Simultaneous Determination of Nine Quinolones in Pure Milk Using PFSPE-HPLC-MS/MS with PS-PAN Nanofibers as a Sorbent

In this study, a packed-fiber solid-phase extraction (PFSPE)-based method was developed to simultaneously detect nine quinolones, including enrofloxacin (ENR), ciprofloxacin (CIP), ofloxacin (OFL), pefloxacin (PEF), lomefloxacin (LOM), norfloxacin (NOR), sarafloxacin (SAR), danofloxacin (DAN), and difloxacin (DIF), in pure milk, using high-performance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS). Polystyrene (PS) and polyacrylonitrile (PAN) were combined to form PS-PAN composite nanofibers through electrospinning. The nanofibers were used to prepare the home-made extraction columns, and the process was optimized and validated using blank pure milk. The analytical method showed high accuracy, and the recoveries were 88.68–97.63%. Intra-day and inter-day relative standard deviations were in the ranges of 1.11–6.77% and 2.26–7.17%, respectively. In addition, the developed method showed good linearity (R2 ≥ 0.995) and low method quantification limits for the nine quinolones (between 1.0–100 ng/mL) for all samples studied. The nine quinolones in the complex matrix were directly extracted using 4.0 mg of PS-PAN composite nanofibers as a sorbent and completely eluted in 100 μL elution solvent. Therefore, the developed PFSPE-HPLC-MS/MS is a sensitive and cost-effective technique that can effectively detect and control nine quinolones in dairy products.


Introduction
Milk is a nutritious liquid food produced by animals that is a key nutritional source for humans. Milk production also significantly contributes to the global economy. Antibiotic treatments protect animals from infections such as mastitis [1]. However, inappropriate use of antibiotics such as quinolones can deteriorate the quality of milk and other dairy products. Since quinolones are effectively used to resist Gram-positive and Gram-negative bacteria, they are widely applied for therapeutic purposes [2]. Animal products may contain some fractions of antibiotics due to excessive antibiotic treatment. The antibiotic remaining in the animal products can potentially have severe consequences on human health via the food chain [3]. Thus, monitoring the residues of quinolones in dairy products is vital to human health. The control of veterinary drug residues is an important measure for ensuring consumer protection [4]. Therefore, an efficient and sensitive method for detecting quinolones should be developed and validated to monitor the residues in food, ensuring the healthy consumption of dairy products.
Single stock solutions of 0.1 mg/mL of the nine quinolones, namely, ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, and DIF in methanol were separately prepared. The working mixture solutions were prepared via appropriate dilution of the stock solutions with 1.0 ng/mL, 5.0 ng/mL, 20.0 ng/mL, 50.0 ng/mL, and 100.0 ng/mL of pure water at 4 • C in a volumetric flask.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 µm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 • C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 µL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Equipment and HPLC-MS-MS Conditions
The separation of the nine quinolones from the pure milk extracts was performed using an HPLC system (Agilent 1200 Series HPLC system, Agilent Technologies, Santa Clara, CA, USA) placed in an Anhui Guoke Testing Technology Co., Ltd. (www.guoketest.com.cn (accessed on: 11 May 2022)) system consisting of a vacuum degasser, an autosampler, and a binary pump, equipped with a reversed-phase SB-C18 analytical column (2.1 × 100 mm, 2.7 μm particle size, Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of 0.1% of formic acid in water (A) and acetonitrile (B). Gradient conditions such as 0-1.0 min (95% A), 1.1-4.0 min (2% A), and 4.1-6.0 min (95% A) were set up, and the flow rate was maintained at 0.3 mL/min.
The mass spectrometric acquisition was performed on an Agilent 6460 triple quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA) equipped with an electrospray ionization source to generate positive ions [M + H] + . In order to achieve the best ionization of the nine quinolones, an ion spray voltage of 4.0 kV in the positive mode was employed. The nebulizing gas was nitrogen (99.999%), and the rate was 7.0 L/min. The temperature of the symmetrical heaters was 300 °C, and the flow rate of the sheath gas was 11 L/min. Finally, the injection volume was 2 μL. The retention time, precursor ion, ionic product, and collision energy for each analyte are listed in Table 1.

