Determination of Tetracycline Antibiotic Residues in Honey and Milk

---

Article pubs.acs.org/JAFC

Determination of Tetracycline Antibiotic Residues in Honey and Milk by Miniaturized Solid Phase Extraction Using Chitosan-Modified Graphitized Multiwalled Carbon Nanotubes Jing-Jing Xu,† Mingrui An,§ Rui Yang,§ Zhijing Tan,§ Jie Hao,⊗ Jun Cao,*,† Li-Qing Peng,† and Wan Cao† †

College of Material Chemistry and Chemical Engineering, Hangzhou Normal University, Hangzhou 310036, China Department of Surgery, University of Michigan Medical Center, Ann Arbor, Michigan 48109, United States ⊗ Periodontics and Oral Medicine, School of Dentistry, University of Michigan, Ann Arbor, Michigan 48109, United States §

S Supporting Information *

ABSTRACT: A rapid, simple, and strongly selective miniaturized solid phase extraction (SPE) technique, requiring only small amounts of sorbent (24 mg) and elution solvent (600 μL), coupled with ultrahigh-performance liquid chromatography and quadrupole time-of-flight mass spectrometry was developed for detecting tetracycline antibiotics. These analytes were extracted from honey and milk using chitosan-modified graphitized multiwalled carbon nanotubes (G-MWNTs) as the solid sorbent and acetonitrile/acetic acid (8:2, v/v) as the eluent in miniaturized SPE. Under the optimum experimental conditions, a satisfactory linearity (r2 > 0.992) was obtained, and the limits of detection were in the range of 0.61−10.34 μg/kg for the analytes. The mean recoveries of the five tetracycline antibiotic residues in the real samples were between 81.5 and 101.4%. The results demonstrated that chitosan-modified G-MWNTs comprise a promising material for the enrichment of tetracycline antibiotics from complex food matrices. KEYWORDS: chitosan, graphitized multiwalled carbon nanotubes, miniaturized solid phase extraction, tetracycline antibiotics, ultrahigh-performance liquid chromatography, quadrupole time-of-flight mass spectrometry

■

INTRODUCTION Following the discovery of the carbon nanostructure (C60fullerene),1 carbon nanotubes (CNTs) found by Ijima in 19912 have been used in many analytical fields due to their extraordinary mechanical, thermal, electric, magnetic, chemical, and optical properties as well as their immense surface area. They can be mainly divided into single-walled carbon nanotubes (SWNTs), double-walled carbon nanotubes, and multiwalled carbon nanotubes (MWNTs) according to the number of graphene layers.3 Among them, graphitized multiwalled carbon nanotubes (G-MWNTs) have attracted special interest in various disciplines, such as dye-sensitized solar cells,4 the pseudostationary phase in electrokinetic chromatography,5 durable cathode-catalyst supports for PEFCs,6 the removal of carcinogenic dyes,7 dispersive cleanup of acetonitrile extracts,8 and oxygen-reduction catalysts.9 Recently, the functionalization of CNTs with polymers has generated remarkable interest because of their potential properties and applications.10−14 One of the polymers that is utilized for the functionalization of carbon nanotubes is chitosan.11,12,15,16 To date, no reports on the use of chitosan for the functionalization of G-MWNTs have been published. Solid phase extraction (SPE) has now widely replaced liquid−liquid extraction (LLE) as the method of choice for relevant analytical processes.17 In recent years, with the aim of reducing the operational cost, miniaturized SPE was developed as an effective and rapid technique. Compared to traditional SPE,18,19 this sample preparation technique presents several advantages including simplicity, rapidity, a small volume of elution solvent, and a low consumption of sorbents, making the © 2016 American Chemical Society

entire procedure inexpensive and not time-consuming. For miniaturized SPE, careful selection of the sorbent material is required, which determines the selectivity and preconcentration efficiency for target analytes in different samples.20,21 In recent years, many sorbent materials have been studied in prior studies, such as mesoporous molecular sieves,21 strata-SAX and strata-XC,22 activated carbon,23 multiwalled carbon nanotubes,24,25 molecularly imprinted polymers (MIP),26 pyridinefunctionalized magnetic nanoporous silica material (PyFe3O4@MCM-41),27 and C18.17 However, to our knowledge, the utilization of chitosan-grafted G-MWNTs as the sorbent for miniaturized SPE has not appeared in the literature. Tetracycline antibiotics are broad-spectrum medicinal drug compounds that are active against a number of Gram-positive and Gram-negative bacteria. 28 Tetracyclines have been successfully applied worldwide in veterinary and human medicine for the treatment and prevention of microbial infections and as additives in animal foodstuffs. The most widely used compounds within this antibiotic group are tetracycline, 1; oxytetracycline, 2; chlortetracycline, 3; metacycline, 4; and doxycycline, 5 (Figure 1). In addition, tetracycline antibiotics can be directly sprayed onto plant leaves and seeds to treat and control infection by Mycoplasma, Erwinia amylovara, and Xanthomonas campestris.27 However, the occurrence of these antibiotic residues in foodstuff samples Received: Revised: Accepted: Published: 2647

