Journal of Chromatography A, 1361 (2014) 23–33

Contents lists available at ScienceDirect

Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma

Fabrication of enrofloxacin imprinted organic–inorganic hybrid mesoporous sorbent from nanomagnetic polyhedral oligomeric silsesquioxanes for the selective extraction of fluoroquinolones in milk samples Hai-Bo He a,b,∗ , Chen Dong a , Bin Li a , Jun-Ping Dong a , Tian-Yu Bo a , Tian-Lin Wang a , Qiong-Wei Yu b , Yu-Qi Feng b,∗∗ a

Department of Chemistry, Shanghai University, Shanghai 200444, China Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), Department of Chemistry, Wuhan University, Wuhan 430072, China b

a r t i c l e

i n f o

Article history: Received 28 May 2014 Received in revised form 28 July 2014 Accepted 29 July 2014 Available online 16 August 2014 Keywords: Fe3 O4 @POSS Molecularly imprinted hybrid materials Magnetic solid-phase extraction Fluoroquinolones Milk samples

a b s t r a c t This paper reports a nanomagnetic polyhedral oligomeric silsesquioxanes (POSS)-directing strategy toward construction of molecularly imprinted hybrid materials for antibiotic residues determination in milk samples. The imprinted polymeric layer was facilely obtained through the copolymerization of active vinyl groups present on the nanomagnetic POSS (Fe3 O4 @POSS) surface and functional monomer (methacrylic acid) binding with template (enrofloxacin). Herein, the octavinyl POSS acted as not only the building blocks for hybrid rigid architectures but also the cross-linker for the formation of effective recognition sites during the imprinting process. The molecularly imprinted Fe3 O4 @POSS nanoparticles (Fe3 O4 @MI-POSS) demonstrated much higher adsorption capacity and selectivity toward enrofloxacin molecules and its analogs than the non-imprinted Fe3 O4 @POSS (Fe3 O4 @NI-POSS) materials. The imprinted particles were applied as a selective sorbent in solid-phase extraction focusing upon sample pretreatment in complex matrices prior to chromatographic analysis. The three FQs (ofloxacin, enrofloxacin, danofloxacin) could be selectively extracted from the biological matrix, while the matrix interferences were effectively eliminated simultaneously under the optimum extraction conditions. A simple, rapid and sensitive method based on the Fe3 O4 @MI-POSS material combined with HPLC-UV detection was then established for the simultaneous determination of three FQs from milk samples. The average recoveries of the three FQs were in the range of 75.6–108.9%. The relative standard deviations of intra- and inter-day ranging from 2.91 to 8.87% and from 3.6 to 11.5%, respectively. The limits of detections (S/N = 3) were between 1.76 and 12.42 ng mL−1 . It demonstrates the effectiveness of trace analysis in complicated biological matrices utilizing magnetic separation in combination with molecularly imprinted solid-phase extraction, the rich chemistry of POSS makes it possible to be an ideal platform for generating molecular imprinted hybrid materials is also exhibited. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Molecular imprinting is a well-known technique for the creation of tailor-made binding sites with memory of the shape, size,

∗ Corresponding author at: Department of Chemistry, Shanghai University, Shanghai 200444, China. Tel.: +086 21 66132930; fax: +86 21 66134594. ∗∗ Corresponding author. Tel.: +86 27 68755595; fax: +86 27 68755595. E-mail addresses: [email protected] (H.-B. He), [email protected] (Y.-Q. Feng). http://dx.doi.org/10.1016/j.chroma.2014.07.089 0021-9673/© 2014 Elsevier B.V. All rights reserved.

and functional groups of the template. The molecular recognition systems are usually created by a polymer network involves the copolymerization of functional monomers and cross-linkers in the presence of the template, and the removal of template generates the recognition sites (cavities) which can selectively rebind the template molecules from a mixture of closely related compounds [1,2]. Compared to other recognition systems, molecularly imprinted polymers (MIPs) possess many promising characteristics, such as low cost and easy synthesis, high stability to harsh chemical and physical conditions and excellent reusability [1,2]. Therefore, MIPs have been widely applied in the fields of

24

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

extraction/separation [3–5], chemical sensors [6] and catalysis [7]. In recent years, MIPs are becoming good alternative solid-phase extraction (SPE) sorbents for sample preparation. The most significant advantages of MIPs as compared to the conventional SPE sorbents, are the superior selective affinities toward the targeted template analyte existing in complex matrix such as biological and food samples [8]. Magnetic SPE (MSPE) is a SPE in dispersion mode based on the use of magnetic sorbents [9]. In this procedure, the magnetic sorbent need not be packed into the SPE cartridge like the traditional SPE, and it can be dispersed into the solution directly and then easily separated from the matrix using an external magnetic field without additional centrifugation or filtration, which makes the sample pretreatment easier and faster. MIP-based magnetic composites can combine the favorable molecular recognition selectivity with additional functions such as magnetic susceptibility. Recent years have witnessed rapidly increasing interest in application of such magnetic MIPs as SPE sorbents in sample preparation [3–5,8,10], because the efficient magnetic separation allows the selective extracting of the target analytes in a convenient and economical way, it is particularly very useful for the analysis of complex samples [3,10–12]. Different methods have been developed to prepare magnetic MIPs. The most commonly used to prepare magnetic MIPs is based on vinyl-functionalized Fe3 O4 magnetic hybrid materials. The imprinting process could be completed easily through a simple free radical polymerization occurring in the presence of the template molecule, functional monomer and cross-linker [2,10]. However, prior to the grafting of MIP layers, time-consuming surface modification of magnetic particles is necessary to produce the base platform on which the imprinted polymer is formed. Typically, the preparation process first involves preparation of the silica shell coated Fe3 O4 magnetic nanoparticles (Fe3 O4 @SiO2 ) through Stöber method and then location of end vinyl groups on the surface of Fe3 O4 @SiO2 by silylation using a silylating agent, such as 3(trimethoxysilyl)propyl methacrylate [2,10,13–22], which is highly susceptible and tedious (such as drying of the solvents, repeating silylation cycles to obtain desirable functionalities). Moreover, the low densities of vinyl groups present on the Fe3 O4 @SiO2 via silylation was disadvantageous for grafting of high-performance imprinting layer [23], thereby largely limited the broad application of magnetic MIPs. Therefore, facile and versatile fabrication schemes to richen the surface end vinyl groups in the magnetic hybrid architecture for preparing efficient magnetic MIPs as SPE sorbents are highly desirable [11]. Polyhedral oligomeric silsesquioxanes (POSS) are organic–inorganic hybrid materials at a molecular level [24]. A typical polyhedral oligomeric silsesquioxane molecule possesses a cubic rigid (T8 ) structure represented by the formula R8 Si8 O12 , where the central inorganic core (Si8 O12 ) is functionalized with organic moieties (R) at each of the eight vertices [25]. POSS are recently regarded as ideal building blocks for constructing a variety of functional composites due to their nano-sized as well as hybrid molecular frameworks [26]. For example, Nischang et al. [27,28] proposed a facile, single-step process by using octavinyl POSS as the sole building blocks in polymerization to construct versatile, high-surface-area, hierarchically structured hybrid materials. And recently a series of novel POSS-based functionalizable porous hybrid monoliths were prepared in a facile, highly flexible pathway wherein POSS were employed as the cross-linker [29,30] and monomer [31], respectively. It is worth mentioning that we have successfully prepared a mesoporous nanomagnetic hybrid material based on octavinyl POSS (Fe3 O4 @POSS) by a facile polymerization step [32]. In fact the polymers formed from such cages possess a number of residual vinyl groups [27,28], therefore the modification of Fe3 O4 @POSS was implemented by introducing

