sites for proteins from surface imprinting on nanoparticles, ACS Appl. Mater. Interfaces, 2015, 7, 28197–28206.
19. J. Cui, Y. Wu, M. Meng, J. Lu, C. Wang, J. Zhao and Y. Yan, Bio-inspired synthesis of molecularly imprinted nanocomposite membrane for selective recognition and separation of artemisinin, J. Appl. Polym. Sci., 2016, 133, 43405.
20. S. H. Yu, B. Quan, A. Jin, K. S. Lee, S. H. Kang, K. Kang, Y. Piao and Y. E. Sung, Hollow nanostructured metal silicates with tunable properties for lithium ion battery anodes, ACS Appl. Mater. Interfaces, 2015, 7, 25725–25732.
21. H. Lv, Achieving hierarchical hollow carbon@Fe@Fe₃O₄ nanospheres with superior microwave absorption properties and lightweight features, J. Mater. Chem. C, 2015, 3, 10232–10241.
22. H. Lun, J. Ouyang and H. Yang, Natural halloysite nanotubes modified as an aspirin carrier, RSC Adv., 2014, 4, 44197–44202.
23. J. Dai, X. Wei, Z. Cao, Z. Zhou, P. Yu, J. Pan, T. Zou, C. Li and Y. Yan, Highly-controllable imprinted polymer nanoshell at the surface of magnetic halloysite nanotubes for selective recognition and rapid adsorption of tetracycline, RSC Adv., 2014, 4, 7967.
24. S. Zuo, W. Liu, C. Yao, X. Li, Y. Kong, X. Liu, H. Mao and Y. Li, Preparation of polyaniline–polypyrrole binary composite nanotube using halloysite as hard-template and its characterization, Chem. Eng. J., 2013, 228, 1092–1097.
25. D. Papoulis, S. Komarneni, D. Panagiotaras, E. Stathatos, D. Toli, K. C. Christoforidis, M. Fernández-García, H. Li, S. Yin and T. Sato, Halloysite–TiO₂ nanocomposites: synthesis, characterization and photocatalytic activity, Appl. Catal., B, 2013, 132–133, 416–422.
26. S. Zuo, W. Liu, C. Yao, X. Li, Y. Kong, X. Liu, H. Mao and Y. Li, Preparation of polyaniline–polypyrrole binary composite nanotube using halloysite as hard-template and its characterization, Chem. Eng. J., 2013, 228, 1092–1097.
27. H. Huang, Y. Ding, X. Chen, Z. Chen and X. Z. Kong, Synthesis of monodisperse micron-sized poly(divinylbenzene) microspheres by solvothermal precipitation polymerization, Chem. Eng. J., 2016, 289, 135–141.
28. L. Su, X. Guo and S. Han, Preparation and evaluation of vanillin molecularly imprinted polymer microspheres by reversible addition–fragmentation chain transfer precipitation polymerization, Anal. Methods, 2014, 6, 2512–2517.
29. C. Zha, W. Wang, Y. Lu and L. Zhang, Constructing covalent interface in rubber/clay nanocomposite by combining structural modification and interlamellar silylation of montmorillonite, ACS Appl. Mater. Interfaces, 2014, 6, 18769–18779.
30. J. Dai, J. Pan, L. Xu, X. Li, Z. Zhou, R. Zhang and Y. Yan, Preparation of molecularly imprinted nanoparticles with superparamagnetic susceptibility through atom transfer radical emulsion polymerization for the selective recognition of tetracycline from aqueous medium, J. Hazard. Mater., 2012, 205–206, 179–188.
31. I. Langmuir, The adsorption of gases on plane surfaces of glass, mica and platinum, J. Am. Chem. Soc., 1918, 40, 1361–1403.
32. H. M. F. Freundlich, Over the adsorption in solution, J. Phys. Chem., 1906, 57, 385–471.
33. L. Zhang, T. Zhu, X. Liu and W. Zhang, Simultaneous oxidation and adsorption of As(III) from water by cerium modified chitosan ultrafine nanobiosorbent, J. Hazard. Mater., 2016, 308, 1–10.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra08042a

---