Layer-by-layer coated molecular-imprinted solid-phase microextraction fibers for the determination of polar compounds in water samples

Abstract

In this study, a novel technique was reported for the selective extraction and analysis of polar compound in water samples using molecular-imprinted solid-phase microextraction (MISPME) combined with dispersive liquid–liquid microextraction (DLLME) with in situ derivatization. 1,2-Benzenediol (1,2-BD) was used as a model compound to present the validity of novel approach of analyzing polar compounds in aqueous samples. The layer-by-layer structure MISPME fiber was first prepared using 1,2-phenylene diacetate (1,2-PDA, derivatization product of 1,2-BD) as the template. Subsequently, 1,2-BD was extracted from water samples by DLLME after in situ derivatization using acetic anhydride as a derivatization agent, dichloromethane as an extraction solvent and methanol as a disperser solvent. Finally, the derivatization product 1,2-PDA in the extraction solvent was selectively enriched by the MISPME fiber, followed by analysis by gas chromatography-flame ionization detection (GC-FID). The preparation and extraction conditions were optimized. The method exhibited a good linear relationship in a wide range from 0.05 to 10 μg mL⁻¹ (r² = 0.9984), which indicated that the established method is applicable for the quantification of 1,2-BD. The limit of detection (LOD) was 0.01 μg mL⁻¹. The precision of the method was evaluated by analyzing a standard solution of 1,2-BD at 0.5 μg mL⁻¹, and the RSD was 5.9%.

---

1 Introduction

Sample preparation is a crucial step in the analysis process and is considered a bottleneck to rapidly obtaining an accurate result. The main objectives of sample preparation are the removal of potential interferents and separation and pre-concentration of analytes from the sample matrix for detection. Traditional methods for sample preparation, such as liquid–liquid extraction (LLE) usually suffer from being time-consuming and labor-intensive and require large quantities of expensive, toxic, and environmentally unfriendly organic solvents. A lot of new sample preparation methods have been developed, such as solid-phase extraction (SPE),¹ liquid-phase microextraction (LPME),² dispersive liquid–liquid microextraction (DLLME)³ and solid-phase microextraction (SPME)⁴ have been developed and used in the analysis of real samples. Among them, SPME, developed by Pawliszyn in 1989,⁵ is a simple, sensitive, time-efficient and solvent-free technique and has been applied widely to the analysis of environmental, food, biological, and pharmaceutical samples. Although a variety of fiber coatings for SPME are now commercially available, the selectivity of the extraction process is low, and the determination of target analytes at trace levels in complex samples is challenging using chromatographic techniques coupled with common detectors.

Molecularly imprinted polymers (MIPs) were used as SPME fibers for the recognition of target analytes.^6–9 MIPs are synthetic materials possessing artificially generated cavities and recognition sites on the polymer surface and are able to selectively separate target molecules from other structural analogue compounds or complex sample matrices. The resulting molecular-imprinted, solid-phase microextraction (MISPME) fibers show specific selectivity towards target analytes in complex sample matrices, such as foods,¹⁰ environmental samples¹¹ and biological fluids.¹² Moreover, the chemical stability and easy-to-prepare characteristics make MIPs suitable as materials for SPME fiber coatings. In spite of the above-mentioned advantages, there are two problems that greatly hinder the development of MISPME. The extraction capacity is limited by the relatively small surface of MISPME fibers, which only offer a limited number of recognition cavities. To improve the extraction efficiency of MISPME fibers, several polymerization methods have been used to prepare novel MISPME fibers.^13–15 Additionally, the application of MISPME fibers is disturbed by the water compatibility problem because MIPs have poor extraction efficiency for target molecules in aqueous environments. To overcome the water-compatibility problem of the MISPME, a novel liquid–liquid–solid microextraction (LLSME), which integrated traditional LLE and MISPME, was proposed to provide a solution.^16,17 In this method, target analytes are first enriched between the aqueous solution and the organic solvent through LLE and are then separated selectively between the organic solvent and the MIP coating through MISPME. However, this method can only be used to extract non-polar target analytes because the analytes must first be transferred from aqueous samples into a non-polar solvent. For some polar analytes, such as phenols, poor selectivity and low extraction efficiency were shown due to their high solubility in aqueous solution.