Preparation of PS-PAN Nanofibers
PS-PAN nanofibers were fabricated as follows. First, PAN (15%, w/v) solution was prepared by dissolving an appropriate amount of PAN in the DMF solvent to form a polymer solution. PS (5%, w/v) was added to the polymer solution after the PAN was completely dissolved. The mixed electrospinning solution was stirred at 6000 rpm at an ambient temperature of 23~25 °C. Then, it was loaded into the glass syringe, which was equipped with a steel needle with a tip diameter of 0.33 mm. Finally, the conditions for electrospinning were as follows: an anodic voltage of 22 kV, a distance between the tip of the needle and the collector of 15 cm, and a mixed electrospinning solution flow rate of 2.0 mL/h. The nanofibers were collected on the collector and dried at room temperature.

Characterization
The surface morphologies of the electrospun fibers were observed using a scanning electron microscope (SEM, Supra55, ZEISS, Jena, Germany) at 20 kV. The average diameter of the electrospun fibers was determined by analyzing 30 single fibers using the Nano Measurer software, version 1.2 (Nanjing, China). The molecular structure of the nanofiber membrane was confirmed using ATR-IR spectroscopy (FTIR, NicoletiS10, Thermo Fisher, Waltham, MA, USA), carried out in the 4000-500 cm −1 range. The X-ray diffraction analysis was carried out using an X-ray diffractometer (XRD, XD-3X, Persee General, Beijing, China), and the XRD spectra were recorded over a 2θ range of 5° to 60°, with a step of 0.02°.

Sample Pretreatment
Pure milk (150 mL, Yili) was purchased from a local supermarket in Chuzhou, China. According to the method described in the literature, one sample was used as the blank sample without quinolones [22]. The preparation of the milk sample included the following steps: (i) 2.0 mL milk was added into a 50 mL centrifuge tube; (ii) nine quinolones (ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, and DIF) with appropriate concentrations (2 ng/mL, 10 ng/mL, and 25 ng/mL) were added to the sample; (iii) 10.0 mL of EDTA-McIlvaine buffer solution (pH 4.0) was added to the mixed solution; (iv) the mixture was sonicated for 20 min and centrifugated at 10,000 rpm for 30 min before the supernatant was separated; and (v) the total supernatant was transferred to a clean tube and stored at 4 °C before the pretreatment using PFSPE.
To effectively implement the extraction, 4.0 mg of nanofibers was packed into the PFSPE column at the tip end of the storage cartridge [23]. Then, the column was activated using 200 μL of methanol and 200 μL of deionized water. A 500 mL volume of supernatant was loaded into the gas-tight plastic syringe and then pushed through the sorbent under atmospheric pressure. Targets adsorbed on the nanofibers were eluted using 100 μL of eluent, as shown in Figure 1. Hence, the extraction recovery was calculated following Equation (1): where A0 is the peak area of 10 μL of the nine quinolones standard solution and A1 is the peak area of 10 μL of the nine quinolones eluent solution.

Preparation of PS-PAN Nanofibers
PS-PAN nanofibers were fabricated as follows. First, PAN (15%, w/v) solution was prepared by dissolving an appropriate amount of PAN in the DMF solvent to form a polymer solution. PS (5%, w/v) was added to the polymer solution after the PAN was completely dissolved. The mixed electrospinning solution was stirred at 6000 rpm at an ambient temperature of 23~25 • C. Then, it was loaded into the glass syringe, which was equipped with a steel needle with a tip diameter of 0.33 mm. Finally, the conditions for electrospinning were as follows: an anodic voltage of 22 kV, a distance between the tip of the needle and the collector of 15 cm, and a mixed electrospinning solution flow rate of 2.0 mL/h. The nanofibers were collected on the collector and dried at room temperature.

