Highly selective amino-functionalized magnetic molecularly imprinted polymers: absorbents for dispersive solid phase extraction and trace level analysis of chlorophenols in seawater

Abstract

A method of dispersive magnetic solid phase extraction (dMSPE) using magnetic molecularly imprinted polymer (nFe₃O₄@NH₂MIP) as an adsorbent coupled with high performance liquid chromatography (HPLC) was developed for the determination of 5 kinds of chlorophenols (CPs) in seawater samples. 8 kinds of nFe₃O₄@NH₂MIP were synthesized via suspension polymerization by using the 5 different kinds of CPs as templates and 4 different kinds of amines for surface grafting. Compared with seven other kinds of MIP materials and the non-imprinted magnetic polymers (nFe₃O₄@NH₂NIPs), the as-prepared nFe₃O₄@TEPA–PCP-MIP showed excellent extraction performance for CPs with high enrichment efficiencies, short times, high recoveries and reusabilities. Various parameters affecting the extraction efficiency of the dMSPE of the 5 target CPs were optimized, including solution pH, extraction time, desorption time, desorption solution, volume of desorption solution, kinds of absorbents, etc. The results show that under the optimal condition, linearities are obtained ranging from 1 ng L⁻¹ to 5000 ng L⁻¹ with correlation coefficients (R) higher than 0.9989 for the target CPs. The recoveries are between 86.5% and 98.8% at the spiked levels with the relative standard deviations (RSDs) in the range of 0.8–8.8%. The limits of detection (LODs) are in the range of 0.18–1.2 ng L⁻¹ and the limits of quantification (LOQs) are between 0.6 ng L⁻¹ and 4.0 ng L⁻¹. The developed method is successfully applied for real sample analyses, which confirm that it can be applied to routine monitoring for the determination of the CPs in seawater samples.

---

1. Introduction

Chlorophenols (CPs) are widely used pesticides, disinfectants, wood preservatives and pulp bleaching agents,¹ resulting in the release of CPs into the environment. Chlorine treatment in drinking water systems can also produce CPs.^2,3 Because of their high toxicity, CPs have been included in the list of priority pollutants by the US Environmental Protection Agency (EPA).⁴ The US EPA has set a maximum residue limit (MRL) of 1 μg L⁻¹ for pentachlorophenol (PCP) in drinking water, while European Union legislation has set a maximum admissible concentration (MAC) of PCP in inland and other surface waters to be 1 μg L⁻¹, as well.⁵ Thus, a rapid, sensitive, and accurate analytic method would be necessary for the determination of CPs in environmental samples. With the awareness of the importance of marine environmental protection, monitoring the trace residual concentration of CPs in seawater has become one of the important issues.^6,7 Over the past decades, gas chromatography (GC) with flame ionization detection (FID),⁸ electron-capture detection (ECD)⁹ and mass spectroscopy (MS)¹⁰ are usually used for the analysis of CPs. However, due to the high polarity of CPs, derivatization is often needed for sample preparation, which is tedious and time-consuming. High-performance liquid chromatography (HPLC) is one of the mainly used methods for the separation and determination of CPs by many researchers.^11–17

As the concentrations of CPs in environmental samples are usually quite low and sample matrix are complex, preconcentration and clean-up of CPs are usually necessary prior to the instrumental analysis. Some sample pretreatment methods have been reported, including liquid-phase microextraction (LPME),¹⁸ dispersive liquid–liquid microextraction (DLLME),¹⁹ liquid–liquid–liquid micro-extraction (LLLME),²⁰ and solid-phase extraction (SPE).²¹ However, among these, reports on monitoring CPs in seawater are still limited.²⁰ Solid-phase micro-extraction (SPME), pioneered by Arthur and Pawliszyn in 1990s,²² has been widely applied for monitoring CPs in environment matrices. With the development of magnetic absorbents, the magnetic solid-phase extraction (MSPE) is mushrooming recently and proved to be a quick, easy, cheap, effective, rugged, and safe (QuEChERS) procedure for extraction and preconcentration of CPs,²³ as well as other pollutants.^24–26

Recently, new separation methods based on the use of magnetic molecularly imprinted nano-materials have found to be simple, convenient, and powerful approaches for the separation and purification of environmental samples, and removal of toxic pollutants in water.^27–29 It would be desirable if three of the promising concepts (molecular imprinting, magnetic separation and solid-phase extraction) were combined. Herein, in this work, we report the design and successful synthesis of a serial of core–shell microspheres, consisting of a Fe₃O₄ nanoparticle core and an outer layer of amino-functionalized MIP shell (nFe₃O₄@NH₂MIP). Since MIP is normally highly selective to the target template, which cannot achieve the request of the multi-residue of CPs determination, different kinds of CPs were used as template and four kinds of amines, i.e., ethylenediamine (EDA), diethylenetriamine (DETA), triethylenetetramine (TETA) and tetraethylenepentamine (TEPA), were used for surface grafting. The tetraethylenepentamine (TEPA) functionalized with PCP as template core–shell molecularly imprinted magnetic polymer (nFe₃O₄@TEPA–PCP-MIP) was proved to have great adsorptive ability toward five chlorophenols (CPs), i.e., 2-chlorophenol (2-CP), 2,4-dichlorophenol (2,4-DCP), 2,4,6-trichlorophenol (2,4,6-TCP), 2,3,4,6-tetrachlorophenol (2,3,4,6-TeCP) and PCP. The nFe₃O₄@TEPAMIP was used as the three-in-one dMSPE for preconcentration of five trace-level CPs in seawater, prior to HPLC analysis. The developed analytical procedure is proven to be effective, fast, and accurate for routine analyses.