To achieve high extraction efficiency for polar analytes in aqueous solution, we had two aims in this work. The first was to increase the surface area of the fiber and to improve the extraction capacity of the MISPME fiber through the design of the layer-by-layer structure. The second was to provide an efficient method for the extraction of polar analytes based on MISPME combined with in situ derivatization of DLLME. This method can solve the low recognition extraction efficiency problem of polar analytes in aqueous samples. In this strategy, 1,2-BD was used as a model compound and 1,3- and 1,4-benzenediol (1,3-, 1,4-BD) as structural analogs to present the validity of novel approach of analyzing polar compounds in aqueous samples. With these goals in mind, an MISPME fiber was prepared with 1,2-PDA (derivatization product of 1,2-BD) as a template. Subsequently, 1,2-BD and its structural analogs were extracted from water samples by DLLME after in situ derivatization using acetic anhydride as a derivatization agent, dichloromethane as an extraction solvent and methanol as a disperser solvent. Finally, the derivatization product in the extraction solvent was selectively extracted by the MISPME, followed by analysis using GC-FID. The parameters affecting the microextraction efficiency were investigated in details and the optimal conditions were established.

The present method synergistically combines the advantages of MISPME and DLLME with in situ derivatization. The results of experiment indicated that the proposed method has some advantages, with a low detection limit and high recognition extraction efficiency for polar analytes. The applicability of the new fiber has been demonstrated by the analysis of real water samples.

2 Experimental

2.1 Chemicals and reagents

1,2-, 1,3-, and 1,4-BD and 1,2-PDA were purchased from Sinopharm Chemical Reagent Co. Ltd (Shanghai, China). Methacrylic acid (MAA) and 2,2-azobisisobutyronitrile (AIBN) was purchased from Xilong Chemical. Co. Ltd (Guangdong, China). Ethyleneglycol dimethacrylate (EDMA) was purchased from Fushun Anxin Chemical Ltd (Shenyang, China). Stock standard solutions of 1,2-, 1,3-, and 1,4-BD were prepared in methanol at a concentration of 1 mg mL⁻¹ and were stored in glass-stoppered bottles in the dark at 4 °C. Fresh working solutions were prepared by the dilution with an appropriate amount of phosphate buffer solution (PBS, pH 7.00). All of the other chemicals were of analytical grade and were used as received. Pure water was obtained from a Simplicity Personal Ultrapure Water System (Millipore, USA).

2.2 Instrumentation

Chromatographic analysis was performed on a Shimadzu GC-14C gas chromatograph equipped with a split/splitless injector and a flame ionization detector (FID) (Shimadzu, Japan). An Rtx-50 fused-silica column (30 m × 0.25 mm I.D.) with 50% phenyl and 50% methyl polysiloxane (film thickness 0.25 μm) (Restek, USA) was used. High purity nitrogen (99.999%) was used as the carrier gas (1.0 mL min⁻¹) and make-up gas (30 mL min⁻¹). The instrumental temperatures were as follows: injector temperature 300 °C; detector temperature 300 °C; initial oven temperature 40 °C for 4 min, increased to 280 °C at a rate of 20 °C min⁻¹, and held for 5 min. Hydrogen and air were used as detector gases at 48 mL min⁻¹ and 450 mL min⁻¹, respectively. The inlet was operated in splitless mode (3 min). Clarity Shimadzu was utilized to control the system and to acquire the analytical data.

2.3 Preparation of the MISPME fiber with a layer-by-layer structure

2.3.1 Preprocessing of stainless steel wire. A homemade SPME device was constructed following the procedure described in our previous study.¹⁷ Prior to coating, bare stainless steel wire with a smooth surface was etched by HF solution for 30 min at room temperature and was sequentially rinsed with acetone, methanol and distilled water in an ultrasonicator for 5 min each before being air-dried at room temperature.