dithiol organic anchors via thiol-ene addition reaction. The resulting Fe3 O4 @POSS-SH turned out to be an ideal single adsorbent for purifying wastewater coexisting with inorganic heavy metal ions and organic dyes at room temperature. The preliminary exploitation demonstrated the tailorability of Fe3 O4 @POSS owing to the residual vinyl groups. On the basis of the previous research work, herein we propose a new strategy for fabrication of magnetic MIPs taking the Fe3 O4 @POSS as the imprinting platform. Nowadays, the extensive use of antibiotics to animals has been of food security concerns. As one of the most important class of synthetic antibacterials, FQs have been used as veterinary drugs for animals worldwide [33]. The presence of FQs in edible animal products is a significant risk which can be directly toxic or cause of pathogen resistance in humans. Therefore, monitoring FQs at trace levels in a sensitive and selective manner is necessary to ensure that food is completely free from antibiotic residues. Many analytical methods for the determination of FQ residues in animal-producing food are described in the scientific literature. Most of these rely on HPLC methods using ultra-violet (UV) [34,35], fluorescence (FD) [36–38] or mass spectrometric (MS) detection [38–40]. It should be noted that due to the complexity of food samples, the development of sample pretreatment procedures prior to instrumental analysis showing improved selectivity seems of overwhelming importance to reduce matrix interferences [41], particularly for where a relatively expensive and complex instrumentation (such as LC–MS) is not available for routine monitoring. As a solution, several extraction processes such as liquid–liquid extraction (LLE), solid phase extraction(SPE), supercritical fluid extraction (SFE), dispersive liquid–liquid microextraction (DLLME), stir bar sorptive extraction (SBSE), microwave-assisted extraction (MAE) and even combination of the basic ones have been suggested [34,41]. Among them, SPE is the most frequently used technique for the preconcentration and cleanup of multiple matrix samples, but also restricted by defective purifying effect caused by those conventional sorbents [34]. Given the superiority of the so-called MSPE as to the sensitivity and labor efficiency required, the exploration of magnetic MIP sorbents using FQs as the templates is highly desirable. In the present study, enrofloxacin was used as the template to demonstrate the feasibility and efficiency of Fe3 O4 @POSS-directing strategy toward construction of molecularly imprinted hybrid materials. The selective extraction performance of the as-prepared enrofloxacin-imprinted nanomagnetic hybrid material was evaluated by extraction of three FQs from milk samples with HPLC-UV analysis. As far as we know, this is a first attempt on manufacturing nanomagnetic molecular imprinted hybrid material based on POSS for sample pretreatment in complex samples. The approach reported here should be adaptable for fabricating various magnetic MIPs hybrid nanoparticles, thereby extensive applications can be find beyond residue analysis in food samples.

2. Experimental 2.1. Chemicals and reagents Octavinyl POSS, ferric chloride (FeCl3 ·6H2 O), ferrous chloride (FeCl2 ·4H2 O), methacrylic acid (MAA), chloroform, and 2,2 azobisobutyronitrile (AIBN) were purchased from Aladdin Chemistry Co. Ltd. AIBN was recrystallized in ethanol before use. Sodium hydroxide (NaOH), disodium hydrogenphosphate (Na2 HPO4 ), sodium phosphate (NaH2 PO4 ), phosphoric acid (H3 PO4 ), acetic acid (HAc) and trifluoroacetic acid (TFA) were obtained from Sinopharm Chemical Reagent Co., Ltd. (China). Methanol (MeOH) and acetonitrile (ACN) of HPLC grade were got from Adamas-beta Co., Ltd. Ofloxacin (OFL), enrofloxacin (ENR), danofloxacin (DAN) and chloramphenicol (CAP) were purchased from Laboratories of

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

25

Fig. 1. Chemical structures of ofloxacin (OFL), danofloxacin (DAN), enrofloxacin (ENR), and chloramphenicol (CAP).

Dr. Ehrenstorfer (Augsburg, Germany). Individual stock solutions of OFL, ENR, and CAP were prepared in a concentration of 500 mg L−1 stock solution in ACN and stock solution of DAN was prepared in water at a concentration of 500 mg L−1 . The working standard solution was diluted with phosphate buffer solution (PBS, 25 mmol L−1 ) to the desired concentration for experiments. All of the above solutions were maintained in refrigerator at 4 ◦ C. Chemical structures of the studied FQs and CAP in this work are shown in Fig. 1.

2.2. Preparation of Fe3 O4 @MI-POSS The preparation protocol of the ENR molecular imprinted hybrid structure magnetic nanoparticles is shown in Fig. 2. The Fe3 O4 @POSS composite was first prepared by surface polymerization of octavinyl POSS on the Fe3 O4 nanoparticles according to our previous work [32]. Briefly, POSS (400 mg) was dissolved in binary solvent mixture of 2700 ␮L of THF and 535 ␮L of PEG 200. Then 255 mg of acrylic acid coated Fe3 O4 nanoparticles was added to the suspension and was ultrasonically mixed for 5 min. The polymerization was initiated by the addition of 96 mg of AIBN at 60 ◦ C for 24 h. The nanocomposites were collected by an external magnet separation then were subjected to extraction with THF in a Soxhlet apparatus until the eluent was transparent. The obtained Fe3 O4 @POSS nanoparticles were dried in vacuum before use. The Fe3 O4 @MI-POSS was synthesized by imprinting polymerization on the surface of Fe3 O4 @POSS, wherein MAA and immobilized POSS were used as the functional monomer and cross-linking agent, respectively. Typically, the template molecule (ENR, 1 mmol) and the functional monomer (MAA, 8 mmol) were mixed adequately in 20 mL of chloroform to form pre-assembly solution. After stirring the mixture for 1 h, 20 mL of chloroform with 500 mg of dispersed Fe3 O4 @POSS and 100 mg of AIBN were added to the mixture. The suspension solution was ultrasonically degassed for 10 min, followed by stirring at 60 ◦ C for 24 h under an N2 atmosphere. After the polymerization, the resultant nanoparticles were magnetically collected and the original ENR template in the organosiliceous cube shell was extracted with MeOHHAc-TFA (90:9:1, v/v/v) under ultrasound. Finally, the obtained Fe3 O4 @MI-POSS was washed with distilled water and MeOH in