2. Experimental

2.1 Chemicals

Ferric chloride (FeCl₃·6H₂O), ferrous sulphate (FeSO₄·7H₂O), oleic acid, polyglycol and ammonium acetate (NH₄Ac) were analytical grade, and purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Styrene (St), divinylbenzene (DVB) and glycidylmethacrylate (GMA) were supplied by Alddin Chemical Reagent Co., Ltd. (Shanghai, China) and purified by vacuum distillation. Benzoyl peroxide (BPO) was purchased from J&K Chemical (99%) and used as an initiator without further purification. EDA, DETA, TETA, TEPA, 2-CP, 2,4-DCP, 2,4,6-TCP, 2,3,4,6-TeCP and PCP were supplied by Aladdin Chemical Reagent Co., Ltd. (Shanghai, China). LC grade of methanol and acetonitrile were obtained from Merck (Darmstadt, Germany).

2.2 Synthesis of nFe₃O₄@TEPA–PCP-MIP

nFe₃O₄@TEPA–PCP-MIP was prepared according to a reported procedure²⁹ after minor modification with PCP as the template. Overall preparation procedure was given in Scheme 1. Briefly, Fe₃O₄ particles were firstly encapsulated by oleic acids to obtain OA-M (oleic acids encapsulated magnetic Fe₃O₄); then OA-M was modified by suspension polymerization with 4 mL (0.04 mol) St, 8 mL (0.05 mol) GMA and 0.5 mL (0.004 mol) DVB, yielding epoxyl-Fe₃O₄-co-poly(DVB-St-GMA)s, named as nFe₃O₄@eO-polymer, which was isolated under magnetic field and washed with water, dispersed in ethanol (Scheme 1(a)).

Scheme 1 Schematic procedure of preparation of nFe₃O₄@TEPA–PCP-MIP.

Then the PCP template molecule solution was prepared by mixing 2.0 mmol L⁻¹ PCP and 200 mmol L⁻¹ TEPA in 50 mL of ethanol and self-assembled via hydrogen-bonding interactions by stirring at 80 °C for 2 h (Scheme 1(b)).

Afterward, the above two solutions were mixed and stirred at 500 rpm at 80 °C under ultrasound. The PCP template molecules were grafted onto the material via ring-opening reaction (Scheme 1(c)). Finally, the template molecules were cleaned with acetic acid/methanol (1 : 1, v/v) for several times under ultrasound until PCP could not be detected by HPLC. The as-prepared nFe₃O₄@TEPA–PCP-MIP were washed with water three times, dried in a vacuum oven at 60 °C and stored in a sealed bottle for further use (Scheme 1(d)).

Other seven kinds of absorbents with different kinds of amines, i.e., nFe₃O₄@EDA–PCP-MIP, nFe₃O₄@DETA–PCP-MIP, and nFe₃O₄@TETA–PCP-MIP and with different kinds of template molecules, i.e., nFe₃O₄@TEPA–2-CP-MIP, nFe₃O₄@TEPA–2,4-DCP-MIP, nFe₃O₄@TEPA–2,4,6-TCP-MIP and nFe₃O₄@TEPA–2,3,4,6-TeCP-MIP were synthesized, respectively.

For comparing, non-imprinted amino-functionalized magnetic polymers, i.e., nFe₃O₄@TEPA-NIP, nFe₃O₄@TETA-NIP, nFe₃O₄@DETA-NIP, and nFe₃O₄@EDA-NIP were synthesized by almost the same procedures described above without the addition of the templates.

2.3 Instruments and HPLC analysis

The morphology and dimensions of as-prepared nFe₃O₄@NH₂MIP were obtained on a Hitachi H-7650 transmission electron microscopy (TEM) (Hitachi, Japan) at an accelerating voltage of 75 kV. Scanning electron microscopy (SEM) was performed using scanning electron microscopy (SEM, JSM-6700F) at an accelerating voltage of 5.0 kV. Sample dispersed at an appropriate concentration in ethanol was cast onto a silicon sheet at room temperature and sputter-coated with gold. The magnetic properties of magnetic particles were measured using a vibrating sample magnetometer (VSM, Lake Shore 7410). Powder X-ray diffraction (XRD) patterns were collected on an X-ray diffractometer (Bruker D8 Advance) with CuKα radiation at λ = 0.154 nm operating at 40 kV and 40 mA. The content of Fe₃O₄ in each of nFe₃O₄@NH₂MIP was calculated from the amount of leached Fe, which was measured by using a spectrophotometer (722, Shanghai, China) according to the standard colorimetric method³⁰ after digesting nFe₃O₄@NH₂MIP in 12 mol L⁻¹ HCl solution. The elementary analysis results of the nitrogen contents in nFe₃O₄@NH₂MIP were measured using an elementary analysis (EA, Thermo Fisher Flash-1112). Fourier Transform Infrared spectrometer (FTIR, Thermo Nicolet, USA) were applied for characterization.

HPLC analysis was performed on an Elite HPLC system including a binary pump and a UV detector (Elite Corporation, PRC), using a ZORBAX SB-C8 (5 μm particle size, 250 mm × 4.6 mm) analytical column. The mobile phase was using a 70 : 30 methanol and 5 mmol L⁻¹ NH₄Ac aqueous solution (v/v), at a flow rate of 1.0 mL min⁻¹. The analytes were detected by DAD at 230 nm. Column was maintained at a temperature of 35 °C to enhance the retention time reproducibility, and the injection volume was 20.0 μL.

2.4 Sample preparation

Seawater samples were collected in glass bottles. After the essential filtration treatment, the collected samples were stored in dark at 4 °C. Those free from the target analytes were used as blanks and for spiked sample preparation.