Characterization
The surface morphologies of the electrospun fibers were observed using a scanning electron microscope (SEM, Supra55, ZEISS, Jena, Germany) at 20 kV. The average diameter of the electrospun fibers was determined by analyzing 30 single fibers using the Nano Measurer software, version 1.2 (Nanjing, China). The molecular structure of the nanofiber membrane was confirmed using ATR-IR spectroscopy (FTIR, NicoletiS10, Thermo Fisher, Waltham, MA, USA), carried out in the 4000-500 cm −1 range. The X-ray diffraction analysis was carried out using an X-ray diffractometer (XRD, XD-3X, Persee General, Beijing, China), and the XRD spectra were recorded over a 2θ range of 5 • to 60 • , with a step of 0.02 • .

Sample Pretreatment
Pure milk (150 mL, Yili) was purchased from a local supermarket in Chuzhou, China. According to the method described in the literature, one sample was used as the blank sample without quinolones [22]. The preparation of the milk sample included the following steps: (i) 2.0 mL milk was added into a 50 mL centrifuge tube; (ii) nine quinolones (ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, and DIF) with appropriate concentrations (2 ng/mL, 10 ng/mL, and 25 ng/mL) were added to the sample; (iii) 10.0 mL of EDTA-McIlvaine buffer solution (pH 4.0) was added to the mixed solution; (iv) the mixture was sonicated for 20 min and centrifugated at 10,000 rpm for 30 min before the supernatant was separated; and (v) the total supernatant was transferred to a clean tube and stored at 4 • C before the pretreatment using PFSPE.
To effectively implement the extraction, 4.0 mg of nanofibers was packed into the PFSPE column at the tip end of the storage cartridge [23]. Then, the column was activated using 200 µL of methanol and 200 µL of deionized water. A 500 mL volume of supernatant was loaded into the gas-tight plastic syringe and then pushed through the sorbent under atmospheric pressure. Targets adsorbed on the nanofibers were eluted using 100 µL of eluent, as shown in Figure 1. Hence, the extraction recovery was calculated following Equation (1): where A 0 is the peak area of 10 µL of the nine quinolones standard solution and A 1 is the peak area of 10 µL of the nine quinolones eluent solution.

Method Validation
Different standard solutions were prepared and verified through PFSPE-HPLC MS/MS to validate the linearity, precision, detection limit (LOD), and quantification (LOQ) limit under optimal conditions. The linearity was determined by analyzing stand ard solutions ranging from 1.0 to 100 ng/mL in blank pure milk. The standard solution with the lowest concentration (2.0 ng/mL) of the nine quinolones was analyzed to obtain the LOD and LOQ, which were calculated for signal-to-noise ratios (S/N) of 3 and 10, re spectively. The relative standard deviation (% RSD) was used as the precision and calcu lated according to the reproducibility (intra-day precision, n = 3 and inter-day precision n = 6). In this study, the matrix-matched slope divided by the solvent-based slope equal the ME, and the formula is: ((slope matrix − matched/slope solvent) × 100).

Characterization of Nanofibers
Compared with PAN nanofibers, no significant morphological change was observed in the SEM images of PAN-PS nanofibers (Figure 2a,b). In addition, the rough surface o the PAN-PS nanofibers, compared with PAN nanofibers, remained unchanged. The di

Method Validation
Different standard solutions were prepared and verified through PFSPE-HPLC-MS/MS to validate the linearity, precision, detection limit (LOD), and quantification (LOQ) limit under optimal conditions. The linearity was determined by analyzing standard solutions ranging from 1.0 to 100 ng/mL in blank pure milk. The standard solution with the lowest concentration (2.0 ng/mL) of the nine quinolones was analyzed to obtain the LOD and LOQ, which were calculated for signal-to-noise ratios (S/N) of 3 and 10, respectively. The relative standard deviation (% RSD) was used as the precision and calculated according to the reproducibility (intra-day precision, n = 3 and inter-day precision, n = 6). In this study, the matrix-matched slope divided by the solvent-based slope equals the ME, and the formula is: ((slope matrix − matched/slope solvent) × 100).

