Fe3O4@SiO2@VAN Nanoadsorbent Followed by GC-MS for the Determination of Polycyclic Aromatic Hydrocarbons at Ultra-Trace Levels in Environmental Water Samples

In the present study, silica-coated magnetic nanoparticles functionalized with vancomycin (Fe3O4@SiO2@VAN) were synthesized. The Fe3O4@SiO2@VAN nanocomposite was used as a sorbent for the magnetic solid-phase extraction (MSPE) of polycyclic aromatic hydrocarbons (PAHs) from environmental water, followed by GC-MS. The nanocomposite was characterized by Fourier-transform infrared spectroscopy, X-ray diffraction, scanning electron microscopy, transmission electron microscopy, vibrating sample magnetometry, and nitrogen sorption. Various experimental parameters were optimized, including extraction condition and desorption condition. Results show that Fe3O4@SiO2@VAN combined the advantages of nanomaterials and magnetic separation technology, showing excellent dispersibility and high selectivity for PAHs in environmental water sample. Under the optimal extraction conditions, an analytical method was established with the sensitive limit of detection (LOD) of 0.03–0.16 μg L−1. The method was successfully applied for the analysis of environmental water samples. The relative standard deviations (%) were in the range of 0.50–12.82%, and the extraction recovery (%) was in the range of 82.48% and 116.32%. MSPE-coupled gas chromatography–mass spectrometry quantification of PAHs is an accurate and repeatable method for the monitoring of PAH accumulation in environmental water samples. It also provides an effective strategy for the tracing and quantification of other environmental pollutants in complex samples.


Introduction
Polycyclic aromatic hydrocarbons (PAHs) are a class of cyclic organic compounds consisting of two or more dense aromatic rings, forming a large and diverse group of organic compounds [1]. PAHs exhibit carcinogenic and teratogenic activities, and their carcinogenicity increases with the number of benzene rings [2,3]. The low water solubility and high octanol-water partition coefficient PAHs present lead to low bioavailability and high levels of accumulation in the environment, resulting in environmentally persistent, bioconcentrated, and difficult to degrade PAHs [4,5]. The European Union Water Framework Directive, the United States Environmental Protection Agency (USEPA), and the European Union (EU) have placed 16 PAHs on the list of priority persistent organic pollutants for control [6][7][8]. The US EPA and the Chinese national environmental quality standards have set recommended PAH concentrations of 0.2 and 2.0 µg L −1 for drinking water, respectively (USEPA 1993, GB5749-2006). PAHs widely exist in the environment, and this condition can lead to water and food pollution; these PAHs can also enter the human body by food chains [9]. Therefore, the sensitivity of the detection methods used for PAHs needs to be improved.
A solvothermal method was employed to prepare Fe 3 O 4 nanoparticles. In a typical synthesis procedure, FeCl 3 ·6H 2 O (2.7 g) was dissolved in ethylene glycol (50 mL) and mechanically stirred vigorously for 30 min. Subsequently, NaAc (7.2 g) and PEG-2000 (2.0 g) were added in the above mixture and stirred well. Then, the homogenized mixture was transferred into the Teflon lined stainless-steel autoclave and then heated, at 200 • C, for 8 h. The Fe 3 O 4 nanoparticles were washed for several times with EtOH and water until reaching neutrality, and then dried, at 60 • C, for 24 h.

Preparation of Core-Shell Fe 3 O 4 @SiO 2 Nanoparticles
The Fe 3 O 4 nanoparticles were coated silica shells by sol-gel co-condensation. In a typical reaction, Fe 3 O 4 (1.0 g) was dispersed in 1 mol L −1 of HCl by ultrasonic for 10 min. Afterward, deionized water (20 mL) and ammonia (2.5 mL) were added. The mixture was agitated for 30 min under vigorous mechanical stirring. Subsequently, 0.5 mL of TEOS was added into the above mixture, and then continuously stirred for 12 h. After the reaction, the Fe 3 O 4 @SiO 2 magnetic particles were obtained. It was washed with ethanol and water repeatedly and finally dried, at 60 • C, for 24 h. Fe 3 O 4 @SiO 2 nanoparticles were eventually obtained.