sequence and dried at 60 ◦ C in a vacuum oven overnight prior to use. For comparison, the non-imprinted Fe3 O4 @POSS composite (Fe3 O4 @NI-POSS) was prepared under the above-described conditions in the absence of the template. 2.3. Instrumentation and analytical conditions The morphologies and dimensions of the nanomagnetic composites were examined by an S-4800 field emission scanning electron microscope (FE-SEM, Hitachi, Japan) at an accelerating voltage of 20 kV. Fourier transform infrared (FT-IR) spectra (KBr pellets) over the range of 400–4000 cm−1 were performed on a Nicolet Avatar 370 spectrometer (Nicolet Thermo, U.S.A.). The nitrogen adsorption/desorption isotherms were measured by using an ASAP 2020 N2 adsorption and desorption analyzer (Micromeritics Co., Ltd., U.S.A.). Surface areas were calculated using the Brunauer–Emmett–Teller (BET) equation from the desorption isotherms. Pore diameters were determined from the desorption branch of the isotherm using the Barrett–Joyner–Halenda (BJH) method. The total pore volumes (desorption cumulative volumes) were calculated at a relative pressure of P/P0 above 0.99. The samples were degassed under vacuum at 120 ◦ C for 5 h prior to measuring at −196 ◦ C. Magnetic properties were analyzed using a vibrating sample magnetometer (VSM, Lakeshore Model-7410, U.S.A.). The adsorption tests were performed on a incubator shaker (Noki NHWY-100B, Shanghai). The HPLC profiles were recorded using a Shimadzu LC-2010AHT (Tokyo, Japan) HPLC system. The analytes were separated on a Hisep C18 -T column (250 × 4.6 mm I.D., 5 ␮m) from Weltech (Wuhan) Co., Ltd. The optimized mobile phase was water-ACN (8:2, containing 0.1% TFA). Aliquots of 10 ␮L sample were injected into the column and the UV wavelength was set at 280 nm. The chromatographic assay was carried out at ambient temperature and the flow rate was maintained at 0.5 mL min−1 . 2.4. Evaluation of adsorption properties The binding experiment was carried out by adding 20.0 mg of Fe3 O4 @MI-POSS or Fe3 O4 @NI-POSS nanoparticles in glass tubes

26

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

Fig. 2. Schematic illustration of molecular imprinting on nanomagnetic POSS.

containing 5.0 mL of ENR stock solution which was prepared in water with concentrations varying from 0.10 to 1.75 mmol L−1 . The solution was incubated for 2 h with continuous shaking in a horizontal shaker at room temperature. After the adsorbents had been isolated by an external magnet, the supernatant was transferred for HPLC analysis. The static adsorption capacities of the Fe3 O4 @MI-POSS and Fe3 O4 @NI-POSS materials were calculated based on the following formula [32]: Qe =

(C0 − Ce )V M

is the equilibrium adsorption capacities, C0 where Qe (mg L−1 ) is the initial concentration, Ce (mg L−1 ) is the equilibrium concentration, V (L) is the volume of the initial aqueous solution and M (g), M (g) is the mass of the adsorbent used. The selectivity of Fe3 O4 @MI-POSS was investigated with ENR, OFL and DAN as the structural analogs of ENR template, and CAP as the reference compound. The experiment was carried out by adding 20.0 mg of Fe3 O4 @MI-POSS or Fe3 O4 @NI-POSS in glass tubes containing 4.00 mL of mixture solution with each standard solution at a concentration of 15 mg L−1 . The solution was incubated for 2 h at room temperature, and then the supernatant was separated and analyzed by HPLC with UV detection at 280 nm. The gradient elution was carried out starting from 100% water-ACN (8:2, containing 0.1% TFA) in first 30 min for FQs and then held 100% ACN in following 20 min for CAP. The flow rate was maintained at 0.5 mL min−1 . The distribution coefficient (Kd , mL g−1 ) suggests the ratio of the binding capacity of sorbent to the free analytes concentration in the supernatant, which is determined by the following equation [42,43]: (C0 − Ce )V Qe = Ce M Ce

k=

(2)

Kd(FQS ) Kd(CAP)

(3)

The relative selectivity coefficient (k ) represents the otherness of two sorbents, which is calculated with the following equation [42,43]:

(1)

(mg g−1 )

Kd =

The selectivity coefficients (k) of the sorbent reveals the otherness of two substances adsorbed by one sorbent, which is defined as the relative distribution coefficients of FQs (Kd(FQS ) ) to that of CAP (Kd(CAP) ) [42,43]:

k =

kFe3 O4 @MI-POSS kFe3 O4 @NI-POSS

(4)

2.5. Molecularly imprinted magnetic solid-phase extraction (MI-MSPE) of FQs from milk samples The milk samples were randomly obtained from the local supermarket (near to Shanghai University) and were stored at −4 ◦ C in dark until use. Prior to extraction, all milk samples were pretreated according to the literatures [44] with a little procedure modification. 0.5 mL of milk samples spiked with known variable amounts of FQs standard solution (the volume added was always less than 2% of the final sample volume to preserve the approximate volume of the samples) were vortexed with 0.5 mL of ACN for 5.0 min. After being equilibrated at room temperature for 15 min, the mixtures were centrifuged for 5 min at 10 000 rpm. The supernatants were totally pipetted and diluted with PBS (pH 6.0, 25 mmol L−1 ) to 10 mL. Blank samples were prepared in the same way as above but without the FQs spiking step. An amount of 60 mg Fe3 O4 @MI-POSS nanoparticles were put into a glass vial and preconditioned in sequence with 3.0 mL of MeOH and 3.0 mL of PBS (25 mmol L−1 , pH 6.0). The supernatant was isolated from the composite with a Nd–Fe–B permanent

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

27

Fig. 3. Representative scanning electron micrographs of (a) Fe3 O4 , (b) Fe3 O4 @POSS, (c) Fe3 O4 @NI-POSS and (d) Fe3 O4 @MI-POSS.

magnet and discarded. Then 10 mL of sample solution was added into the bottle and the mixture was shaking for 30 min at room temperature. Subsequently, the MI-material captured FQs were collected rapidly from the solution under external magnetic field and washed with 3.0 mL of water. Finally, the FQs were desorbed from the MI-material with 2.0 mL of MeOH–HAc–TFA (90:9:1, v/v/v) under the action of ultrasound for 10 min. The desorption solution was collected under an external magnetic field and injected directly into the HPLC. 3. Results and discussion 3.1. Preparation and characterization of Fe3 O4 @MI-POSS In general, the Fe3 O4 @POSS, Fe3 O4 @MI-POSS and the corresponding Fe3 O4 @NI-POSS were prepared via a simple free-radical polymerization process occurred in octavinyl POSS. As illustrated in Fig. 1, firstly, a functionalizable organosiliceous hybrid magnetic material was constructed by surface polymerization of POSS on the Fe3 O4 nanoparticles as reported by our group earlier [32]. The resultant Fe3 O4 @POSS can be a promising support of molecularly imprinted material. The MI-material could be tailored by a simple modification of the residual vinyl groups and subsequent copolymerization with the functional monomers being attached to the specific template molecule. There are silica cages that are connected through formed alkyl linkages between cages, which would be ideally suitable for the formation and stabilization of delicate recognition sites. Herein, a complex is prepared by self-assembling of the template of ENR and the functional monomer MAA due to the formation of hydrogen bond interactions. Then the ENR template molecule was introduced into the cross-linked silica layer by initiating a second polymerization reaction of organic monomers. Finally, the template embedded nanoparticles were washed several times with MeOH–HAc–TFA (90:9:1, v/v/v) under ultrasound. The eluate was collected and monitored by HPLC analysis. The template was estimated to be effectively removed from the Fe3 O4 @POSS when no ENR peak was detected in the eluate (namely, washing solvent