2.3.2 Preparation of the SiO₂ coating. The SiO₂ coating was obtained by the sol–gel coating process, which was previously described in detail.¹⁸

2.3.3 Preparation of the MIP coating. The MIP coating was prepared through the photochemical-induced radical copolymerization of MAA and EDMA using 1,2-PDA as a template. For this purpose, 194 mg (1 mmol) 1,2-PDA and 0.5 mL MAA were dissolved in 8 mL acetonitrile and were stirred in ultrasonic bath for 30 min. Then, 200 mg of preprocessed carbon nanotubes (CNTs, preprocessing of CNTs described in ref. 17) was added to the solution and was dispersed for 5 min by ultrasonic agitation. Then, 5 mL EDMA and 40 mg of AIBN were added and mixed thoroughly for 5 min by ultrasonic agitation. Finally, the fibers were dipped vertically into the solution and were placed under the ultraviolet irradiation for 1 h until an MIP coating was formed. Then, the fibers were removed and dried at 80 °C to produce a thin-layer MIP coating.

2.3.4 Preparation of the MISPME fiber with a layer-by-layer structure. To obtain an MISPME fiber with a layer-by-layer structure, the fiber was coated alternately and repeatedly with SiO₂ coating and MIP coating through operations (2) and (3) described above to produce a coating of approximately 200 μm in thickness. The prepared fiber was placed in a desiccator at room temperature for 24 h. Finally, the fiber was soaked in a methanol : acetic acid (9 : 1, v/v) solution for 1 h to remove the template molecules. This procedure was performed repeatedly until no response of the template molecule was detected.

The non-imprinted solid-phase microextraction (NISPME) fibers were also prepared according to the above procedures except in the absence of template molecule during polymerization.

2.4 Derivatization/DLLME–MISPME procedure

A schematic representation of the derivatization/DLLME–MISPME procedure is shown Fig. 1. Firstly, 10 mL of working solution was placed in a 20 mL glass test tube with conical bottom. Then, 0.8 mL acetic anhydride (derivatization agent) was transferred to the tube. The tube was tightly capped and shaken vigorously for 1 min to mix the phases. Then, the tube was placed in a 50 °C water bath for derivatization. After 30 min, 1.00 mL methanol (disperser solvent) and 1.00 mL dichloromethane (extraction solvent) were rapidly injected into the solution using a 5 mL glass syringe, and the mixture was shaken vigorously for 30 s to form a cloudy solution. Then, the mixture was centrifuged for 5 min at 4000 rpm. After centrifugation, dichloromethane was settled in the bottom of the vial. Finally, the extraction solution was transferred into a glass syringe and stored in a 1.5 mL plastic test tube.

Fig. 1 Scheme of the derivatization/DLLME–MISPME procedure.

Subsequently, the MISPME or NISPME fiber was immersed in the above extraction solution to extract analytes under ultrasonic conditions at 30 °C for 15 min. After extraction, the fiber was then thermally desorbed in the GC injection port at 300 °C for 3 min.

3 Results and discussion

3.1 Preparation of the MISPME fibers

To obtain homogenous MIP coatings with high extraction ability, the surface characteristics of the support stainless steel wire are important. The wire was etched with HF solution for 30 min to acquire a rough surface and to enhance the binding stability between the wire and coating.

In MISPME, increasing the coating thickness of the extraction fiber results in larger recognition extraction efficiency. Unfortunately, it is difficult to obtain thick MIP coatings because of the weak adhesion of the pre-polymer solution to the surface of the wire. In this strategy, a layer-by-layer structure of the coating was designed, and the SiO₂ coating of the porous structure was used as a support to enhance the thickness of the fiber coating. As a result, the fiber was coated alternately and repeatedly with the SiO₂ coating and the MIP coating until the desired thickness was obtained.

To obtain the best extraction efficiency, the composition of the MIP coating was optimized. The optimization results were determined by the chromatographic peak area of the template molecule (1,2-PDA) obtained by MISPME and NISPME with 1 μg mL⁻¹ of 1,2-PDA standard solution. Firstly, the incorporation of CNTs increased the chromatographic peak area due to the increased surface area of the MIP coating. However, when the amount of CNTs was more than 200 mg, the lifespan of the fiber was shortened because of fracturing of the fiber coating. Therefore, 200 mg of CNTs was adopted in the following studies. Secondly, the quantity of functional monomer (MAA) and cross-linker (EDMA) in the pre-polymer solution were also optimized. As summarized in Table 1, the MIP coating synthesized with an MAA : EDMA at ratio of 1 : 10 (v/v) exhibited the highest extraction efficiency for 1,2-PDA.