Synthesis of Fe 3 O 4 @SiO 2 @VAN Nanoparticles
Firstly, vancomycin (VAN, 13.82 g) and γ-glycidoxypropyltrimethoxysilane (1.88 mL) were dissolved in DMF (150 mL), and stirred, at 80 • C, for 24 h. Subsequently, Fe 3 O 4 @SiO 2 (2.5 g) was added to the above mixture then stirred, at 80 • C, for 24 h. Afterward, the obtained Fe 3 O 4 @SiO 2 @VAN nanoparticles were washed for several times with water and ethanol. The resulting Fe 3 O 4 @SiO 2 @VAN nanoparticles were dried, at 60 • C, for 24 h. The procedure for the fabrication of Fe 3 O 4 @SiO 2 @VAN magnetic adsorption material is shown in Scheme 1. mL) were dissolved in DMF (150 mL), and stirred, at 80 °C, for 24 h. Subsequently, Fe3O4@SiO2 (2.5 g) was added to the above mixture then stirred, at 80 °C, for 24 h. Afterward, the obtained Fe3O4@SiO2@VAN nanoparticles were washed for several times with water and ethanol. The resulting Fe3O4@SiO2@VAN nanoparticles were dried, at 60 °C, for 24 h. The procedure for the fabrication of Fe3O4@SiO2@VAN magnetic adsorption material is shown in Scheme 1.

Extraction Procedure
In total, 50 mg Fe3O4@SiO2@VAN nanoparticles was added to 30 mL of 100 µ g L −1 PAH solution. The mixture was stirred in a vortex with a stirrer at 250 rpm for 1 h to favor the extraction of the target analytes. Then, the sorbent was isolated from the solution by using a magnet, and then transferred into a 1.5 mL glass vial with 1 mL of n-hexane. The sorbent was sonicated for 10 min, and then filtered through a 0.22 μm organic phase membrane. The desorption solution was collected in a 1.5 mL glass vial for further GC-MS analysis.

Instrument Conditions
Chromatographic separation and identification of PAHs were carried out on a GC-MS system (Clarus 600, USA) by using a capillary column DB-5 (30 m × 0.25 mm, 0.25 μm film thickness, Agilent J&W). The carrier gas was helium at a flow rate of 1.2 mL min −1 . The injector and ion source temperatures were set at 250 and 260 °C, respectively. The initial temperature of column was set at 80 °C for 2 min, further raised to 180 °C at 20 °C min −1 and held for 2 min, and then raised to 230 °C at 3 °C min −1 and held for 2 min. Finally, the temperature was increased to 300 °C at 30 °C min −1 , and then held for 8 min. The temperature of the GC inlet was 250 °C. The solvent delay time was 5 min. The transmission line temperature was 280 °C . Mass spectrometry detection was carried out in the selected ion detection mode.

Characterization of Materials
The surface morphology and microstructure of Fe3O4, Fe3O4@SiO2, and Fe3O4@SiO2@VAN nanoparticles were observed using SEM and TEM. The SEM images of Scheme 1. Schematic procedure for Fe 3 O 4 @SiO 2 @VAN magnetic composite particles.

Extraction Procedure
In total, 50 mg Fe 3 O 4 @SiO 2 @VAN nanoparticles was added to 30 mL of 100 µg L −1 PAH solution. The mixture was stirred in a vortex with a stirrer at 250 rpm for 1 h to favor the extraction of the target analytes. Then, the sorbent was isolated from the solution by using a magnet, and then transferred into a 1.5 mL glass vial with 1 mL of n-hexane. The sorbent was sonicated for 10 min, and then filtered through a 0.22 µm organic phase membrane. The desorption solution was collected in a 1.5 mL glass vial for further GC-MS analysis.

Instrument Conditions
Chromatographic separation and identification of PAHs were carried out on a GC-MS system (Clarus 600, USA) by using a capillary column DB-5 (30 m × 0.25 mm, 0.25 µm film thickness, Agilent J&W). The carrier gas was helium at a flow rate of 1.2 mL min −1 . The injector and ion source temperatures were set at 250 and 260 • C, respectively. The initial temperature of column was set at 80 • C for 2 min, further raised to 180 • C at 20 • C min −1 and held for 2 min, and then raised to 230 • C at 3 • C min −1 and held for 2 min. Finally, the temperature was increased to 300 • C at 30 • C min −1 , and then held for 8 min. The temperature of the GC inlet was 250 • C. The solvent delay time was 5 min. The transmission line temperature was 280 • C. Mass spectrometry detection was carried out in the selected ion detection mode.