Characterization of Nanofibers
Compared with PAN nanofibers, no significant morphological change was observed in the SEM images of PAN-PS nanofibers (Figure 2a,b). In addition, the rough surface of the

Molecular Interactions of Nanofibers
The Fourier transform infrared spectrometry (FTIR) spectra of the PAN and PAN/PS nanofibers are shown in Figure 3

Molecular Interactions of Nanofibers
The Fourier transform infrared spectrometry (FTIR) spectra of the PAN and PAN/PS nanofibers are shown in Figure 3. The characteristic peaks of PAN nanofibers were composed of conjugated C−N stretching at 2242 cm −1 and −CH 2 − stretching vibrations at 1450 cm −1 and 2491 cm −1 . After blending the PS, as shown in the spectrum for PAN/PS nanofibers, many new peaks appeared in the IR spectra of the PAN/PS nanofibers. The new peaks at 698 cm −1 and 756 cm −1 were absorption peaks due to C−H vibrations in the PS benzene ring. The above results indicate that the combination of PAN and PS does not produce new chemical bonds but exists as PAN/PS composite nanofibers.

X-ray Diffraction Analysis (XRD)
X-ray diffractograms of the PAN nanofibers, PS nanofibers, and PAN/PS nanofibers are shown in Figure 4. The broad diffraction peaks at 2θ = 10° and 20° are the diffraction peaks of PS nanofibers. The sharp diffraction peaks at 2θ = 16° and 26° indicate the highly crystalline nature of PAN nanofibers. The XRD pattern of PAN/PS nanofiber membranes exhibited a similar pattern to the PAN nanofibers and PS nanofibers. Sharp diffraction peaks were observed at 2θ = 16° and 26°. The broad peak at 2θ = 10°, corresponding to PS, did not appear in the XRD pattern of PAN/PS nanofibers. Moreover, the lower intensity of the peaks at 2θ = 16° and 26° suggested that the overall crystallinity of the PAN/PS nanofibers was slightly lower than that of PS but higher than that of PAN nanofibers. The results showed that pure PAN and PS particles were physically mixed and generated strong molecular interactions.

X-ray Diffraction Analysis (XRD)
X-ray diffractograms of the PAN nanofibers, PS nanofibers, and PAN/PS nanofibers are shown in Figure 4. The broad diffraction peaks at 2θ = 10 • and 20 • are the diffraction peaks of PS nanofibers. The sharp diffraction peaks at 2θ = 16 • and 26 • indicate the highly crystalline nature of PAN nanofibers. The XRD pattern of PAN/PS nanofiber membranes exhibited a similar pattern to the PAN nanofibers and PS nanofibers. Sharp diffraction peaks were observed at 2θ = 16 • and 26 • . The broad peak at 2θ = 10 • , corresponding to PS, did not appear in the XRD pattern of PAN/PS nanofibers. Moreover, the lower intensity of the peaks at 2θ = 16 • and 26 • suggested that the overall crystallinity of the PAN/PS nanofibers was slightly lower than that of PS but higher than that of PAN nanofibers. The results showed that pure PAN and PS particles were physically mixed and generated strong molecular interactions.

X-ray Diffraction Analysis (XRD)
X-ray diffractograms of the PAN nanofibers, PS nanofibers, and PAN/PS nanofibers are shown in Figure 4. The broad diffraction peaks at 2θ = 10° and 20° are the diffraction peaks of PS nanofibers. The sharp diffraction peaks at 2θ = 16° and 26° indicate the highly crystalline nature of PAN nanofibers. The XRD pattern of PAN/PS nanofiber membranes exhibited a similar pattern to the PAN nanofibers and PS nanofibers. Sharp diffraction peaks were observed at 2θ = 16° and 26°. The broad peak at 2θ = 10°, corresponding to PS, did not appear in the XRD pattern of PAN/PS nanofibers. Moreover, the lower intensity of the peaks at 2θ = 16° and 26° suggested that the overall crystallinity of the PAN/PS nanofibers was slightly lower than that of PS but higher than that of PAN nanofibers. The results showed that pure PAN and PS particles were physically mixed and generated strong molecular interactions.