Further dMSPE procedure was shown in Scheme 2: in a 1000 mL flask, 500 mL of sample solution and 20.0 mg of nFe₃O₄@TEPA–PCP-MIP was added. After the mixture being shaken for 10 min, a magnet was placed close to the bottom of flask to collect the magnetic adsorbent for several seconds, and the supernatant was discarded. Then, 1.0 mL of desorption solution (methanol containing 2 mmol L⁻¹ NaOH) was added into the flask to elute CPs by shaking for 4 min. Similarly, desorption solution was also separated under a magnet, then the desorption solution was adjusted to neutral by hydrochloric acid and transferred to a 1.5 mL centrifuge tube and evaporated to dryness under a mild nitrogen stream at room temperature. Finally, the residue was dissolved with 0.5 mL methanol, filtrated via 0.2 μm filter membrane, and analyzed by HPLC.

Scheme 2 Schematic representation of dMSPE process and analysis.

3. Results and discussion

3.1 Characterization of nFe₃O₄@NH₂MIP

The TEM and SEM images of nFe₃O₄@TEPA–PCP-MIP were shown in Fig. S1 (ESI†). It revealed that the nFe₃O₄@TEPA–PCP-MIP microspheres had a multidispersed spherical shape with a rough surface, which could help to adsorption, with an average diameter of around 200 nm. Fig. S2a† showed the XRD analysis of nFe₃O₄@TEPA–PCP-MIP as well as the bare Fe₃O₄. It indicated that nFe₃O₄@TEPA–PCP-MIP had retained the spinel structure of Fe₃O₄, in which the identical peaks for Fe₃O₄ located at 30.1°, 35.5°, 43.1°, 53.4°, 57.0° and 62.6°, corresponding to their indices (220), (311), (400), (422), (511) and (400)³¹ appeared. The paramagnetic property of the nFe₃O₄@TEPA–PCP-MIP was verified by the magnetization curve measured by VSM, shown in Fig. S2b.† Its saturation moment obtained from the hysteresis loop was found to be 32.6 emu g⁻¹. The nFe₃O₄@TEPA–PCP-MIP was expected to respond well to magnetic field without any permanent magnetization, therefore making the solid and liquid phases separate easily. For comparing, the saturation moments, Fe₃O₄ contents and nitrogen contents of the nFe₃O₄@NH₂MIP with different surface grafting amines and molecular printing templates were also investigated. Results were shown in Table 1. The Fe₃O₄ content of the nFe₃O₄@NH₂MIP decreased from 40.4% to 35.4%, while the nitrogen content of which increased from 4.83 to 7.88 mol g⁻¹ for the amino group varying from EDA to TEPA with the growth of the amino chain, thus leading the variation of the saturation moments decreasing from 40.4 emu g⁻¹ to 32.6 emu g⁻¹. Meanwhile, the Fe₃O₄ content and the nitrogen content of the nFe₃O₄@NH₂MIP kept constant at almost 35% and 7.8%, respectively, for those prepared with different kinds of CPs as templates, thus their saturation moments were at around 32 emu g⁻¹. The results showed that with a relatively larger amine molecular anchored onto the polymer, the content of Fe₃O₄ in the nFe₃O₄@NH₂MIP decreased, which led a slight decrease of the saturation moment of the nFe₃O₄@NH₂MIP with increase of the amino chain growth, while almost no obvious difference of the saturation moment for the nFe₃O₄@NH₂MIP with different kinds of CPs were used as template since effects of the molecular printing templates on the magnetic properties of the MIP materials were slight.

Table 1 Content of Fe₃O₄, N and saturation moments of nFe₃O₄@NH₂MIP

Absorbents	Content of Fe3O4 (%)	Content of N (mmol g^−1)	Saturation moments (emu g^−1)
OA-M	97.53	—	73.98
nFe3O4@EDA–PCP-MIP	40.4	4.83	37.8
nFe3O4@DETA–PCP-MIP	39.5	5.45	36.3
nFe3O4@TETA–PCP-MIP	38.6	6.22	35.7
nFe3O4@TEPA–PCP-MIP	35.4	7.88	32.6
nFe3O4@TEPA–2,3,4,6-TeCP-MIP	35.2	7.85	32.3
nFe3O4@TEPA–2,4,6-TCP-MIP	35.0	7.78	32.5
nFe3O4@TEPA–2,4-DCP-MIP	35.2	7.82	32.8
nFe3O4@TEPA–2-CP-MIP	35.1	7.74	32.2

---

FTIR was used to further confirm the synthesis route of nFe₃O₄@TEPA–PCP-MIP. As shown in Fig. 1, the FTIR spectra of nFe₃O₄@eO-polymer (Fig. 1a) had the characteristic absorptions of C=O groups at ∼1728 cm⁻¹, C=C groups of benzene at ∼1603 cm⁻¹, C–O–C groups at ∼1269 cm⁻¹ and ∼1147 cm⁻¹, and Fe₃O₄ at ∼589 cm⁻¹. After the reaction with PCP–TEPA template molecule (Fig. 1b), a new peak located at 722 cm⁻¹, attributed to the stretching vibration of C–Cl bond of PCP, can be clearly observed.^32,33 Broad peaks appeared at ∼3380 cm⁻¹ and ∼1631 cm⁻¹ can be assigned to be the stretching and bending vibrations of the –NH and –NH₂ groups. After the template molecule (PCP) was cleaned (Fig. 1c), the characteristic bands at ∼1631 cm⁻¹ disappeared along with the appearance of the bands at ∼1590 cm⁻¹, which may be attributed to the breaking the hydrogen bonds (–N–H⋯O) between amino groups and the PCP. The formation of these hydrogen bonds between –NH₂ groups and PCP might lead the weakening of the N–H bonding and in a large shift (40–70 cm⁻¹).³⁴

Fig. 1 FTIR curves of nFe₃O₄@eO-polymer (a), nFe₃O₄@TEPA–PCP-MIP (b, with template) and (c, without template).