Characterization of Materials
The surface morphology and microstructure of Fe 3 O 4 , Fe 3 O 4 @SiO 2 , and Fe 3 O 4 @SiO 2 @VAN nanoparticles were observed using SEM and TEM. The SEM images of Fe 3 O 4 , Fe 3 O 4 @SiO 2 , and Fe 3 O 4 @SiO 2 @VAN nanoparticles are displayed in Figure 1A-C. It can be seen that Fe 3 O 4 nanoparticles are uniform in size, spherical in shape, and approximately 200 nm in diameter. After being coated with a thin, transparent cladding, a silica layer, a typical morphology with core/shell structure, was obtained. The dark Fe 3 O 4 @SiO 2 @VAN nanoparticles were obtained after being coated with VAN on the surface of Fe 3 O 4 @SiO 2 , and the diameter of Fe 3 O 4 @SiO 2 @VAN roughly 210-230 nm. The TEM images of Fe 3 O 4 , Fe 3 O 4 @SiO 2 , and Fe 3 O 4 @SiO 2 @VAN nanoparticles are shown in Figure 1D-F. The figure shows that the dark Fe 3 O 4 surfaces are covered by transparent silica layer alone, forming a uniform amorphous shell that can protect the core-shell of the Fe 3 O 4 magnetic nanoparticles and makes further modification more effective. The thin and transparent shell, approximately 10 nm thick, is the silica layer. After Fe 3 O 4 @SiO 2 composites were wrapped by VAN, the surfaces of the microspheres became rough. Therefore, the magnetic nanosorbent facilitates the adsorption of the target analyte, thus increasing the adsorption ability.
layer, a typical morphology with core/shell structure, was obtained. The dark Fe3O4@SiO2@VAN nanoparticles were obtained after being coated with VAN on the surface of Fe3O4@SiO2, and the diameter of Fe3O4@SiO2@VAN roughly 210-230 nm. The TEM images of Fe3O4, Fe3O4@SiO2, and Fe3O4@SiO2@VAN nanoparticles are shown in Figure  1D-F. The figure shows that the dark Fe3O4 surfaces are covered by transparent silica layer alone, forming a uniform amorphous shell that can protect the core-shell of the Fe3O4 magnetic nanoparticles and makes further modification more effective. The thin and transparent shell, approximately 10 nm thick, is the silica layer. After Fe3O4@SiO2 composites were wrapped by VAN, the surfaces of the microspheres became rough. Therefore, the magnetic nanosorbent facilitates the adsorption of the target analyte, thus increasing the adsorption ability. The FT-IR results are shown in Figure S1. The peaks at 578 and 1623 cm −1 are related to the Fe-O and hydroxyl groups on the surface of Fe3O4 nanoparticles, respectively (spectrum (a)). The absorption peaks at 462, 796, and 1101 cm −1 were observed in spectrum (b) along with corresponding bending vibration mode Si-O-Si and the presence of Si-O-Si symmetrical and asymmetrical stretching vibration [24,25]. In addition, the peak occurring around 956 cm −1 expressed the stretching vibration absorption of Fe-O-Si, confirming that the Fe3O4@SiO2 has been successfully prepared [26]. The Fe3O4@SiO2@VAN FT-IR spectrum (spectrum (c)), which formed after the Fe3O4@SiO2 NPs were coated with VAN exhibited an absorption peak at 1230 cm −1 relates to benzene that belong to VAN molecule [27]. In addition, the peak of the stretching vibration peaks at 3280, 1660, and 1047 cm −1 correspond to vancomycin, indicating that VAN successfully bonded with Fe3O4@SiO2 [28].
The phases and crystalline structures of Fe3O4, Fe3O4@SiO2, and Fe3O4@SiO2@VAN nanocomposites were confirmed by XRD nanocomposites. As illustrated in Figure 2A, the formation of Fe3O4 material was confirmed based on the peaks of 2θ at 30.28°, 35.60°, 43.26°, 53.73°, 57.28°, and 62.81°, which were assigned to the (220), (311), (400), (422), (511) and (440) of standard data of Fe3O4 diffraction peak [29][30][31], respectively. After silica shell loading with Fe3O4 nanocomposites, peak alterations were observed with the reduction in the peak intensity of Fe3O4@SiO2. Therefore, the silica shell did not change the crystal structure of Fe3O4. Moreover, a broad peak can be seen in the 2θ of 19-27° in The FT-IR results are shown in Figure S1. The peaks at 578 and 1623 cm −1 are related to the Fe-O and hydroxyl groups on the surface of Fe 3 O 4 nanoparticles, respectively (spectrum (a)). The absorption peaks at 462, 796, and 1101 cm −1 were observed in spectrum (b), along with corresponding bending vibration mode Si-O-Si and the presence of Si-O-Si symmetrical and asymmetrical stretching vibration [24,25]. In addition, the peak occurring around 956 cm −1 expressed the stretching vibration absorption of Fe-O-Si, confirming that the Fe 3 O 4 @SiO 2 has been successfully prepared [26]. The Fe 3 O 4 @SiO 2 @VAN FT-IR spectrum (spectrum (c)), which formed after the Fe 3 O 4 @SiO 2 NPs were coated with VAN, exhibited an absorption peak at 1230 cm −1 relates to benzene that belong to VAN molecule [27]. In addition, the peak of the stretching vibration peaks at 3280, 1660, and 1047 cm −1 correspond to vancomycin, indicating that VAN successfully bonded with Fe 3 O 4 @SiO 2 [28].
The phases and crystalline structures of Fe 3 O 4 , Fe 3 O 4 @SiO 2 , and Fe 3 O 4 @SiO 2 @VAN nanocomposites were confirmed by XRD nanocomposites. As illustrated in Figure 2A [29][30][31], respectively. After silica shell loading with Fe 3 O 4 nanocomposites, peak alterations were observed with the reduction in the peak intensity of Fe 3 O 4 @SiO 2 . Therefore, the silica shell did not change the crystal structure of Fe 3 O 4 . Moreover, a broad peak can be seen in the 2θ of 19-27 • in Fe 3 O 4 @SiO 2 @VAN nanoparticles pattern, which was related to VAN, indicating that the Fe 3 O 4 @SiO 2 surface was successfully coated with VAN. ence of mesoporous structure. Compared with Fe3O4@SiO2, the ratio of mesopores on the surface of Fe3O4@SiO2@VAN nanomaterials noticeably increased, which could offer more potential interaction sites and spatial match effects for the extraction of PAHs with high steric hindrance. Table S1 summarizes some textural parameters. It shows that the pore volumes of Fe3O4@SiO2@VAN (0.108 cc g −1 ) are less than that of uncoated Fe3O4 nanomaterials 0.128 (cc g −1 ), which is due to the pore blockage of the Fe3O4@SiO2@VAN surface.