fixed at 2 mL) within the limits of detection. The resultant material was expected to possess a rigid three-dimensional network with accessible binding cavities based on the ENR molecule. The FT-IR spectra of Fe3 O4 @POSS in Fig. S1(a) with intense bands at 3050, 1600, 1410, and 1275 cm−1 revealed the existence of a number of residual vinyl groups. As displayed in Fig. S1(b) and (c) as ) of the two derived materials, the alkyl bands at 2930 cm−1 (CH 2 increased, while the bands associated to residual vinyl groups reduced, evidencing the occurrence of further polymerization reaction. The C O stretching band at 1735 cm−1 , COO− asymmetric stretching band at 1620 cm−1 , appear obviously in the Fig. S1(b) and (c) spectra, implying the successful graft of MAA by modification of Fe3 O4 @ POSS. The spectra of Fig. S1(b) and (c) are found to be very similar and no other obvious peaks related to ENR are observed, further evidence that the ENR template has been removed effectively by Soxhlet extraction with THF and ultrasonic wash with MeOH-HAc-TFA (90:9:1, v/v/v). Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2014.07.089. The typical SEM images of the as-prepared Fe3 O4 , Fe3 O4 @POSS, Fe3 O4 @NI-POSS and Fe3 O4 @MI-POSS materials are shown in Fig. 3 at 80 000 × magnification. All the SEM micrographs show some agglomerates with distribution of bead sizes, which may be ascribed to magnetic dipole interaction. However, there are obvious differences in appearances between the pristine Fe3 O4 and the modified Fe3 O4 materials, suggesting the polymerizations in different extends have occurred on the Fe3 O4 surface. Fig. 3(a) indicates that the Fe3 O4 material is nearly spherical in appearance with an average diameter of 10 nm, while Fig. 3(b) shows rough surfaces of the Fe3 O4 @POSS with the average particle size about 20 nm. The distinguished differences in appearance between the Fe3 O4 @POSS (in Fig. 3(b)) and its derived hybrid materials (in Fig. 3(c) of Fe3 O4 @NI-POSS and Fig. 3(d) of Fe3 O4 @MI-POSS) could also be observed, revealing further polymerizations occurred on the Fe3 O4 @POSS in the presence of MAA and AIBN. Fig. 3(c) and (d) illustrate that there is no obvious difference in the appearance of the

28

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

Fig. 4. Nitrogen adsorption–desorption isotherms of (a) Fe3 O4 @POSS, (b) Fe3 O4 @MI-POSS and (c) Fe3 O4 @NI-POSS (the insets are corresponding pore size distribution curves).

two Fe3 O4 @ POSS derived material, both of which show a agglomerated bed-like appearance. And such agglomerated beds are made up of sphere-like nanoparticles with average sizes of about 30 nm. It should be pointed out that the influence on material’s architecture resulting from imprinting would not be expected to be seen with SEM on the nanoscale. Therefore, further characterizations on the two materials by nitrogen adsorption/desorption analyses were carried out. As shown in Fig. 4 that probed by nitrogen adsorption/desorption analysis, the Fe3 O4 @POSS and its derivatives (Fe3 O4 @MI-POSS and Fe3 O4 @NI-POSS) behave the type IV isotherms with steep hysteresis loop at relative pressure (P/P0 ) of 0.45, indicating that all of the three materials have mesoporous structures that originate from the highly cross-linked POSS units. Data gathered in Table 1 illustrate that as compared to the pristine Fe3 O4 @POSS, the derived materials has notable shrinking BET surface areas and pore volumes (from 0.71 to 0.44 and 0.37 cm3 g−1 , respectively), implying the formation of “second” polymeric layer on the Fe3 O4 @POSS surface during the latter copolymerization of POSS and MAA, the pore “cavity and wall” may be blocked to some extent as a result. This case is similar to what has been observed in previous works reported by our group [32] and Hao et al. [45]. The Fe3 O4 @MI-POSS material has superior physical properties over the Fe3 O4 @NI-POSS material, such as larger surface area, pore volume and pore size. These maybe result from the steric effect owned by anchored template molecule during imprinting process. The extra space may be favor for better recognition toward the corresponding template molecule and its related analogs. And so as a consequence, the Fe3 O4 @MI-POSS material is expected to bring out better adsorption performance than the Fe3 O4 @MI-POSS material.

Field dependent magnetization measurements on the samples were conducted to study their magnetic behaviors. Fig. S2 shows the hysteresis loops of the magnetite particles recorded at room temperature. They exhibit neither coercivity nor remanence, indicating these nanocomposites are superparamagnetic. The magnetic saturation (Ms) values are 29.82, 26.06 and 20.42 emu g−1 for the Fe3 O4 @POSS, Fe3 O4 @MI-POSS and Fe3 O4 @NI-POSS, respectively. The saturation magnetization decrease in the magnetization value can be attributed to POSS coating of different thickness, but the materials still remained strong magnetic response and could be isolated within 30 s by use of an external magnet. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.chroma. 2014.07.089. 3.2. Static adsorption and selectivity evaluation The binding isotherms plotted in Fig. 5 indicated that the amount of ENR adsorbed increased with increasing of the initial concentration of ENR solution and then it reached to an equilibrium state. Meanwhile, the Fe3 O4 @MI-POSS exhibited significantly higher ENR loading amount than the Fe3 O4 @NI-POSS. For the NI-material, when the concentration of loading solution reached 1.0 mmol L−1 , the adsorption capacity was close to saturation. When the concentration of loading solution reached to 1.75 mmol L−1 , the maximum adsorption capacities of MImaterial and NI-material estimated from Fig. 5 were 0.187 and 0.104 mmol g−1 , respectively. The superior physical properties of MI-material (such as larger surface area and pore volume) as shown

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

29

Table 1 Porous properties of the prepared Fe3 O4 @POSS, Fe3 O4 @MI-POSS and Fe3 O4 @NI-POSS probed by nitrogen adsorption/desorption. Magnetic nanoparticles

BET surface area (m2 g−1 )

Pore volume (cm3 g−1 )

Average pore size (nm)