Table 1 Effect of the amount of raw materials on the peak area of 1,2-BD

Amount of 1,2-BD (mg)	Amount of CNTs (mg)	Volume of MAA (mL)	Volume of EDMA (mL)	Peak area of 1,2-BD
194	0	0.5	5.0	124.214
194	100	0.5	5.0	156.242
194	200	0.1	1.0	32.014
194	200	0.2	2.0	61.278
194	200	0.3	3.0	100.368
194	200	0.4	4.0	142.214
194	200	0.5	5.0	210.276
194	200	0.6	6.0	180.245

---

3.2 Morphological study of the MISPME fibers

The morphological characteristics and thickness of the prepared fibers were investigated using scanning electron microscopy (SEM) (Fig. 2). Fig. 2A shows that a homogeneous and dense coating was attached to the support wire. As observed in Fig. 2B, the CNTs were well-dispersed in the MIP coating, which increased the surface area and extractive capacity of the fiber coating. Fig. 2C further shows that the coating possessed a homogeneous, porous and layer-by-layer structure, and the thickness of the coating was approximately 100 μm.

Fig. 2 SEM images of the MISPME fiber under the magnifications of 170 (A), 2000 (B) and the cross section under the magnifications of 700 (C).

3.3 Investigation of the selectivity of the in situ derivatization/DLLME–MISPME procedure

It is important to conduct experimental investigation of selectivity of the derivatization/DLLME–MISPME procedure. For this purpose, the MISPME and NISPME fibers were separately used for the extraction of a target molecule (1,2-BD) and its structural analogs (1,3-, 1,4-BD) from aqueous solutions through the derivatization/DLLME–MISPME method. The results are shown in Fig. 3 by means of the chromatographic peak area. The chromatographic peak area of 1,2-BD with the MISPME fibers was approximately 4-times higher than that of the NISPME fibers. For two reference compounds, however, the chromatographic peak areas of 1,3- and 1,4-BD extracted by MISPME fibers were comparable to those of the NISPME fibers. The results indicated that the derivatization/DLLME–MISPME procedure had high selectivity towards the target molecule.

Fig. 3 Comparison of the extraction efficiency of MISPME, NISPME and direct injection (1.00 μL) for 1,2-BD and its structural analogs (0.5 μg mL⁻¹). The small figure is a typical chromatogram. (a) Direct injection (1.00 μL) after DLLME; (b) NISPME; (c) MISPME. Peak order: (1), 1,2-BD; (2), 1,3-BD; (3), 1,4-BD.

3.4 Optimization of the derivatization/DLLME procedure

To achieve the best extraction efficiency of the derivatization/DLLME–MISPME procedure, under fixed the condition of MISPME procedure, parameters related to the derivatization/DLLME procedure, such as the type and volume of derivatization agent, the derivatization temperature/time, the type and volume of the disperser solvent/extraction solvent, pH, and solution ionic strength, were systematically studied.

3.4.1 Selection of the type and volume of derivatization reagent. It is difficult to analyze phenolic compounds using GC because the hydroxyl group interacts with the stationary phase of the GC column, resulting in poor peak shapes. Therefore, derivatization is usually required before the GC analysis of these compounds. Phenolic compounds can be transferred from aqueous solution to non-polar organic solvent after derivatization through DLLME, which is beneficial for the subsequent MISPME procedure because MIPs have poor extraction efficiency for the target molecule in an aqueous environment. Acetic anhydride quickly reacts with hydroxyl groups and has been used for the derivatization of phenolic compounds.¹⁹ Therefore, the derivatization was performed using acetic anhydride as the derivatization agent in this experiment.