Optimization of Extraction Conditions
Before the analytical method validation, the key parameters of extraction condition (time, salt addition and pH of the sample) and desorption condition (solvent, time) were optimized to obtain the best extraction efficiency. The initial experimental conditions were extraction aqueous solution concentration of 0.1 mg L −1 . The extraction aqueous solution It is critical that the magnetic nanomaterials should possess magnetic property for their rapid separation in removing process. Therefore, the magnetization curves of nanomaterials were obtained using vibrating sample magnetometer (VSM), as illustrated in Figure 2B. The saturation magnetization values (Ms) of Fe 3 O 4 , Fe 3 O 4 @SiO 2 , and Fe 3 O 4 @SiO 2 @VAN nanoparticles were 77.6, 68.3, and 44.18 emu g −1 , respectively, indicating that the magnetic response of nanomaterials reduced gradually with further modification, directly confirming the functional materials are modified on magnetic Fe 3 O 4 individually. In addition, all of these three kinds of samples show the fast magnetic response, which provides the samples with ability of magnetic separation and collection. The Fe 3 O 4 @SiO 2 @VAN nanocomposite material could be dispersed evenly in the solution matrix and quickly separated from aqueous phase by a magnet within 10 s (inset in Figure 2B).
The porous structure, pore distribution and specific surface area of as-prepared three nanomaterials were investigated using BET method. As shown in Figure 2C,D, the nitrogen adsorption-desorption isotherms can be classified as IV curves, confirming the existence of mesoporous structure. Compared with Fe 3 O 4 @SiO 2 , the ratio of mesopores on the surface of Fe 3 O 4 @SiO 2 @VAN nanomaterials noticeably increased, which could offer more potential interaction sites and spatial match effects for the extraction of PAHs with high steric hindrance. Table S1 summarizes some textural parameters. It shows that the pore volumes of Fe 3 O 4 @SiO 2 @VAN (0.108 cc g −1 ) are less than that of uncoated Fe 3 O 4 nanomaterials 0.128 (cc g −1 ), which is due to the pore blockage of the Fe 3 O 4 @SiO 2 @VAN surface.