3.2 Method optimization

Various parameters affecting the extraction efficiency of the dMSPE of the five target CPs were optimized, including solution pH, extraction time, desorption time, desorption solution, volume of desorption solution, kinds of absorbents, etc. Each experiment was performed in triplicate.

3.2.1 Optimization of solution pH. Since the basic principle of the dMSPE procedure is based on adsorption and desorption of the target analytes and absorbents, the solution pH always plays an important role in the dMSPE process. In this work, the effect of pH on the extraction efficiencies of CPs at 5.0 μg L⁻¹ was investigated with pH values ranging from 2.0 to 10.0, and the results were shown in Fig. 2. The extraction capacity of CPs was dependent on solution pH. With the solution pH increasing, the extraction capacities firstly increased with the solution pH ranging from 2.0 to 4.0, and reached a plateau value with pH ranging from 4.0 to 8.5, then sharply decreased at the pH > 8.5. The dependence of CPs extraction on solution pH could be explained from the perspectives of surface charge of the adsorbent and the state of CPs at various pH values.²⁴ Under acidic conditions (pH < 4.0), the amino groups on the surface of nFe₃O₄@TEPA–PCP-MIP are easy to be protonated, and the main formation was –NH₃⁺ without lone pair electrons and it was difficult to form hydrogen bond (–O–H⋯N) with CPs. The main driving forces for CPs extraction were only π–π interaction between the benzene ring of CPs and those in the polymer matrix. At pH ranging from 4.0 to 8.5, the main formation of the surface groups might be –NH₂. Both hydrogen bond (–O–H⋯N) and π–π interaction coexisted to obtain the highest extraction capacity of CPs. When pH > 8.5, OH⁻ ions may be adsorbed to the surface of nFe₃O₄@TEPA–PCP-MIP, which contributed to the negatively charged sites of the nFe₃O₄@TEPA–PCP-MIP. Meanwhile, CPs molecules presented in an ionic state (deprotonation of hydroxyl group, –O⁻), and there would be repulsion between these sites and the deprotonation state of CPs, finally resulted in difficulties to form hydrogen bonds (–O–H⋯N) with amino groups on the surface of nFe₃O₄@TEPA–PCP-MIP.

Fig. 2 Effect of pH on the extraction efficiencies of CPs at 5.0 μg L⁻¹.

Since the pH values of the real seawater samples were at 7.6–8.2, no pH adjustment was need for real sample preparation as high extraction efficiencies can be obtained in this pH range.

3.2.2 Optimization of extraction time. The effect of extraction time was investigated in the range of 1–30 min. The results showed that the extraction process was initially rapid within the first 5 min for all the five CPs and reached equilibrium at 8 min with no further increase out to 30 min (shown in Fig. 3), to shorten the analysis time, all further extractions of CPs were carried out using 10 min as the optimized extraction time. However, in the case of nFe₃O₄@TEPA-NIP, 90 min were needed to reach extraction equilibrium with low adsorption capacities (34–50% of those of nFe₃O₄@TEPA–PCP-MIP) (shown in Fig. S3†). This revealed that the presence of surface imprinting cavities in absorbents could effectively shorten the extraction time, implying that the surface imprinting and uniform structures of nFe₃O₄@TEPA–PCP-MIP allowed efficient mass transport, thus overcoming some drawbacks of traditionally non-imprinting materials. Similar phenomenon was found in our previous work²⁹ and literature reports.³⁵

Fig. 3 Effect of extraction time on the extraction efficiencies of CPs at 5.0 μg L⁻¹.

3.2.3 Optimization of desorption parameters. In order to obtain high desorption efficiencies, some desorption parameters were optimized.

First, three kinds of solvents were selected as desorption solution for five CPs, including pure methanol, 2% ammonium hydroxide–methanol solution (v/v), and 2 mmol L⁻¹ NaOH methanol solution. The results of desorption efficiencies were shown in Fig. 4a. When pure methanol was used, only the recoveries of 2,4,6-TCP, 2,3,4,6-TeCP and PCP can reach above 80%, the recoveries of the other two kinds of CPs were quite low (<80%), which indicated that pure methanol might not be a suitable solvent for the desorption all the five studied CPs. The results confirmed the above discussion on adsorption mechanism that besides π–π interactions among the benzene rings of CPs and polymer matrices, hydrogen-bonding interactions among amino groups and the phenolic hydroxyl group of CPs were formed. Since there are more substitutional groups (–Cl) in 2,4,6-TCP, 2,3,4,6-TeCP and PCP, due to their greater steric and stronger electron withdrawing effects, and their adsorption on the materials might be more relying on π–π interactions. Therefore, most 2,4,6-TCP, 2,3,4,6-TeCP and PCP could be desorbed by pure methanol, while much stronger interaction (hydrogen-bonding interaction) with 4-CP and 2,4-DCP on the adsorbent, which was more polar and highly hydrophilic, thus they were not easily desorbed by neuter pure methanol. To improve the desorption efficiencies, 2% (v/v) ammonium hydroxide methanol solution was used as a modifier. As shown in Fig. 4a, with the ammonium hydroxide adding into methanol, the absolute recoveries of 2,4-DCP obviously increased compared with pure methanol, which is contributed to that alkali can break the formation of hydrogen bonds, while the desorption recovery of 2-CP (74.5%) had not been obviously improved. This might be due to the fact that much 2-CP was lost during solvent evaporation owing to its high volatility. This can be overcome by directly adding 2 mmol L⁻¹ NaOH in methanol, which can convert 2-CP into ionized form. Therefore, 2 mmol L⁻¹ NaOH in methanol solution was used to elute the five CPs, and high desorption efficiencies and stable recoveries were obtained.

Fig. 4 Desorption parameters optimization: (a) desorption solvents (b) desorption time, (c) volume of desorption solvent (CPs spiked concentrations at 5.0 μg L⁻¹).