To study the effect of derivatization reagent volume, the experiments were performed with different volumes (0, 200, 400, 600, 800, and 1000 μL) of acetic anhydride. Other experimental conditions were fixed. The results (Fig. 4A) showed that the peak area increased significantly in the range of 0 to 800 μL, and no obvious increase was obtained after 1000 μL; therefore, 800 μL of acetic anhydride was selected for this study.

Fig. 4 Effects of (A) derivatization reagent volume, (B) derivatization temperature, (C) derivatization time on the derivatization/DLLME–MISPME procedure for 0.5 μg mL⁻¹ 1,2-BD. Other experimental conditions were fixed.

3.4.2 Selection of the derivatization temperature/time. The effects of the derivatization temperature were investigated in the range of 10 to 90 °C (Fig. 4B). The highest peak areas for the target analytes were achieved at 50 °C. Therefore, a derivatization temperature of 50 °C was chosen for the following experiments. The influence of the derivatization time was tested in the range of 5 to 60 min (Fig. 4C). The peaks of the target analytes increased with increasing time up to 30 min. Therefore, 30 min was used for further study.

3.4.3 Selection of the type and volume of the extraction solvent. The selection of an appropriate extraction solvent is critical for a derivatization/DLLME–MISPME procedure. Tobiszewski et al. performed rankings of pairs of eight extraction and three dispersive solvents used in DLLME for chlorophenols extraction from water samples.²⁰ In our experiment, four water-immiscible organic solvents, including CH₂Cl₂, CS₂, CHCl₃ and CCl₄, were investigated as the possible extraction solvents, and the results are shown in Fig. 5A. CH₂Cl₂ exhibited the overall best extraction efficiency. Therefore, CH₂Cl₂ was used as the extraction solvent in this experiment.

Fig. 5 Effect of (A) the type of extraction solvent, (B) the volume of the extraction solvent, (C) the type of the disperser solvent, (D) the volume of the disperser solvent, (E) pH, and (F) salt addition on the derivatization/DLLME–MISPME procedure for 0.5 μg mL⁻¹ 1,2-BD. Other experimental conditions were fixed.

To study the effect of the extraction solvent volume on the performance of the presented derivatization/DLLME–MISPME procedure, different volumes of CH₂Cl₂ (0.5, 1, 1.5 and 2 mL) were tested, and the results are shown in Fig. 5B. Based on the results, the concentration of analyte in the extraction solvent decreased with increasing extraction solvent volume, resulting in decreased peak areas. Unfortunately, the MISPME fiber cannot be fully immersed in the extraction solvent when the volume of the extraction solvent is less than 1 mL. For further experiments, 1 mL of CH₂Cl₂ was chosen.

3.4.4 Selection of the type and volume of the disperser solvent. Miscibility in both the extraction solvent and aqueous phase is an essential factor in the selection of a disperser solvent. Acetone, acetonitrile, and methanol were tested using 1 mL of each disperser solvent. Fig. 5C shows that methanol resulted in the highest peak area. Methanol was therefore selected as the disperser solvent in following experiments.

To determine the optimized volume of disperser solvent, various experiments were performed using different volumes of methanol (0, 0.50, 1.0, 1.5 and 2.0 mL) as the disperser solvent. The results (Fig. 5D) showed that the peak area increased with increasing volume of methanol. However, the volume of the sediment phases decreased with increasing methanol, and the sediment phases were not sufficient for use in the following MISPME procedure when the amount of methanol was higher than 1.00 mL. Taking both the extraction efficiency and the volume of sediment phase into account, 1.00 mL methanol was chosen for further work.

3.4.5 Effect of pH/salt addition. The effect of the aqueous solution pH was examined within the range of 2 to 8 using phosphate buffer. Fig. 5E indicated that the peaks increase with the pH from 2 to 7.0 and then decrease at higher pH values. Consequently, the pH of the sample was adjusted to 7.0 to enhance the extraction efficiency. The salt addition was studied by adding NaCl in the range of 0 to 10% (w/v) to the sample solution. As shown in Fig. 5F, the analytical signals were slightly decreased with the increase of the concentration of NaCl. Hence, all experiments were performed without the addition of salt.

3.5 Optimization of the MISPME procedure

To optimize the MISPME procedure, the effects of the extraction parameters, such as the extraction temperature, extraction time, desorption temperature and desorption time, were examined.