Fe3 O4 @POSS Fe3 O4 @MI-POSS Fe3 O4 @NI-POSS

625.96 357.06 294.75

0.71 0.44 0.37

4.70 4.56 3.92

Table 2 Selectivity of Fe3 O4 @MI-POSS.a Analytes

Qe (␮mol g−1 ) Fe3 O4 @MI-POSS

Fe3 O4 @NI-POSS

Fe3 O4 @MI-POSS

Fe3 O4 @NI-POSS

Fe3 O4 @MI-POSS

Fe3 O4 @NI-POSS

OFL DAN ENR CAP

6.24 6.48 7.89 1.80

2.32 3.91 4.28 1.77

150.26 196.02 189.09 38.78

55.99 118.29 102.53 38.12

3.87 5.05 4.88

1.47 3.10 2.69

a

−Qe =

(C0 −Ce )V M

; Kd =

(C0 −Ce )V Ce M

Kd (mL g−1 )

=

Qe Ce

; k=

Kd(FQ ) S Kd(CAP)

; k =

kFe O @MI-POSS 3 4 kFe O @NI-POSS 3 4

2.64 1.63 1.81

.

in Table 1 would be responsible for its larger adsorption capacity to ENR than that of NI-material. The molecular selectivity of the imprinted composite was confirmed by comparing the results of competitive assays. The structurally similar compounds OFL, DAN were selected to act as competitors and CAP as the reference compound. The distribution coefficient (Kd ), the selectivity coefficient (k), and the relative selectivity coefficient (k ) were obtained from these competitive experiments and are listed in Table 2. Kd indicates the adsorption ability of a substance. Obviously, the adsorption amounts of ENR and its analogs on the MI-material were higher than those on the NI-material. This may be related to the larger pore diameter of the MI-material for better site accessibility as discussed on Table 1. However, there was no obvious difference in the case of CAP, revealing the existence of special binding sites for template analogs on the Fe3 O4 @MI-POSS. The selectivity factor (k) is an indication of the selectivity of the sorbent over the related compounds. The much larger k values of Fe3 O4 @MIPOSS (3.87–5.05) than those of the Fe3 O4 @NI-POSS (1.47–3.10) suggest that the Fe3 O4 @MI-POSS has higher affinities for the template’s structurally related compounds over another typical antibiotic, the CAP, that encountered different recognition. A Comparison of k values for the MI-material with the corresponding control blank sample allows an estimation of the effect of imprinting on a particular FQ. The values of k are in the range of 1.81–2.64, which are higher than 1, further confirming that the 0.20 Fe3O4@MI-POSS Fe3O4@NI-POSS 0.15

a certain amount of ENR templates have been successfully incorporated into organic–inorganic networks by imprinting, and the imprinted cavities and specific binding sites in a predetermined orientation have formed after removal of template. The relative selectivity coefficient (k ) of OFL is higher than ENR may be contributed from the non-specific interactions relating to the silica cages that are connected through formed alkyl linkages between cages. 3.3. Optimization of MI-MSPE conditions In order to evaluate the applicability of the Fe3 O4 @MI-POSS material for separation and enrichment of FQs in milk samples, the parameters affecting the performance of the extraction including the amount of Fe3 O4 @MI-POSS, extraction time, sample pH and elution solvents were investigated. The extraction conditions were optimized by analyzing spiked milk samples at a concentration of 1000 ng mL−1 . When one parameter was changed, the other parameters were fixed at their optimized values. 3.3.1. The amount of Fe3 O4 @MI-POSS Different amounts of Fe3 O4 @MI-POSS sorbent ranging from 30 to 80 mg in 10 mL of PBS solution containing known amount of FQs sample were applied to extract the FQs from spiked 0.5 mL of milk sample (Fig. 6a). The minimum amount of sorbent required to get efficient recovery was then investigated. The results indicated that 60 mg of Fe3 O4 @MI-POSS was enough, and satisfactory recoveries ranging from 79.5 to 94.5% were obtained. Further increasing the amount of Fe3 O4 @MI-POSS sorbent gave no improvement for recoveries. Therefore, the amount of sorbent used was fixed at 60 mg. 3.3.2. Extraction time The effect of the extraction time from 10 to 120 min on the recoveries of FQs was investigated (Fig. 6b). An increase in recoveries from 67.2–90.4% to 80.7–100.8% was observed for the selected FQs with increasing contact time up to 30 min. Further increase in contact time does not result in a significant increment in recoveries but leads to a plateau. Therefore, in this work, the extraction time of 30 min was sufficient to achieve satisfactory recovery.

-1

Q (mmol g )

k

k

0.10

0.05

0.00 0.0

0.5

1.0

1.5

2.0

-1

Initial concentration (mmol L ) Fig. 5. Curves of adsorption isotherm of ENR onto Fe3 O4 @MI-POSS and Fe3 O4 @NIPOSS.

3.3.3. pH of sample In view of the pKa values of the three FQs studies here ranging from 6.07 to 6.81 for the carboxylic functions and from 8.04 to 8.56 for the ammonium form [38,46,47], the extraction behavior may be pH-dependent. The influence of the sample pH was conducted

30

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

100 100

(c)

OFL DAN ENR

(a) 80

60

Recoveries (%)

Recoveries (%)

80

OFL DAN ENR

40

60

40

20

20

0 30

40

50

60

70

0

80

3

4

5

Fe3O4@POSS (mg)

100

(d)

(b)

7

OFL DAN ENR

100

80

80

OFL DAN ENR

60

Recoveries (%)

Recoveries (%)

6

pH

40

20

60

40

20

0

20

40

60

80

100

120

0

ACN

,v/v) MeOH v/v/v) Ac(95:5 (90:9:1, MeOH-H Ac-TFA MeOH-H

Extraction time (min)

Desorption solution Fig. 6. The effects of (a) Fe3 O4 @MI-POSS amount, (b) extraction time, (c) pH of sample and (d) elution solvents on the recoveries of FQs (n = 3).

interactions between the sorbent and the template molecules could be suppressed effectively in the presence of organic acid. Better recoveries (80.9–94.3%) were obtained by using 2.0 mL of MeOH–HAc–TFA (90:9:1, v/v/v) as desorption solvent than that 1000 1

2

3

800 (c)

Intensity(µV)

in phosphate matrix solution (25 mmol L−1 ) over the pH 3.0–7.0 range, and the extraction efficiency for the FQs was illustrated in Fig. 6c. As it was shown in Fig. 6c, the extraction efficiencies for the FQs reached maximums at pH 6.0. It might be ascribed that the carboxyl groups (pKa = 5.5) of Fe3 O4 @MI-POSS are ionized and present negative charges at pH 6.0, while the FQs exist in cationic forms from pH 4.0 to 6.0, leading to strong ion-exchange interactions and thus higher extraction efficiencies. The lower extraction efficiencies at lower pH (pH ≤5) should be ascribed to the weakened electrostatic interactions between the protonated analytes and the undissociated extraction phase. When the pH is above 6.0, the ionization degrees of the analytes are decreased and FQs are converted into the intermediate form accordingly, resulting in decreased ion-exchange interactions and thus lower extraction efficiency. It should be noted that hydrogen bonding and hydrophobic interaction may also contribute to the molecular recognition thus obtain good extraction efficiencies [36,47,48]. Considering the extraction efficiency, sample solution with pH 6.0 was applied to extract the FQs in real samples.