Comparison of Adsorption Efficiencies between PAN Nanofibers and PAN/PS Composite Nanofibers
First, the effect of the nanofiber characteristics on the efficiency of extraction was investigated. In this study, the extraction recoveries of the two types of nanofibers prepared with different frameworks were evaluated. The extraction recoveries of PAN/PS composite nanofibers for nine quinolones (ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, and DIF) were 97.5%, 99.2%, 93.2%, 98.8%, 93.2%, 98.3%, 87.6%, 98.6%, and 89.1%, respectively, and these values were much higher than those for pure PAN nanofibers. The target analytes were difficult to detect due to the polymeric nature of PAN and the existence of strong chemical bonds between its molecules. Moreover, the result could be explained using the interaction mechanism between PAN-PS and the analytes, due to the conjugated π-π structure in the PS backbone, which showed a strong interaction between the analytes. A hydrogen bonding interaction could also exist between the polymer and analytes.

Comparison of Adsorption Efficiencies between PAN Nanofibers and PAN/PS Composite Nanofibers
First, the effect of the nanofiber characteristics on the efficiency of extraction was vestigated. In this study, the extraction recoveries of the two types of nanofibers prepa with different frameworks were evaluated. The extraction recoveries of PAN/PS com site nanofibers for nine quinolones (ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, a DIF) were 97.5%, 99.2%, 93.2%, 98.8%, 93.2%, 98.3%, 87.6%, 98.6%, and 89.1%, respective and these values were much higher than those for pure PAN nanofibers. The target a lytes were difficult to detect due to the polymeric nature of PAN and the existence strong chemical bonds between its molecules. Moreover, the result could be explain using the interaction mechanism between PAN-PS and the analytes, due to the conjuga π-π structure in the PS backbone, which showed a strong interaction between the a lytes. A hydrogen bonding interaction could also exist between the polymer and analy After extraction, quinolones were desorbed from the sorbents using eluent prior HPLC-MS/MS analysis. To determine the optimal elution solvent, 100 μL of 70% (v methanol, 80% (v/v) methanol, 90% (v/v) methanol, and 1% (v/v) sulfuric acid/70% me anol were evaluated (Figure 5b). Low recoveries were observed when the various conc trations of methanol were used, but an increase in recovery was observed after 1% sulfu acid was added (Figure 5c). Compared with the desorption ability of 70% methanol, formic acid/70% methanol exhibited higher recovery rates (82.4-100.7%). The amount of PAN-PS nanofibers played a vital role in the recovery of the n quinolones. If there are insufficient nanofibers, the target analytes cannot be complet adsorbed; however, excessive nanofibers may lead to incomplete elution. As shown Figure 5d, as the amount of nanofibers increased from 3.0 mg to 4.0 mg, the recovery the nine targets increased significantly. However, for amounts of nanofibers ranging fr 5.0 mg to 8.0 mg, the elution rates of the nine targets decreased significantly. Therefo 4.0 mg of 15% polyacrylonitrile and 5% polystyrene composite nanofibers was selec The amount of PAN-PS nanofibers played a vital role in the recovery of the nine quinolones. If there are insufficient nanofibers, the target analytes cannot be completely adsorbed; however, excessive nanofibers may lead to incomplete elution. As shown in Figure 5d, as the amount of nanofibers increased from 3.0 mg to 4.0 mg, the recovery of the nine targets increased significantly. However, for amounts of nanofibers ranging from Foods 2022, 11, 1843 9 of 13 5.0 mg to 8.0 mg, the elution rates of the nine targets decreased significantly. Therefore, 4.0 mg of 15% polyacrylonitrile and 5% polystyrene composite nanofibers was selected for solid-phase extraction in this study.
To explore the effect of salt addition on the extraction of the nine quinolones, different concentrations (0%, 2%, 4%, 6%, 8%, and 10.0% w/v of the analytes) were added to the samples. The addition of salt to the samples reduced the solubility of the analytes and the amount of the analytes extracted [24]. Figure 5e shows that there were no significant changes in extraction efficiency. Hence, adjustment of the ionic strength was not necessary in this study.
To exhibit the reproducibility of the synthesis of sorbent, the extraction recoveries of the nine quinolones were tested. As shown in Figure 5f, the extraction recoveries were almost the same under the same conditions of synthesis and pretreatment of the nine quinolones. The results indicate that the PAN/PS composite nanofibers exhibited good chemical and mechanical stability and reproducibility.