Optimization of Extraction Conditions
Before the analytical method validation, the key parameters of extraction condition (time, salt addition and pH of the sample) and desorption condition (solvent, time) were optimized to obtain the best extraction efficiency. The initial experimental conditions were extraction aqueous solution concentration of 0.1 mg L −1 . The extraction aqueous solution and the desorbing solution volumes were 30 and 1.0 mL, respectively. The extraction and desorbing time were 30 and 10 min, respectively. All experiments were tested six times in parallel. The extraction amount (Q, µg) of analyte was calculated using Equation (1). The extraction recovery was calculated using Equation (2).
where C 0 (mg L −1 ) is the initial concentration of PAHs before extraction, C e (mg L −1 ) is the equilibrium concentration of extraction solution of PAHs after by Fe 3 O 4 @SiO 2 @VAN, V (mL) is the volume of the extraction solution of PAHs, Q 0 (µg) is the initial quantity of PAHs, and Q (µg) is the equilibrium binding quantity of PAHs on the Fe 3 O 4 @SiO 2 @VAN.

Extraction and Desorption Time Optimization
Extraction time should be sufficient to achieve the extraction equilibrium. Thus, the effect of extraction time for six PAHs was studied in the range of 2-120 min. It can be seen that the extraction efficiency increased in the first 30 min and basically stabilized afterward ( Figure S2A). This phenomenon is caused by the saturation of analytes on the sorbent sites over time. The desorption time profile on extraction performance was attained at a time duration of 2-20 min ( Figure S2B). The highest recovery obtained when the desorption time was 5 min.

Desorption Solvent, pH, and Salt Addition Optimization
To select the best desorption solvent, we investigated different eluents such as dichloromethane, n-hexane, ethanol, and acetonitrile and evaluated their extraction efficiencies. According to the results in Figure S2C, dichloromethane exhibited the best desorption performance. Therefore, dichloromethane was chosen as the best desorption solvent. In water solution, pH is an essential factor for the extraction performance by affecting the stability of the sorbent, the charge on its surface and analytes. In order to obtain the optimum pH, we studied various pH values in the range of 2.0-12.0. Since the PAHs are neutral compounds, the pH of the sample may not have a significant effect on the extraction efficiency of PAHs, but pH can affect the morphology and dispersion of the magnetic materials in the aqueous phase. Acidic and basic conditions may affect the dispersion of the sorbent in water and some leaching of Fe 3 O 4 @SiO 2 @VAN magnetic material. The best extraction efficiency was obtained at pH 7.0, and this phenomenon is due to the zwitterionic form of PAHs, which has a strong binding affinity with the cavity of VAN ( Figure S2D). In the ionic strength range of 0-25%, The influence of ionic strength was examined ( Figure S2E). The presence of NaCl would decrease the solubility of the analytes by salting-out effect and would thus be conducive to the extraction of PAHs by Fe 3 O 4 @SiO 2 @VAN. However, after the salt concentration exceeded 20%, analyte recoveries were reduced in viscosity of the solution, thus hindering the mass transfer of analytes. Therefore, 20% NaCl was added to the extraction solution.