The effect of the extraction temperature was studied by plotting the peak areas of the target analyte as a function of the extraction temperature. As shown in Fig. 6A, the extraction efficiency increased with increasing temperature up to 30 °C and decreased thereafter. Thus, 30 °C was identified as the optimal extraction temperature. The effect of the extraction time was studied in the time range from 5 to 50 min. Fig. 6B shows that the peak areas of the target analyte increased gradually with the increase of extraction time from 5 to 15 min and remained almost unchanged with further increase of time. Therefore, the extraction time was set at 15 min.

Fig. 6 Effect of (A) extraction temperature, (B) extraction time, (C) desorption temperature, and (D) desorption time on the derivatization/DLLME–MISPME procedure for 0.5 μg mL⁻¹ 1,2-BD. Other experimental conditions were fixed.

The desorption temperature and time in the GC injector were studied over the range of 100 to 350 °C and 1–5 min, respectively (Fig. 6C and D). The desorption efficiency increased with increasing temperature and reached a maximum at 300 °C. The desorption amounts increased with increasing desorption time from 1 to 3 min and then remained constant at longer times. Based on these results, 300 °C and 3 min were chosen as the optimum desorption temperature and time, respectively.

3.6 Method validation

A series of standard solutions of 1,2-BD at different concentrations were extracted using the MISPME fiber and the derivatization/DLLME–MISPME procedure under optimized conditions. The samples were then subjected to GC analysis (Table 2). A good linear relationship (r² = 0.9984) between the peak heights and the concentrations in a wide range (0.05–10 μg mL⁻¹) proved that the established method is applicable for the quantification of the analyte. The limit of detection (LOD), calculated at a signal-to-noise ratio (S/N) of 3, was 0.01 μg mL⁻¹.

Table 2 The linear equation, correlation coefficients, linear range LOD and RSD of the method

Analyte	Regression equation^a	Correlation coefficient	Linear range (μg mL^−1)	LOD (μg mL^−1)	RSD^b (%)
					Single fiber (n = 5)	Fiber to fiber (n = 5)
a x, concentration of analyte (μg mL^−1) and y, peak area.b Concentration of 0.5 μg mL^−1.
1,2-BD	y = 34.49x + 3.96	0.9984	0.05–10	0.01	5.9	10.7

---

The precision of the method was evaluated by analyzing six replicates of a standard solution of 1,2-BD at 0.5 μg mL⁻¹, the RSD of single fiber was 5.9% and the RSD of fiber to fiber was 10.7%.

3.7 Analysis of real samples

To demonstrate the applicability of the derivatization/DLLME–MISPME, the method was used to analyze 1,2-BD in real water samples obtained from Jiulong River of Zhangzhou, inner lakes in school and sewage outfall of paper mills, spiked at concentration levels of 0.1 μg mL⁻¹ by adding standard solutions of 1,2-, 1,3-, and 1,4-BD to the water samples. The recoveries and precision were obtained based on the analysis of water samples with 6 replicates. As listed in Table 3, the recoveries were between 79.8 and 110%, with RSD values between 8.5 and 10.7%. These results indicate that the MISPME fiber can be used for the selective extraction of 1,2-BD in real water samples. Typical chromatograms of the analysis of 1,2-BD in spiked water samples using the proposed method are shown in Fig. 7.

Table 3 Determination results of 1,2-BD in real samples (n = 6)

Analyte	Sample 1			Sample 2			Sample 3
	Detected (μg mL^−1)	Recovery^a (%)	RSD (%)	Detected (μg mL^−1)	Recovery (%)	RSD (%)	Detected (μg mL^−1)	Recovery (%)	RSD (%)
a Recoveries determined with the derivatization/DLLME–MISPME procedure at spiked levels of 0.1 μg mL^−1.b Not detected.
1,2-BD	ND^b	79.8	8.5	ND	110	10.7	0.086	105	10.5

---

Fig. 7 Chromatograms of 1,2-BD in (a) water samples; (b) water samples spiked at 0.1 μg mL⁻¹ of 1,2-, 1,3-, and 1,4-BD; (c) 0.1 μg mL⁻¹ of 1,2-, 1,3-, and 1,4-BD standard solution. (A) Jiulong River of Zhangzhou; (B) Inner lakes in School; (C) sewage out fall of paper mills. Peak order: (1), 1,2-BD.