The volume of desorption solution (2 mmol L⁻¹ NaOH) was also investigated ranging from 0.5 to 3 mL. The results showed that 1 mL of the desorption solution was enough to elute these target CPs (Fig. 4b). To shorten the evaporation time, 1 mL of the desorption solution was selected for the further experiments.

The effect of desorption time was also studied and the results were shown in Fig. 4c. Although 2 min was found enough to elute 2,4,6-TCP, 2,3,4,6-TeCP, and PCP, while longer desorption time (4 min) was required for completely desorption of 2-CP and 2,4-DCP. Thus, 4 min of desorption was selected.

3.2.4 Optimization of extraction volume. The optimization of extraction volume for CPs was investigated using a series of different volume aqueous solutions (ranging from 50 to 800 mL) spiked with 0.5 μg of each the CPs. The amount of nFe₃O₄@TEPA–PCP-MIP added was kept 20.0 mg. Insufficient recovery was considered to occur when recovery was below 90%. The results were shown in Fig. 5. The results showed insufficient recovery occurred when seawater volume was above 500 mL for 2,4,6-TCP, 2,3,4,6-TeCP, and PCP, thus the optimized extraction volume for CPs was suggested to be 500 mL. For the five CPs, complete desorption was obtained with 1 mL of methanol containing 2 mmol L⁻¹ NaOH solution. By drying the desorption solution with a nitrogen flow and re-dissolving the CPs in 0.5 mL methanol. A theoretical values of the enrichment factor (EF), defined as the ratio between the analyte concentration in the final extract and its initial concentration in the sample, for each CP would be expected to be 1000. Bearing on mind that a dSPME process takes place in the proposed method, the actual EF value was calculated in this work by comparison of the slopes of the calibration lines obtained after and before enrichment.³⁶ The results were listed in Table 2, ranging from 486 to 496, for the 5 studied CPs.

Fig. 5 Extraction volume optimization.

Table 2 LODs, LOQs and linear range of the 5 kinds of studied CPs

CP	Linear range (ng L^−1)	Linear regression	R	LODs (ng L^−1)	LOQs (ng L^−1)	EF
4-CP	2–5000	y = 50.126x + 83.318	1.0000	0.25	0.9	486
2,4-DCP	5–5000	y = 33.469x + 33.924	0.9999	1.20	4.0	496
2,4,6-TCP	2–5000	y = 36.048x − 7.631	0.9999	0.54	2.0	490
2,3,4,6-TeCp	2–5000	y = 59.366x + 70.093	0.9989	0.36	1.2	486
PCP	1–5000	y = 60.563x + 132.32	0.9998	0.18	0.6	492

---

3.3 Reproducibility and reusability of nFe₃O₄@TEPA–PCP-MIP

Reproducibility and the reusable of this laboratory-made material nFe₃O₄@TEPA–PCP-MIP were evaluated by comparing the extraction efficiency. Five batches of nFe₃O₄@TEPA–PCP-MIP were prepared and the reproducibility of the extraction was investigated. The results were shown in Fig. 6a. The mean extraction recoveries of five batches of absorbents toward five CPs were in range of 92.8–100.5% with RSDs ranging from 0.6% to 1.5%. The low RSDs (<5%) indicated that good reproducibility was achieved using the self-made nFe₃O₄@TEPA–PCP-MIP for the extraction of CPs. The post-desorbed nFe₃O₄@TEPA–PCP-MIP was reused, and results were shown in Fig. 6b. The results indicated that nFe₃O₄@TEPA–PCP-MIP could be used for at least five cycles with a loss of less than 2.5% upon recovery on average. No obvious decrease in the extraction efficiency was found, implying that the nFe₃O₄@TEPA–PCP-MIP was stable and could be recycled.

Fig. 6 Reproducibility (a) and reusability (b) of nFe₃O₄@TEPA–PCP-MIP.

3.4 Comparing with other laboratory-made MIP materials

It is of great importance to assess the selective recognition towards the template molecule for a novel imprinted material. Adsorption selectivity is known to be related to the size, shape, and functionality of the template molecule of the imprinted cavities in imprinted materials. Unlike single target selective recognition, a high recognition of analogues CPs is desirable for multi-residue determination of CPs. Thus, other seven kinds of MIP absorbents with different kinds of amines, i.e., nFe₃O₄@EDA–PCP-MIP, nFe₃O₄@DETA–PCP-MIP, and nFe₃O₄@TETA–PCP-MIP and with different kinds of template molecules, i.e., nFe₃O₄@TEPA–2-CP-MIP, nFe₃O₄@TEPA–2,4-DCP-MIP, nFe₃O₄@TEPA–2,4,6-TCP-MIP and nFe₃O₄@TEPA–2,3,4,6-TeCP-MIP were prepared. Their selectivity to CPs and recoveries were investigated.

The recoveries obtained with nFe₃O₄@TEPA–PCP-MIP as the uSMPE sorbent for the studied 5 CPs ranged between 92.4% and 97.4% with the RSD values of targets were lower than 3%. However, the recoveries obtained with the other 7 kinds of MIP absorbents varied from 58.6% to 95.2% with the RSD values varied from 5.2% to 12.4% (Fig. 7). The results showed the extraction recoveries were improved significantly with a broader range of recognition of analogues CPs. This could be attributed to the fact that in extracting process, specific recognition sites to the studied CPs generated on the surface of nFe₃O₄@TEPA–PCP-MIP, thus CPs was strongly bound to the specific binding sites. While the recognition sites of the other kinds of MIPs were not complementary to all of them, leading much less recoveries than that of nFe₃O₄@TEPA–PCP-MIP.

Fig. 7 Comparing with other laboratory-made MIP materials.