Selectivity Evaluation
Under the optimal conditions of 100 µg L −1 concentration of extraction solution, 30 mL volume, 30 min extraction time, 1 mL volume of desorbing solution in dichloromethane, and 20% NaCl addition to the extraction solution, five substances were selected for the selectivity study of the non-structural analogues, such as bisphenol A, bisphenol B, pyridine, and phenol, and structural analogues (one of the polycyclic aromatic hydrocarbons) such as acenaphthylene. The enrichment factor (EF) and selectivity coefficient (SC) were calculated using Equations (3) and (4) as follows: where C i (µg L −1 ) is the initial concentration before extraction, and C f (µg L −1 ) is the concentration of analytes after Fe 3 O 4 @SiO 2 @VAN extraction. EF p and EF c are the enrichment factors for PAHs and its competitors. As shown in Figure 3, the magnetic sorbent material has the best extraction effect for the target analytes (weak polarity and consisting of dense aromatic rings), which all exceeded 90%, whereas no extraction effect was observed for pyridine and phenol, which are more polar and consist of one aromatic ring. For the homologue acenaphthene, due to its higher structural similarity, the vancomycin magnetic material (Fe 3 O 4 @SiO 2 @VAN) had a better extraction effect on it with an extraction recovery of 45%, but considering its shorter benzene ring chain, the extraction effect was significantly lower than that of the six target analytes. Meanwhile, we list the solubility coefficient (S w ) and octanol-water partition coefficient (logK ow ) of PAHs in Table S2. As the number of thickened benzene rings increases, the S w decreases and hydrophobicity of PAHs increases, which can cause the Fe 3 O 4 @SiO 2 @VAN material to have an excellent adsorption capacity for aromatics with multi-benzene rings due to hydrophobic interactions. Therefore, the sorbent has a good selectivity for PAHs, which can be attributed to the unique multi-ring cavity structure and hydrophobic interaction between them. For the determination of the selectivity of Fe 3 O 4 @SiO 2 @VAN, Table S3 exhibits where Ci (μg L −1 ) is the initial concentration before extraction, and Cf (μg L −1 ) is the concentration of analytes after Fe3O4@SiO2@VAN extraction. EFp and EFc are the enrichment factors for PAHs and its competitors. As shown in Figure 3, the magnetic sorbent material has the best extraction effect for the target analytes (weak polarity and consisting of dense aromatic rings), which all exceeded 90%, whereas no extraction effect was observed for pyridine and phenol, which are more polar and consist of one aromatic ring. For the homologue acenaphthene, due to its higher structural similarity, the vancomycin magnetic material (Fe3O4@SiO2@VAN) had a better extraction effect on it with an extraction recovery of 45%, but considering its shorter benzene ring chain, the extraction effect was significantly lower than that of the six target analytes. Meanwhile, we list the solubility coefficient (Sw) and octanol-water partition coefficient (logKow) of PAHs in Table S2. As the number of thickened benzene rings increases, the Sw decreases and hydrophobicity of PAHs increases, which can cause the Fe3O4@SiO2@VAN material to have an excellent adsorption capacity for aromatics with multi-benzene rings due to hydrophobic interactions. Therefore, the sorbent has a good selectivity for PAHs, which can be attributed to the unique multi-ring cavity structure and hydrophobic interaction between them. For the determination of the selectivity of Fe3O4@SiO2@VAN, Table S3 exhibits

Establishment of Analytical Methods
To further explore the application of Fe3O4@SiO2@VAN for the detection of PAHs in real samples, a matrix matching standard analytical method was established. The linear range was 0.1-200 μg L −1 . The linear equations and correlation coefficients are shown in

Establishment of Analytical Methods
To further explore the application of Fe 3 O 4 @SiO 2 @VAN for the detection of PAHs in real samples, a matrix matching standard analytical method was established. The linear range was 0.1-200 µg L −1 . The linear equations and correlation coefficients are shown in Table 1. When the signal-to-noise ratio (S/N) is 3 and 10, respectively, the limit of detection (LOD) and limit of quantification (LOQ) of PAHs are obtained. The resulting LODs were in the range of 0.03-0.16 µg L −1 , and LOQs were in the range of 0.090-0.48 µg L −1 . The US EPA and Chinese national environmental quality standards have set the recommended PAH concentrations to 0.2 and 2.0 µg L −1 for drinking water, respectively (USEPA 1993, GB5749-2006), which are higher than the highest LOD (0.16 µg L −1 ) obtained from our proposed method. Therefore, the proposed method can meet the detection of allowable PAHs content in water samples.