4 Conclusions

To achieve the highly selective extraction and analysis of polar analytes in aqueous solution, the research had a dual focus in this paper. The first was to increase the surface area of the fiber and to improve the extraction efficiency of the MISPME fiber through the design of a layer-by-layer structure fiber. The second was to provide a novel, sensitive analytical method based on MISPME combined with DLLME with in situ derivatization to overcome the low recognition extraction efficiency problem of polar analytes in aqueous samples. The results indicated that the proposed method has some advantages, with a low detection limit and high recognition extraction efficiency for polar analytes.

Acknowledgements

This work was supported by the National Natural Scientific Foundation of China (No. 21105088) the Natural Science Foundation of Fujian Province, China (No. 2013J01062) and the Natural Science Foundation of Zhangzhou City, China (No. ZZ2016J30) which are gratefully acknowledged.

References

1. B. Jurado-Sanchez, E. Ballesteros and M. Gallego, J. Chromatogr. A, 2013, 1318, 65–71.
2. A. Sarafraz-Yazdi and A. Amiri, TrAC, Trends Anal. Chem., 2010, 29, 1–14.
3. M. I. Leong, M. R. Fuh and S. D. Huang, J. Chromatogr. A, 2014, 1335, 2–14.
4. H. Bagheri, H. Piri-Moghadam and M. Naderi, TrAC, Trends Anal. Chem., 2012, 34, 126–139.
5. R. P. Berladi and J. Pawliszyn, Water Pollut. Res. J. Can., 1989, 24, 179–191.
6. M. S. Zhang, J. B. Zeng, Y. R. Wang and X. Chen, J. Chromatogr. Sci., 2013, 51, 577–586.
7. E. H. Koster, C. Crescenzi, H. W. den, K. Ensing and G. J. de Jong, Anal. Chem., 2001, 73, 3140–3145.
8. E. Turiel, J. L. Tadeo and A. Martin-Esteban, Anal. Chem., 2007, 79, 3099–3104.
9. E. Turiel and A. Martin-Esteban, Anal. Chim. Acta, 2010, 668, 87–99.
10. A. R. Khorrami and E. Narouenezhad, Talanta, 2011, 86, 58–63.
11. F. Tan, H. X. Zhao, X. N. Li, X. Quan, J. W. Chen, X. M. Xiang and X. Zhang, J. Chromatogr. A, 2009, 1216, 5647–5654.
12. D. L. Deng, J. Y. Zhang, C. Chen, X. L. Hou, Y. Y. Su and L. Wu, J. Chromatogr. A, 2012, 1219, 195–200.
13. M. K. Y. Li, N. Y. Lei, C. B. Gong, Y. J. Yu, K. H. Lam, M. H. Lam, H. X. Yu and P. K. S. Lam, Anal. Chim. Acta, 2009, 633, 197–203.
14. D. Djozan, B. Ebrahimi, M. Mahkam and M. A. Farajzadeh, Anal. Chim. Acta, 2010, 674, 40–48.
15. A. Ameli and N. Alizadeh, Anal. Chim. Acta, 2011, 707, 62–68.
16. X. G. Hu, T. T. Ye, Y. Yu, Y. J. Cao and C. J. Guo, J. Chromatogr. A, 2011, 1218, 3935–3939.
17. Q. S. Zhong, Y. F. Hu, Y. L. Hu and G. K. Li, J. Chromatogr. A, 2012, 1241, 13–20.
18. M. S. Zhang, J. R. Huang, J. B. Zeng and C. M. Zhang, RSC Adv., 2014, 4, 12313–12320.
19. A. M. Ferreira, M. E. Laespada, J. L. Pavón and B. M. Cordero, J. Chromatogr. A, 2013, 1296, 70–83.
20. P. Bigus, J. Namiesnik and M. Tobiszewski, J. Chromatogr. A, 2016, 446, 21–26.

---