Method Validation
The reproducibility, linearity, and LOD were investigated under optimal conditions. Milk samples were randomly obtained from a local market in Chuzhou (China). According to the national standard method, one of the milk samples without the nine quinolones was selected as a blank for calibration and validation.
Under optimum conditions, everyday snack samples spiked with the nine quinolones were employed for analyzing the methodological parameters. As shown in Table 2, good linearity was found in the range of 1.0−100 ng/mL for the nine quinolones. The coefficients of correlation (R 2 ) ranged from 0.9992 to 0.9999. The LODs of all analytes were in the range of 0.16-0.39 ng/mL, and the LOQs were in the range of 0.53−1.29 ng/mL, indicating that the method can effectively determine the nine quinolones in plain milk samples. As shown in Tables 2 and 3, a high efficiency of extraction was found for all analytes in the pure milk sample. Recoveries were obtained over a reasonable range from 88.68% to 97.63% for the nine analytes in the spiked concentration samples. The intra-day range for the three spiked concentration samples was from 1.11% to 6.77%, and the inter-day RSD was in the range 2.26−7.17%. All these values are below 12%.

Matrix Effect
The extraction and analysis of the nine quinolones in pure milk could be affected by other factors. If the value of the matrix effect (ME) is 100%, this shows that there is no interference from the matrix; however, if the value is more than 100%, it shows that the ME enhances the analysis signal. Finally, if the value is less than 100%, the ME reduces the signal [25]. The results for the ME were evaluated and are summarized in Table 4. Although the ME values of the nine analytes were less than 100%, they all ranged from 94.2% to 98.3%. Therefore, the effect of matrix interference on the analysis signal of pure milk was insignificant.

Comparison of This Method with Other Measurement Methods
The advantages of the developed method for detecting the nine quinolones in plain milk are summarized in Table 5. The developed method was compared with various preparation methods in the existing literature. Firstly, the fabricated PS-PAN nanofibers were used to analyze the nine quinolones in plain milk for the first time. Secondly, the sample pretreatment required no evaporative drying with nitrogen or treatment of the filter membrane. Generally, the procedure is simplified and easy to master.

Application to Real Samples
The samples were then spiked with ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, and DIF standards at 50 ng/mL levels, to assess the matrix effects. Nine quinolones were detected in milk samples, as shown in Figure 6. This results show that the developed method can effectively detect quinolones in samples where they cannot be detected using the standard method.

Conclusions
The PFSPE-based PAN-PS composite nanofibers were used as sorbents to simultaneously extract and analyze nine quinolones (ENR, CIP, OFL, PEF, LOM, NOR, SAR, DAN, and DIF) in pure milk, using HPLC-MS/MS. After the optimization of the process, the method showed good linearity, precision, LOD, and LOQ, with high accuracy, and low ME values. The analytical method showed high accuracy, with recoveries of 88.68−97.63%. Intra-day and inter-day relative standard deviations were in the ranges of 1.11−6.77% and 2.26−7.17%, respectively. In addition, the developed method showed good linearity (R 2 ≥ 0.995) and low method quantification limits for the nine quinolones (between 1.0−100 ng/mL) in all samples studied. Moreover, compared with the sample pretreatment in conventional methods, the sample pretreatment in the developed method required no evaporative drying under nitrogen or filter membrane treatment. The results show that PFSPE-HPLC−MS/MS is a sensitive and cost−effective technique for detecting nine quinolones in plain milk. However, there are also some limitations. For example, the column packing is not automated. If it could be automated, the consistency of detection would be improved considerably.