600

400

(b)

200 (a) 0

3.3.4. Elution solvent In order to obtain the highest recoveries of FQs, a series of solvents, ACN, MeOH, MeOH–HAc (95:5, v/v) and MeOH–HAc–TFA (90:9:1, v/v/v) were used to optimize the eluting condition (Fig. 6d). The poor recoveries (22.1–71.6%) were found by using ACN and MeOH. As shown in Fig. 7d, the addition of organic acid obviously helped to increase the recoveries. This was probably due that the hydrogen bond formation as well as the hydrophobic

0

5

10

15

20

25

Time (min) Fig. 7. HPLC-UV chromatograms of (a) milk sample spiked FQs at concentration of 200 ng mL−1 before MI-MSPE (injection sample matrix: PBS solution), (b) milk sample spiked FQs at concentration of 200 ng mL−1 after MI-MSPE (injection sample matrix: desorption solution) and (c) standard mixture solution at concentration of 50 ng mL−1 (Injection sample matrix: PBS solution). Peaks: (1) OFL; (2) DAN; (3) ENR. Chromatographic and MI-MSPE conditions were outlined in Sections 2.3 and 2.5.

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

31

Table 3 Relative recoveries (n = 3) and RSD (%) of three FQs at different spiking level in milk. Spiking level (ng mL−1 )

OFL

1000 100 50

DAN

ENR

Relative recovery (%)

RSD (%)

Relative recovery (%)

RSD (%)

Relative recovery (%)

RSD (%)

101.5 43.2 30.7

7.8 3.6 2.1

96.3 83.6 52.9

1.6 7.3 3.5

114.5 90.7 103.8

2.7 14.9 2.7

Table 4 Analytical parameters for the proposed MI-SPE method of determining three FQs in milk samples. FQs

Linearity range (ng mL−1 )

Calibration curves

LODa (ng mL−1 )

LOQb (ng mL−1 )

r

OFL DAN ENR

50–1000 50–1000 50–1000

y = 3.055x + 340.432 y = 5.079x − 27.733 y = 5.493x + 35.664

1.76 12.42 9.20

5.84 41.38 30.64

0.9979 0.9969 0.9989

a b

Detection limits (LODs) were calculated as three times the signal-to-noise ratio (S/N = 3). Quantification limits (LOQs) were calculated as ten times the signal-to-noise ratio (S/N = 10).

Table 5 The intra- and inter-assay recoveries and precisions obtained by analyzing of three FQs in milk sample using the proposed MI-MSPE method. FQs

Intra-assay (n = 5) 50 ng mL−1

OFL DAN ENR

Inter-assay (n = 3) 100 ng mL−1

1000 ng mL−1

50 ng mL−1

100 ng mL−1

1000 ng mL−1

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

Recovery (%)

RSD (%)

96.4 77.2 93.3

5.2 4.2 5.0

100.4 85.9 95.1

7.8 5.2 5.4

92.9 79.9 108.8

8.8 4.3 2.9

85.9 75.6 90.8

4.8 11.5 3.8

87.4 103.1 86.2

3.8 5.5 10.1

93.6 75.9 86.5

7.9 3.5 3.7

(73.6–93.7%) obtained by using MeOH-HAc (95:5, v/v). Thus, 2.0 mL of MeOH–HAc–TFA (90:9:1, v/v/v) with sonication for 10 min was chosen for the desorption stage.

3.4. Analytical performance The typical chromatograms of milk samples spiked with FQs obtained by direct injection and injection after MI-MSPE are displayed in Fig. 7a and b, respectively. By comparison, the matrix has been cleaned up and there is no interfering peaks when the samples were treated by the proposed Fe3 O4 @MI-POSS material, demonstrating that the selectivity of the MI-MSPE method based on the Fe3 O4 @MI-POSS material is good in some extent. Furthermore, Fig. 7b displayed better peak shapes and obvious enhancement in all FQs peak heights in comparison with the chromatogram of direct injection (Fig. 7a). None of the studied FQs were detected in the blank chromatogram. Therefore, the MI-MSPE-HPLC method based on the Fe3 O4 @MI-POSS material was evaluated by analysis of FQs spiked blank milk samples in regard to the linearity, limit of detection (LOD) and limit of quantification (LOQ), accuracy and precision. The relative recoveries were calculated by comparing the peak area ratios of FQs extracts from the spiked milk samples to those obtained from the working standard solutions at the same concentration for evaluation the matrix effect. As shown in Table 3, the relative recoveries of three FQs range from 30.7 to 114.5% in milk matrices with acceptable RSDs in the range of 1.6 to 14.9%. The results revealed that the determination of FQs was affected by the interferences from real samples to some extent. On the one hand, protein and fat from the samples would reduce the interaction between analytes and the extraction material. On the other hand, the coextracted components from complex matrices would suppress the signals of analytes. Therefore, the matrix-matched calibration curves were chosen as reference curves throughout this study to provide reliable quantification. Under the optimized conditions described in the Section 3.3, the calibration curves were built by plotting the peak area of analytes

extracted from milk versus the spiked concentration. As listed in Table 4, the correlation coefficients (r) ranging from 0.9969 to 0.9989 are obtained for all the analytes in the linear concentration range of 50–1000 ng mL−1 . LODs and LOQs were calculated as the concentration corresponding to signal 3 and 10 times the standard deviation of the baseline noise, respectively. The LODs for three FQs in milk were found to be 1.8–12.4 ng mL−1 and the LOQs were found to be 5.8–41.4 ng mL−1 . In this study, the accuracy of the method was measured and expressed as recovery. The precision of the method was accessed by determining intra-day (five times in one day) and inter-day (over three days) RSDs at three different spiked levels. As summarized in Table 5, the recoveries for analytes tested are in the range of 75.6–108.8%, the RSDs of intra- and inter-day tests ranging from 2.9 to 8.8% and from 3.5 to 11.5%, respectively. The results demonstrate that the precision and accuracy of the present MI-MSPE-HPLC method are acceptable for routine monitoring purposes. 3.5. Application of the MI-MSPE-HPLC to milk samples To demonstrate the feasibility of the established MI-MSPE-HPLC system, five pure milk samples collected from local different markets (near to Shanghai University) were analyzed. No FQs residues at detectable levels were found in these samples. The recovery studies were then carried out by spiking the milk samples with FQs at the level of 50–1000 ng mL−1 . The recoveries of FQs from 78.4 to 102.7% with RSDs less than 9.6% were obtained. The results indicated that the proposed method based on the Fe3 O4 @MI-POSS was applicable for the extraction and determination of FQs in milk samples. 4. Conclusion In this study, a novel route for fabrication of magnetic molecular imprinted hybrid sorbents toward the sample pretreatment for antibiotics residue analysis in complex matrices has been developed. The ENR imprinted material based on nanomagnetic