February 15, 2016 March 10, 2016 March 12, 2016 March 12, 2016 DOI: 10.1021/acs.jafc.6b00748 J. Agric. Food Chem. 2016, 64, 2647−2654

Article

Journal of Agricultural and Food Chemistry

Figure 1. Chemical structures of the five analytes. (HPLC grade) were provided by Sigma-Aldrich Shanghai Trading Co., Ltd. Double-distilled water (Wahaha Group Co., Ltd., Hangzhou, China) was employed throughout the experiment. The tested standards, including 1−5 (Figure 1), were obtained from Shanghai Winherb Medical Technology Co., Ltd. (Shanghai, China). The purity of each standard was >98%, and their chemical structures are given in Figure 1. The stock standard solutions were prepared in chromatographic grade methanol at the concentration level of 50 μg/mL and stored at 4 °C before use. The working solutions were prepared daily by the subsequent dilution of the stock standard solution in doubledistilled water. Acacia honey, orange blossom honey, and pure milk were purchased from a local supermarket (Hangzhou, China). Apparatus and Chromatographic Conditions. The surface morphologies of the sorbents were investigated using an HT7700 scanning electron microscope (SEM) (Hitachi, Tokyo, Japan). Transmission electron microscopy (TEM) measurements were performed on a Supra55 microscope (Zeiss, Oberkochen, Germany) with an acceleration voltage of 100 kV. The analysis was performed on a model 1290 ultrahigh-performance liquid chromatography system (Agilent Technologies, Santa Clara, CA, USA) coupled with a 6530 QTOF mass spectrometer (Agilent) via an electrospray ionization (ESI) source working in the positive ion mode. The UHPLC instrument was equipped with a binary solvent delivery pump, an autosampler, a thermostated column compartment. The column used was a 50 mm × 4.6 mm i.d., 5 μm, RP SB-C18 at a column temperature of 40 °C (Agilent Technologies). The mobile phases consisted of aqueous solutions containing 0.1% (v/v) formic acid (A) and methanol (B), and a gradient elution of 30−30% B from 0 to 2 min, 30−40% B from 2 to 3 min, 40−50% B from 3 to 4 min, 50−75% B from 4 to 7 min, and 75−100% B from 7 to 8 min was used. The injection volume was 2 μL at a flow rate of 0.4 mL/min. High-purity nitrogen (N2) was used as the nebulizing and drying gas. The optimum parameters were as follows: capillary voltage, 3500 V; drying gas temperature, 350 °C; drying gas flow, 12 L/min; nebulizer pressure, 45 psig; fragmentor, 170 V; OCT/RF, 750 V; skimmer voltage, 65 V; and collision energy, 10− 25 V. All of the data were acquired and processed by Mass Hunter software (version B 05.00, Qualitative Analysis), and the mass range was m/z 100−700. The surface tension measurements were carried out for different ratios of acetonitrile to acetic acid by the maximum bubble method using a DMPY-2C bubble tensiometer (Nanjing Wanghe Technology Co., Ltd., Nanjing, China). The temperatures of the measurements were kept constant at 298.15 K, and the surface tension reported in this study was the averaged value of three consecutive readings. Isothermal Absorption Experiment. The isothermal adsorption experiment was performed for each chitosan-modified carbon sample by measuring concentrations of the tested analytes at equilibrium. Approximately 24 mg of sorbents (20 mg of carbon materials and 4 mg of chitosan) (amounts for CG-MWNTs were 10, 20, 30, 40, and 50 mg) and 10 mL of an adsorptive aqueous solution of five tetracycline