Detection of Actual Samples
Over the past several decades, the rapid population growth and economic development in the Dianchi Lake basin have resulted in the discharge of many contaminants into the lake, and PAHs have become the main pollutant. Accordingly, the Kunming government has spent enormous human and financial resources to rectify the Dianchi Lake by building Laoyuhe constructed wetlands. Plants play an important role in the biogeochemical cycle of environmental pollutants. PAHs in the rhizosphere soil have strong bioaccumulation potential for root absorption. Therefore, Metasequoia glyptostroboides is used for the treatment of polluted Dianchi Lake because of its strong root system. The overall structure of the artificial wetland in the Laoyuhe River is shown in Figure 4, Urban wastewater flows from point 1 into the Luyu River Park, is degraded by M. glyptostroboides when it flows through point 2, and then flows into Dianchi at point 3.
To evaluate the above analytical method for genuine samples by using Fe 3 O 4 @SiO 2 @VAN as sorbents for MSPE in combination with GC-MS for detection, we obtained environmental water samples from three different points of the Laoyuhe River in Kunming City (Figure 4).
To eliminate possible matrix effects, a standard addition method was adopted for the quantitative determination of PAHs. Three aliquots of the lake sample were analyzed in parallel. The results showed that only pyrene was found in three sampling points with different concentrations. The concentrations of pyrene from the inlet to the outlet are 0.35, 0.26, and 0.14 µg L −1 , respectively. The gradual decrease in PAH concentration from the inlet to the outlet in Laoyuhe River suggests that M. glyptostroboides is an effective pathway for perennial tree species for the removal of PAHs absorbed into them, and the Yunnan government has achieved significant and positive results in the management of Dianchi Lake. Furthermore, the analysis of PAHs in environmental water samples has a good recovery, with the range of 82.48-116.32%, and the RSD% (n = 5) does not exceed 12.82% ( Table 2). The chromatogram of sample is displayed in Figure S3.

Comparison with Other Reported Methods
The developed method by using Fe 3 O 4 @SiO 2 @VAN as MSPE adsorbent material was compared with previously reported methods (Table 3). Results show that the developed approach for PAHs analysis was more sensitive than those in previously reported methods. The proposed method exhibited low LODs, eco-friendliness, and anti-matrix interference for monitoring of target PAHs. However, the new approach showed the longest extraction time values. This fact can be ascribed to the use of extraction method. Although sonication can accelerate the adsorption of analytes to the material, it may also cause irreversible damage to the material. Therefore, we choose a gentle method (stirring) for the adsorption of analytes to the material.

Conclusions
In this research, we proposed an easy-to-operate, high sensitivity, and anti-matrix interference MSPE method based on a functionalized magnetic nanoadsorbent (Fe 3 O 4 @SiO 2 @VAN) followed by GC-MS for the preconcentration and detection of PAHs in natural water samples. The synthesized novel magnetic material exhibit nano-and mesoporous structures, as well as superparamagnetic and excellent hydrophobic properties. High selectivity for the target analytes and good dispersibility in hydrophilic matrices were significant advantages of the nanoadsorbent. Under the most favorable extraction conditions, wide linear range, low LODs, good RSDs, and high relative recovery were obtained. Finally, this method was applied for determining PAHs in environmental water samples, and satisfactory results were achieved. These remarkable features provide the great potential of Fe 3 O 4 @SiO 2 @VAN for the extraction of other hydrophobic analytes from water samples.  Table S1: Specific surface area and pore structure parameters of Fe 3 O 4 , Fe 3 O 4 @SiO 2 and Fe 3 O 4 @SiO 2 @VAN; Table S2: Physico-chemical characteristics of nine PAHs listed by USEPA, Table S3: Selectivity adsorption parameters of Fe 3 O 4 @SiO 2 @VAN for PAHs.