32

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33

POSS was facilely prepared via surface vinyl grafted polymerization technique as an example. The resulting material behaved proper magnetism and showed good selectivity and high adsorption capacity to FQs. By employing it as a SPE material, an efficient and convenient MI-MSPE-HPLC method was established for the separation and enrichment of three FQs from milk samples, which provided the low-cost and facile tools for monitoring of FQs residues in food safety analysis. More importantly, the proposed imprinting technique at the surface of Fe3 O4 @POSS by utilizing the rich chemistry of POSS should be generally applicable to imprinting various organic molecules, thus extensive applications involving Fe3 O4 @POSS derived molecular imprinted materials are foreseeable. Acknowledgments This work has been supported by funds from the National Science Foundation of China (No. 20905046 and No. 21005057) and the Key Laboratory of Analytical Chemistry for Biology and Medicine of Ministry of Education in Wuhan University (ACBM2014002). The authors gratefully thank the Instrumental Analysis & Research Center of Shanghai University for instrumentation. References [1] L. Chen, S. Xu, J. Li, Recent advances in molecular imprinting technology: current status, challenges and highlighted applications, Chem. Soc. Rev. 40 (2011) 2922–2942. [2] A. Poma, A.P. Turner, S.A. Piletsky, Advances in the manufacture of MIP nanoparticles, Trends Biotechnol. 28 (2010) 629–637. [3] Y.L. Hu, J.L. Pan, K.G. Zhang, H.X. Lian, G.K. Li, Novel applications of molecularlyimprinted polymers in sample preparation, TrAC Trends Anal. Chem. 43 (2013) 37–52. [4] E. Turiel, A. Martín-Esteban, Molecularly imprinted polymers for sample preparation: a review, Anal. Chim. Acta 668 (2010) 87–99. [5] A. Martín-Esteban, Molecularly-imprinted polymers as a versatile, highly selective tool in sample preparation, TrAC Trends Anal. Chem. 45 (2013) 169–181. [6] M. Trojanowicz, Enantioselective electrochemical sensors and biosensors: a mini-review, Electrochem. Commun. 38 (2014) 47–52. [7] J.L. Zhang, M.X. Zhang, K.J. Tang, F. Verpoort, T.L. Sun, Polymer-based stimuliresponsive recyclable catalytic systems for organic synthesis, Small 10 (2014) 32–46. [8] M. Lasáková, P. Jandera, Molecularly imprinted polymers and their application in solid phase extraction, J. Sep. Sci. 32 (2009) 799–812. [9] X.S. Li, G.T. Zhu, Y.B. Luo, B.F. Yuan, Y.Q. Feng, Synthesis and applications of functionalized magnetic materials in sample preparation, TrAC Trends Anal. Chem. 45 (2013) 233–247. [10] L.G. Chen, B. Li, Application of magnetic molecularly imprinted polymers in analytical chemistry, Anal. Methods 4 (2012) 2613–2621. [11] M. Zhao, C. Zhang, Y. Zhang, X.Z. Guo, H.S. Yan, H.Q. Zhang, Efficient synthesis of narrowly dispersed hydrophilic and magnetic molecularly imprinted polymer microspheres with excellent molecular recognition ability in a real biological sample, Chem. Commun. 50 (2014) 2208–2210. [12] J.L. Cao, X.H. Zhang, X.W. He, L.X. Chen, Y.K. Zhang, The synthesis of magnetic lysozyme-imprinted polymers by means of distillation–precipitation polymerization for selective protein enrichment, Chem. Asian J. 9 (2013) 526–533. [13] M. Zhang, Y.Z. Wang, X.P. Jia, M.Z. He, M.L. Xu, S. Yang, C.J. Zhang, The preparation of magnetic molecularly imprinted nanoparticles for the recognition of bovine hemoglobin, Talanta 120 (2014) 376–385. [14] M.J. Ding, X.L. Wu, L.H. Yuan, S. Wang, Y. Li, R.Y. Wang, T.T. Wen, S.H. Du, X.M. Zhou, Synthesis of core–shell magnetic molecularly imprinted polymers and detection of sildenafil and vardenafil in herbal dietary supplements, J. Hazard. Mater. 191 (2011) 177–183. [15] S. Azodi-Deilami, M. Abdouss, D. Kordestani, Synthesis and characterization of the magnetic molecularly imprinted polymer nanoparticles using N, Nbis methacryloyl ethylenediamine as a new cross-linking agent for controlled release of meloxicam, Appl. Biochem. Biotechnol. 172 (2014) 3271–3286. [16] Q.P. You, M.J. Peng, Y.P. Zhang, J.F. Guo, S.Y. Shi, Preparation of magnetic dummy molecularly imprinted polymers for selective extraction and analysis of salicylic acid in actinidia chinensis, Anal. Bioanal. Chem. 406 (2014) 831–839. [17] Y. Li, M.J. Ding, S. Wang, R.Y. Wang, X.L. Wu, T.T. Wen, L.H. Yuan, P. Dai, Y.H. Lin, X.M. Zhou, Preparation of imprinted polymers at surface of magnetic nanoparticles for the selective extraction of tadalafil from medicines, ACS Appl. Mater. Interfaces 3 (2011) 3308–3315. [18] F.F. Chen, X.Y. Xie, Y.P. Shi, Magnetic molecularly imprinted polymer for the selective extraction of sildenafil, vardenafil and their analogs from herbal medicines, Talanta 115 (2013) 482–489.