Further investigation on the imprinting factor (α) of the nFe₃O₄@NH₂MIP was carried out. The imprinting factor (α) is defined as:

where, the equilibrium adsorption capacity (q, μg g⁻¹) was calculated using the following equation.
where C₀ and C_e represent solution concentrations before and after adsorption (μg L⁻¹), V is the volume of solution (mL), m is the adsorbent dosage (mg), respectively.

To evaluate the selectivity of the synthesized nFe₃O₄@NH₂MIPs, the uptake capacities of the sorbent for 5 CPs were tested, rebinding experiments were performed using a 5 CPs mixture with the initial concentration of 5 μg L⁻¹, and the q and α were calculated. The results were listed in Table S1.† Generally, the cavities of imprinted sorbent created by the imprinted molecules can lead to much greater affinity for the template molecule, in comparison with non-imprinted sorbent. Interestingly, unlike other nFe₃O₄@NH₂MIPs, nFe₃O₄@TEPA–PCP-MIP showed a high α to all the 5 CPs (2.18–3.03). This interesting phenomenon might be attributed to the fact that PCP might act as a “Master key” for all the studied 5 kinds of CPs, thus leading high adsorption/desorption properties, as well as relatively similar recoveries to the five all CPs. In literature, “Master key”-strategy has been realized by dual or multi-template MIP.³⁵

3.5 Method linear range, limits of detection, limits of quantification, precision, and accuracy

The calibration curves were constructed by plotting the mean peak areas (y) versus concentrations of CPs (x, ng L⁻¹) with triplicate measurements. The results were listed in Table 2. Linear ranges for the five CPs were in range of 1–5000 ng L⁻¹ with coefficients of correlation (R) higher than 0.9989. The limits of detection (LODs) and limits of quantification (LOQs) were calculated as the concentrations of the analytes at a signal-to-noise ratio (S/N) of 3 and 10, respectively. The results, listed in Table 2, showed that the LODs and LOQs of the five target CPs range from 0.18 to 1.20 ng L⁻¹ and 0.6 to 4.0 ng L⁻¹, respectively.

The method precision and accuracy were measured based on the analyses of five CPs spiked in seawater samples. The results were summarized in Table 3. It shows that the majorities of mean recoveries are in the range of 86.5–98.8% at the three spiked levels with associated intra-day relative standard deviations (RSDs) ranging from 0.8% to 7.5% and inter-day RSDs ranging from 1.2% to 8.8%.

Table 3 Precision and accuracy of the five CPs in synthesized sea water samples (n = 5)

CP	Spiked (ng L^−1)	Recovery (%)	RSD, %
			Intra-day	Inter-day
4-CP	2.0	88.7	4.2	5.8
	10.0	96.8	0.9	1.5
	50.0	94.2	0.8	1.6
2,4-DCP	5.0	86.5	3.2	5.8
	10.0	94.2	2.8	2.5
	50.0	88.8	1.7	2.8
2,4,6-TCP	2.0	94.8	5.6	7.2
	10.0	98.4	3.5	4.8
	50.0	96.3	1.5	2.4
2,3,4,6-TeCP	2.0	95.2	7.5	8.8
	10.0	96.8	2.6	3.8
	50.0	97.2	1.6	2.8
PCP	1.0	98.2	2.8	3.5
	10.0	97.5	2.2	2.5
	50.0	98.8	1.5	2.2

---

3.6 Comparing with other methods

Comparing of the present method and literature reported ones was listed in Table 4. It shows that the present method has lower LODs than the listed methods based on GC-FID, GC-ECD and GC-MS, along with avoiding the complicated and tedious derivatization process.^37,38 Moreover, SPME and LPME technologies often needed for CPs enrichment before GC or LC analysis, which are time-consuming and generate large volumes of waste solvents.^16,19,40 Although MSPE was applied for CPs enrichment by some researchers,^12,14,41 their LODs were poor ranging from 110 to 350 ng L⁻¹. The LODs of the present work were comparable to our previous work using uMSPE-LC-MS procedure, in which an amino-functionalized magnetic polymer modified graphene oxide as magnetic solid-phase extraction materials was used for enrichment.²³ Comparing with the literature, the proposed method is simple, fast, solvent-saving, sensitive and accurate for the analysis of CPs in seawater samples.

Table 4 Comparison with the current analytical technique for the determination of CPs in water samples^a

No	Sample type	Enrichment material	Analytic method	Number of Cl atoms	LOD (ng L^−1)	Ref.
a Abbreviations: MW, microwave; HS, headspace; SPME, solid-phase microextraction; LPME, liquid-phase microextraction; LLLME, liquid–liquid–liquid microextraction; SPE, solid-phase extraction; MSPE, magnetic solid-phase extraction.
1	Well, tap water	Electropolymerized aniline-based fiber	HS-SPM:GC-FID	1–3, 5	700–58 000	37
2	River, lake, waste water	CHCl3	LPME-tosylation-GC-MS	1–3, 5	200–280	38
3	Sea, ground, wastewater	Carbowax/TPR-100 (CW/TPR) fiber	SPME-LC/DAD	1–3, 5	1000–6000	39
4	Lake, ground, wastewater	Decanoic acid vesicle-based coacervates	LPME-LC/DAD	1–3, 5	100–300	40
5	Seawater	Ionic liquid [BMIM][PF6]	LLLME-LC/UV	1–3, 5	20–100	19
6	River, lake, drinking water	Oasis HLB SPE	SPE-LC/MS/MS	1–3, 5	1–7	16
7	Wastewater	Magnetic polysulfone microcapsule	MSPE-LC/UV	1	170–220	41
8	River, spring, tap water	Ionic liquid-functionalized magnetic microspheres	MSPE-LC/DAD	1, 2, 5	200–350	14
9	River, tap, wastewater	CTAB–Fe3O4	MSPE-LC/MS	1–3, 5	110–150	12
10	River, tap water, well water	NH2-MP@GO	MSPE-LC/MS	1–5	0.16–2.8	23
11	Seawater	nFe3O4@TEPA–PCP-MIP	MSPE-LC/UV	1–5	0.18–1.2	This work