has harmful effects on public health, including allergic reactions in some hypersensitive individuals, liver damage, accumulative toxicity, yellowing of teeth, transferring drug-resistant bacteria from food to humans, and gastrointestinal disturbance. The determination of tetracycline antibiotics as contaminants in foodstuff samples is considerably important. To ensure human food safety and avoid the unnecessary exposure of customers to antibiotic drugs, in Western countries, the governments have established monitoring programs to determine the tetracycline antibiotic levels in meat, honey, and milk. The European Union (EU) has adopted a maximum residue limit (MRL) of 100 μg/ kg for tetracycline antibiotics in foodstuffs of animal origin.29 This low MRL value demands the development of analytical methods that are sensitive enough to determine these residues in foodstuffs. Hence, it is necessary to develop a simple, rapid, and efficient method for the analysis of these tetracycline antibiotic residues in different samples. In this work, a novel method for the analysis of five tetracycline antibiotics, including 1−5 (Figure 1) was proposed. Finally, the method involves sample preconcentration by miniaturized SPE using chitosan-modified-G-MWNTs as the solid sorbent and analytical determination by ultrahighperformance liquid chromatography and quadrupole time-offlight mass spectrometry (UHPLC-Q-TOF/MS). Moreover, various variables, such as the amount of sorbent, elution solutions, and CNT types, were optimized in detail. The developed method was completely validated including measurement of the linearity, precision, accuracy, decision limit (CCα), and detection capability (CCβ), according to EU Decision 2002/657/EC.30 Finally, the present method was used for the determination of tetracycline residues in honey and milk samples.

■

MATERIALS AND METHODS

Chemicals/Materials. Medium molecular weight chitosan (deacetylation, 75−85%; viscosity, 200−800 cps) was provided by ANPEL Laboratory Technologies Inc. (Shanghai, China). G-MWNTs, carboxyl multiwalled carbon nanotubes (C-MWNTs), carboxyl graphitized multiwalled carbon nanotubes (CG-MWNTs), and hydroxyl graphitized multiwalled carbon nanotubes (HG-MWNTs) of 50 μm length and 8−15 nm o.d. were all purchased from Nanjing Jicang Nano Technology (Nanjing, China). Multiwalled carbon nanotubes (MWNTs) (0.1−10 μm long, 1−3 nm i.d., 3−20 nm o.d.) were obtained from Sigma-Aldrich Shanghai Trading Co., Ltd. (Shanghai, China). Methanol and acetonitrile (HPLC grade) were supplied by Tedia Co. Inc. (Fairfield, OH, USA). Formic acid and acetic acid 2648

DOI: 10.1021/acs.jafc.6b00748 J. Agric. Food Chem. 2016, 64, 2647−2654

Article

Journal of Agricultural and Food Chemistry

Figure 2. SEM images: (A) MWNTs; (B) C-MWNTs; (C) G-MWNTs; (D) CG-MWNTs; (E) HG-MWNTs; (F) chitosan-modified MWNTs; (G) chitosan-modified C-MWNTs; (H) chitosan-modified G-MWNTs; (I) chitosan-modified CG-MWNTs; (J) chitosan-modified HG-MWNTs. TEM images: (K) MWNTs; (L) C-MWNTs; (M) G-MWNTs; (N) CG-MWNTs; (O) HG-MWNTs; (P) chitosan-modified MWNTs; (Q) chitosanmodified C-MWNTs; (R) chitosan-modified G-MWNTs; (S) chitosan-modified CG-MWNTs; (T) chitosan-modified HG-MWNTs. antibiotics with concentrations varying from 20 to 200 μg/mL were introduced into well-closed flasks and shaken for 4 h at 25 °C. The sorbent was separated, and the residual target compounds in the supernatant were detected by UHPLC. The amounts of the five tetracycline antibiotics bound to carbon materials were calculated by subtracting the free mass from the initial mass of tetracycline antibiotics. The adsorption capacity (Q) of the carbon materials was calculated according to the equation

Q = (C0 − Ce)V /m

enhancement for each tetracycline antibiotic. The MEs of the target analytes on UHPLC-Q-TOF-MS were calculated by comparing the slope of the standard addition plot with the slope of the standard calibration plot. The ME values were calculated as follows: ME (%) − slope of tetracycline antibiotics after extraction spike/slope of tetracycline antibiotics in standard solution × 100%.

■

(1)

where C0 is the initial concentration (μg/mL), Ce is the equilibrium concentration (μg/mL), V is the volume of solution (mL), and m is the dosage of carbon materials (mg). The maximum adsorption capacity (Qm) of each carbon material was expressed as