[19] G.F. Ma, L.G. Chen, Determination of chlorpyrifos in rice based on magnetic molecularly imprinted polymers coupled with high-performance liquid chromatography, Food Anal. Methods 7 (2013) 377–388. [20] H.Z. Lan, N. Gan, D.D. Pan, F.T. Hu, T.H. Li, N.B. Long, L. Qiao, An automated solid-phase microextraction method based onmagnetic molecularly imprinted polymer as fiber coating for detection of trace estrogens in milk powder, J. Chromatogr. A 1331 (2014) 10–18. [21] A. Mehdinia, T. Baradaran Kayyal, A. Jabbari, M.O. Aziz-Zanjani, E. Ziaei, Magnetic molecularly imprinted nanoparticles based on grafting polymerization for selective detection of 4-nitrophenol in aqueous samples, J. Chromatogr. A 1283 (2013) 82–88. [22] F.F. Chen, X.Y. Xie, Y.P. Shi, Preparation of magnetic molecularly imprinted polymer for selective recognition of resveratrol in wine, J. Chromatogr. A 1300 (2013) 112–118. [23] Y.K. Lv, Y. Ma, X.B. Zhao, C.L. Jia, H.W. Sun, Grafting of norfloxacin imprinted polymeric membranes on silica surface for the selective solid-phase extraction of fluoroquinolones in fish samples, Talanta 89 (2012) 270–275. [24] J.E. Mark, Some interesting things about polysiloxanes, Acc. Chem. Res. 37 (2004) 946–953. [25] R.M. Laine, Nanobuilding blocks based on the [OSiO1.5 ]x (x = 5 6, 8, 10) octasilsesquioxanes, J. Mater. Chem. 15 (2005) 3725–3744. [26] K. Tanaka, Y. Chujo, Advanced functional materials based on polyhedral oligomeric silsesquioxane (POSS), J. Mater. Chem. 22 (2012) 1733–1746. [27] I. Nischang, O. Brüggemann, I. Teasdale, Facile, Single-step preparation of versatile, high-surface-area, hierarchically structured hybrid materials, Angew. Chem. Int. Ed. 50 (2011) 4592–4596. [28] F. Alves, P. Scholder, I. Nischang, Conceptual design of large surface area Porous polymeric hybrid media based on polyhedral oligomeric silsesquioxane precursors: preparation, tailoring of porous properties, and internal surface functionalization, ACS Appl. Mater. Interfaces 5 (2013) 2517–2526. [29] M.H. Wu, R.A. Wu, R.B. Li, H.Q. Qin, J. Dong, Z.B. Zhang, H.F. Zou, Polyhedral oligomeric silsesquioxane as a cross-linker for preparation of inorganic–organic hybrid monolithic columns, Anal. Chem. 82 (2010) 5447–5454. [30] J.J. Ou, Z.B. Zhang, H. Lin, J. Dong, M.H. Wu, H.F. Zou, Preparation and application of hydrophobic hybrid monolithic columns containing polyhedral oligomeric silsesquioxanes for capillary electrochromatography, Electrophoresis 33 (2012) 1660–1668. [31] H. Lin, J.J. Ou, Z.B. Zhang, J. Dong, H.F. Zou, Ring-opening polymerization reaction of polyhedral oligomeric silsesquioxanes (POSSs) for preparation of well-controlled 3D skeletal hybrid monoliths, Chem. Commun. 49 (2013) 231–233. [32] H.B. He, B. Li, J.P. Dong, Y.Y. Lei, T.L. Wang, Q.W. Yu, Y.Q. Feng, Y.B. Sun, Mesostructured nanomagnetic polyhedral oligomeric silsesquioxanes (POSS) incorporated with dithiol organic anchors for multiple pollutants capturing in wastewater, ACS Appl. Mater. Interfaces 5 (2013) 8058–8066. [33] P. Regal, M. Díaz-Bao, R. Barreiro, A. Cepeda, C. Fente, Application of molecularly imprinted polymers in food analysis: clean-up and chromatographic improvements, Cent. Eur. J. Chem. 10 (2012) 766–784. [34] H. Wang, R. Wang, Y. Han, Preparation of molecular imprinted microspheres based on inorganic–organic co-functional monomer for miniaturized solidphase extraction of fluoroquinolones in milk, J. Chromatogr. B 949-950 (2014) 24–29. [35] H.B. Zheng, J.Z. Mo, Y. Zhang, Q. Gao, J. Ding, Q.W. Yu, Y.Q. Feng, Facile synthesis of magnetic molecularly imprinted polymers and its application in magnetic solid phase extraction for fluoroquinolones in milk samples, J. Chromatogr. A 1329 (2014) 17–23. [36] M.M. Zheng, R. Gong, X. Zhao, Y.Q. Feng, Selective sample pretreatment by molecularly imprinted polymer monolith for the analysis of fluoroquinolones from milk samples, J. Chromatogr. A 1217 (2010) 2075–2081. [37] J. Blesa, L.J.G. Silva, C.M. Lino, G. Font, A. Pena, Comparison of three solid-phase extraction processes in quantification of ciprofloxacin and enrofloxacin in pork meat, J. Sep. Sci. 35 (2012) 832–838. [38] J.L. Urraca, M. Castellari, C.A. Barrios, M.C. Moreno-Bondi, Multiresidue analysis of fluoroquinolone antimicrobials in chicken meat by molecularly imprinted solid-phase extraction and high performance liquid chromatography, J. Chromatogr. A 1343 (2014) 1–9. [39] C. Blasco, Y. Pico´, Development of an improved method for trace analysis of quinolones in eggs of laying hens and wildlife species using molecularly imprinted polymers, J. Agric. Food Chem. 60 (2012) 11005–11014. [40] S.B. Turnipseed, S.B. Clark, J.M. Storey, Analysis of veterinary drug residues in frog legs and other aquacultured species using liquid chromatography quadrupole time-of-flight mass spectrometry, J. Agric. Food Chem. 60 (2012) 4430–4439. [41] A. Speltini, M. Sturini, F. Maraschi, A. Profumo, A. Albini, Analytical methods for the determination of fluoroquinolones in solid environmental matrices, TrAC Trends Anal. Chem. 30 (2011) 1337–1350. [42] L.G. Chen, X.P. Zhang, L. Sun, Y. Xu, Q.L. Zeng, H. Wang, H.Y. Xu, A.M. Yu, H.Q. Zhang, L. Ding, Fast and selective extraction of sulfonamides from honey based on magnetic molecularly imprinted polymer, J. Agric. Food Chem. 57 (2009) 10073–10080. [43] D.M. Han, G.Z. Fang, X.P. Yan, Preparation and evaluation of a molecularly imprinted sol–gel material for on-line solid-phase extraction coupled with high performance liquid chromatography for the determination of trace pentachlorophenol in water samples, J. Chromatogr. A 1100 (2005) 131–136.

H.-B. He et al. / J. Chromatogr. A 1361 (2014) 23–33 [44] H.Y. Yan, M.L. Tian, K.H. Row, Determination of enrofloxacin and ciprofloxacin in milk using molecularly imprinted solid-phase extraction, J. Sep. Sci. 31 (2008) 3015–3020. [45] Y.J. Hao, Y.Z. Chong, S.R. Li, H.Q. Yang, Controlled synthesis of Au nanoparticles in the nanocages of SBA-16: improved activity and enhanced recyclability for the oxidative esterification of alcohols, J. Phys. Chem. C 116 (2012) 6512–6519. [46] H.B. He, X.X. Lv, Q.W. Yu, Y.Q. Feng, Multiresidue determination of (fluoro)quinolone antibiotics in chicken by polymer monolith microextraction and field-amplified sample stacking procedures coupled to CE-UV, Talanta 82 (2010) 1562–1570.

33

[47] D.L. Xiao, P. Dramou, N.Q. Xiong, H. He, D.H. Yuan, H. Dai, H. Li, X.M. He, J. Peng, N. Li, Preparation of molecularly imprinted polymers on the surface of magnetic carbon nanotubes with a pseudo template for rapid simultaneous extraction of four fluoroquinolones in egg samples, Analyst 138 (2013) 3287–3296. [48] M.M. Zheng, G.D. Ruan, Y.Q. Feng, Evaluating polymer monolith in-tube solid-phase microextraction coupled to liquid chromatography/quadrupole time-of-flight mass spectrometry for reliable quantification and confirmation of quinolone antibacterials in edible animal food, J. Chromatogr. A 1216 (2009) 7510–7519.

Fabrication of enrofloxacin imprinted organic-inorganic hybrid mesoporous sorbent from nanomagnetic polyhedral oligomeric silsesquioxanes for the selective extraction of fluoroquinolones in milk samples.

This paper reports a nanomagnetic polyhedral oligomeric silsesquioxanes (POSS)-directing strategy toward construction of molecularly imprinted hybrid ...
3MB Sizes 1 Downloads 6 Views