---

3.7 Application

The proposed method was examined by analyzing the CPs from real seawater samples. Five different kinds of seawater collected from Ningbo Coast, Zhejiang, China, labeled as S1–S5, were tested. The results were listed in Table 5. The obtained results show that PCP and 2,4,6-TCP were detected in S1 at concentration of 12.5 ng L⁻¹ and 8.6 ng L⁻¹, respectively. Liquid chromatograms of blank (S5), spiked sample (2 μg L⁻¹ in S5), and S1, after dMSPE treatment, were shown in Fig. 8. Compared with the National Standard method of China,⁴² USEPA method 8270d,⁴³ and our previous work (NH₂-MP@GO based uMSPE-LC-LC-MS method),²³ it demonstrated that the developed method was applicable for real seawater sample analyses with a much cheap, accurate and reliable analytical procedure.

Table 5 Real sample determination of CPs in seawater samples

CP	LOQs (ng L^−1)		Concentration found (ng L^−1)
			S1	S2	S3	S4	S5
4-CP	This work (dMSPE-LC)	0.9	ND	ND	ND	ND	ND
	Ref. 23 (uMSPE-LC-MS)	9.2	ND	ND	ND	ND	ND
	Ref. 42 (LLE-derivatization-GC-ECD)	NA	ND	ND	ND	ND	ND
	Ref. 43 (LLE-SPE-GC-MS)	NA	ND	ND	ND	ND	ND
2,4-DCP	This work (dMSPE-LC)	4.0	ND	ND	ND	ND	ND
	Ref. 23 (uMSPE-LC-MS)	3.6	ND	ND	ND	ND	ND
	Ref. 42 (LLE-derivatization-GC-ECD)	NA	ND	ND	ND	ND	ND
	Ref. 43 (LLE-SPE-GC-MS)	10 000	ND	ND	ND	ND	ND
2,4,6-TCP	This work (dMSPE-LC)	2.0	8.6	ND	ND	ND	ND
	Ref. 23 (uMSPE-LC-MS)	1.8	8.5	ND	ND	ND	ND
	Ref. 42 (LLE-derivatization-GC-ECD)	NA	ND	ND	ND	ND	ND
	Ref. 43 (LLE-SPE-GC-MS)	10 000	ND	ND	ND	ND	ND
2,3,4,6-TeCp	This work (dMSPE-LC)	1.2	ND	ND	ND	ND	ND
	Ref. 23 (uMSPE-LC-MS)	1.2	ND	ND	ND	ND	ND
	Ref. 42 (LLE-derivatization-GC-ECD)	NA	ND	ND	ND	ND	ND
	Ref. 43 (LLE-SPE-GC-MS)	10 000	ND	ND	ND	ND	ND
PCP	This work (dMSPE-LC)	0.6	12.5	ND	ND	ND	ND
	Ref. 23 (uMSPE-LC-MS)	0.6	12.3	ND	ND	ND	ND
	Ref. 42 (LLE-derivatization-GC-ECD)	10	12.6	ND	ND	ND	ND
	Ref. 43 (LLE-SPE-GC-MS)	50 000	ND	ND	ND	ND	ND

---

Fig. 8 HPLC of seawater sample (a, S5), (b, S1) and spiked sample (2 μg L⁻¹ in S5) (c).

4. Conclusions

Eight kinds of amino-functionalized core–shell CP-molecularly imprinted magnetic polymer (nFe₃O₄@NH₂MIP) were successfully synthesized. In which, nFe₃O₄@TEPA–PCP-MIP was proved to be the most valuable candidate as the three-in-one dMSPE adsorbent for preconcentration of five trace-level CPs in seawater in one step. The developed analytical procedure was proven to be effective, fast, and accurate for routine analyses. The present magnetic imprinting mode opens attractive perspectives for the development of dMSPE-based preconcentration systems for persistent toxic substances (PTS). Such systems could be invaluable for the rapid analytical control of PTS in seawater, as well as environmental pollution monitoring and remediation. Recently, dual or multi-template MIP strategy has attracted attention for satisfactory cleanup and enrichment efficiency to some pollutants in seawater, such as polycyclic aromatic hydrocarbons (PAHs),³⁵ which inspired our interest for further investigation. Studies on multi-template MIP strategy for trace analysis are ongoing in laboratory. Favourable results are being expected.

Acknowledgements

We would like to thank the National Natural Science Foundation of Zhejiang Province (LY14B04003), the National Natural Science Foundation of Ningbo (2014A610092), the National College Students' innovation and entrepreneurship training program (201513022009), the Xinmiao Students' innovation training program of Zhejiang Province (2015R401181) for the financial support.

References

1. F. Galán-Cano, R. Lucena, S. Cárdenas and M. Valcárcel, J. Chromatogr. A, 2012, 1229, 48.
2. J. Michalowicz and W. Duda, Pol. J. Environ. Stud., 2007, 16, 347.
3. M. D. Scrimshaw and J. N. Lester, Environ. Technol., 1996, 17, 951.
4. Toxic Substance Control Act, U.S. Environmental Protection Agency, Washington, DC, 1979.
5. J. A. Padilla-Sánchez, P. Plaza-Bolños, R. Romero-González, N. Barco-Bonilla and J. L. Martínez-Vidal, Talanta, 2011, 85, 2397.
6. J. L. Zhang, R. Ye, X. X. Xu, L. J. Yu, S. S. Lv, L. N. Zhu, S. Ye, H. Y. Shen and Q. H. Xia, Acta Mater. Compositae Sin., 2015, 32, 1201.
7. Z. B. Zhang, A. D. Chen, L. S. Liu and Z. D. Wang, Principles and applications of marine chemistry: chemical Chinese offshore marine, Marine Press, Beijing, 1999, p. 336.
8. X. Li, Z. Zeng and J. Zhou, Anal. Chim. Acta, 2004, 509, 27.
9. F. Bianchi, M. Careri, C. Mucchino and M. Musci, Chromatographia, 2002, 55, 595.
10. M. Llompart, M. Lourido, P. Landín, C. García-Jares and R. Cela, J. Chromatogr. A, 2002, 963, 137.
11. Q.-Z. Feng, L.-X. Zhao, W. Yan, J.-M. Lin and Z.-X. Zheng, J. Hazard. Mater., 2009, 167, 282.
12. H. Bagheri, A. Mohammadi and A. Salemi, Anal. Chim. Acta, 2004, 513, 445.
13. J. Li, X. Zhao, Y. Shi, Y. Cai, S. Mou and G. Jiang, J. Chromatogr. A, 2008, 1180, 24.
14. F. E. O. Suliman, S. S. Al-Kindi, S. M. Z. Al-Kindy and H. A. J. Al-Lawati, J. Chromatogr. A, 2006, 1101, 179.
15. F. Yang, R. Shen, Y. Long, X. Sun, F. Tang, Q. Cai and S. Yao, J. Environ. Monit., 2011, 13, 440.
16. J. L. M. Vidal, A. B. Vega, A. G. Frenich, F. J. E. González and F. J. A. Liebanas, Anal. Bioanal. Chem., 2004, 379, 125.
17. S. Marchese, A. Gentili, D. Perret, M. Sergi and S. Notari, Chromatographia, 2004, 59, 411.
18. S. Almeda, L. Nozal, L. Arce and M. Valcárcel, Anal. Chim. Acta, 2007, 587, 97.
19. M. Saraji and M. Marzban, Anal. Bioanal. Chem., 2010, 396, 2685.
20. L. Guo and H. K. Lee, J. Chromatogr. A, 2011, 1218, 4299.
21. A. H. El-Sheikh, R. W. Al-Quse, M. I. El-Barghouthi and F. S. Al-Masri, Talanta, 2010, 83, 667.
22. C. L. Arthur and J. Pawliszyn, Anal. Chem., 1990, 62, 2145.
23. S. D. Pan, L. X. Zhou, Y. G. Zhao, X. H. Chen, H. Y. Shen, M. Q. Cai and M. C. Jin, J. Chromatogr. A, 2014, 1362, 34.
24. X. H. Chen, Y. G. Zhao, Y. Zhang, H. Y. Shen, S. D. Pan and M. C. Jin, Anal. Methods, 2015, 7, 5838.
25. J. H. Li, R. C. Dong, X. Y. Wang, H. Xiong, S. F. Xu, D. Z. Shen, X. L. Song and L. X. Chen, RSC Adv., 2015, 5, 10611.
26. X. Q. Wu, X. Y. Wang, W. H. Lu, X. R. Wang, J. H. Li, H. Y. You, H. Xiong and L. X. Chen, J. Chromatogr. A, 2016, 1435, 30.
27. L. X. Chen, S. F. Xu and J. H. Li, Chem. Soc. Rev., 2011, 40, 2922.
28. L. X. Chen, X. Y. Wang, W. H. Lu, X. Q. Wu and J. H. Li, Chem. Soc. Rev., 2016, 45, 2137.
29. H. Y. Shen, Z. X. Chen, Z. H. Li, M. Q. Hu, X. Y. Dong and Q. H. Xia, Colloids Surf., A, 2015, 481, 439.
30. H. Y. Shen, S. D. Pan, Y. Zhang, X. L. Huang and H. X. Gong, Chem. Eng. J., 2012, 183, 180.
31. H. Deng, X. L. Li, Q. Peng, X. Wang, J. P. Chen and Y. D. Li, Angew. Chem., Int. Ed., 2005, 44, 2782.
32. J. Chen, M. A. Hamon, H. Hu, Y. Chen, A. M. Rao, P. C. Eklund and R. C. Haddon, Science, 1998, 282, 95.
33. H. Y. Shen, Z. J. Wang, A. M. Zhou, J. L. Chen, M. Q. Hu, X. Y. Dong and Q. H. Xia, RSC Adv., 2015, 5, 22080.
34. Y. G. Zhao, H. Y. Shen, S. D. Pan, M. Q. Hu and Q. H. Xia, J. Mater. Sci., 2010, 49, 5291.
35. X. L. Song, J. H. Li, S. F. Xu, R. J. Ying, J. P. Ma, C. Y. Liao, D. Y. Liu, J. B. Yu and L. X. Chen, Talanta, 2012, 99, 75.
36. F. Pena-Pereira, N. Cabaleiro, I. D. L. Calle, M. Costas, S. Gil, I. Lavilla and C. Bendicho, Talanta, 2011, 85, 1100.
37. H. Bagheri, A. Mir and E. Babanezhad, Anal. Chim. Acta, 2005, 532, 89.
38. Y. C. Fiamegos, A. P. Kefala and C. D. Stalikas, J. Chromatogr. A, 2008, 1190, 44.
39. C. M. Santana, M. E. T. Padrón, Z. S. Ferrera and J. J. S. Rodríguez, J. Chromatogr. A, 2007, 1140, 13.
40. F. J. López-Jiménez, S. Rubio and D. Pérez-Bendito, J. Chromatogr. A, 2008, 1195, 25.
41. X. Liu, J. Yin, L. Zhu, G. Zhao and H. Zhang, Talanta, 2011, 85, 2451.
42. Water quality-Determination of Pentachlorophenol by Gas Chromatography, Environmental Protection Agency of China, Beijing, 2010.
43. EPA method 8270d, U.S. Environmental Protection Agency, Washington, DC, 2007.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16413g

---