Q m = (Q m − Q )/Kd

RESULTS AND DISCUSSION

Characterization of the Sorbent Materials. SEM was used to capture possible morphological changes in the five types of CNTs investigated, including MWNTs, C-MWNTs, G-MWNTs, CG-MWNTs, and HG-MWNTs. The SEM micrographs of pristine and functionalized MWNTs shown in Figure 2A−E suggest a smooth surface of pristine MWNTs, which becomes quite irregular after functionalization because of a very strong matrix−MWNT interaction. After surface decoration with chitosan (Figure 2F−J), the surfaces of the carbon nanotubes changed considerably and are covered with chitosan protuberances. In addition, TEM was performed to observe the surface microstructures of these CNTs (Figure 2K−O) as well as chitosan-modified pristine and functionalized CNTs (Figure 2P−T). It appeared that the walls of these five investigated MWNTs were relatively smooth and clean, whereas the chitosan-modified MWNTs (Figure 2P−T) appeared stained with an extra phase that was supposed to mainly come from the modified chitosan. Thus, chitosan can be deemed to have successfully modified the surfaces of these pristine MWNTs. Optimization of Miniaturized SPE. Effect of the Amount of Sorbent. To obtain a high extraction efficiency, different amounts of chitosan-modified CG-MWNTs (mixing ratio = 1:5, w/w) were tested for their ability to extract tetracycline antibiotics from sample solutions. Specifically, doses of CGMWNTs ranging from 10 to 50 mg were applied in miniaturized SPE, whereas the amounts of chitosan ranged from 2 to 10 mg. As shown in Figure 3A, the results indicated

(2)

where Kd is the dissociation constant (μg/mL). The values of Kd and Qm can be calculated from the slope and intercept of the linear line in the plot of Q/C versus Q. Miniaturized SPE Procedure. Before extraction, quantities of 20 mg of carbon nanotubes and 4 mg of medium molecular weight chitosan were placed in an agate mortar, and then they were blended together using an agate pestle to obtain an apparently homogeneous mixture. Once fully dispersed (130 s), the mixture was transferred into an empty 3 mL SPE cartridge containing upper and lower sieve plates at each end of the column. After the cartridge had been preconditioned with 5 mL of methanol and 5 mL of double-distilled water sequentially, 20 mL of model sample (2 μg/mL of each analyte in an aqueous mixture), honey, or pure milk (2 mL of each sample, 800 μL of stock standard solutions and 17.2 mL of water) was passed through the cartridge. During this procedure, the flow rate was maintained at 2 mL/min. Then, elution was performed with 600 μL of organic solvent using an aspirator pump. Finally, the eluting fraction was collected and centrifuged at 13000 rpm for 5 min, followed by the injection of 2 μL into the UHPLC-Q-TOF/MS for analysis of the tetracycline antibiotics. Matrix Effect. The assessment of the matrix effect (ME, %) was quantitatively investigated by the evaluation of signal suppression/ 2649

DOI: 10.1021/acs.jafc.6b00748 J. Agric. Food Chem. 2016, 64, 2647−2654

Article

Journal of Agricultural and Food Chemistry

Figure 3. Optimization of the miniaturized SPE process: (A) effect of the amount of CG-MWNTs and chitosan on the extraction efficiencies of tetracycline antibiotics; (B) effect of the amount of CG-MWNTs on the extraction efficiencies of tetracycline antibiotics; (C) effect of the amount of chitosan on the extraction efficiencies of tetracycline antibiotics; (D) effect of the type of elution solvent on the extraction efficiencies of tetracycline antibiotics; (E) selection of CNTs.

observations mentioned above, 4 mg of chitosan and 20 mg of CG-MWNTs were chosen as the optimal sorbent amounts. Selection of the Elution Solvent. The selection of an appropriate elution solvent is of great importance in miniaturized SPE because the elution solvent determines the selectivity, enrichment factor, and elution efficiency toward the target analytes. In miniaturized SPE, the elution solvent must possess a high affinity toward the target compounds and fast kinetics for their quantitative recoveries within a short time. To determine the most appropriate elution solvent, acetonitrile/ acetic acid (9:1, v/v), acetonitrile/acetic acid (8:2, v/v), methanol/acetic acid (8:2, v/v), acetonitrile, and methanol were investigated, with the volume fixed at 600 μL. As shown in Figure 3D, for all of the target compounds, the best elution efficiencies of the compounds were achieved by using acetonitrile/acetic acid (8:2, v/v), followed by acetonitrile/ acetic acid (9:1, v/v), methanol/acetic acid (8:2, v/v), acetonitrile, and methanol under the same extraction and elution conditions. The strong interactions between the tetracycline antibiotics and acetonitrile may be attributed to the π−π interactions between them. Hence, acetonitrile was conceivably able to solubilize tetracyclines better than methanol. For tetracycline antibiotics, which are water-soluble compounds, the addition of acetic acid may be more favorable for desorption than acetonitrile only, leading to a slightly better desorption efficiency. However, an increase in the acetic acid (in the ratio of acetonitrile and acetic acid) would make the surface tension increase, which would increase the difficulty of elution. The results of the maximum bubble method to measure the surface tension of the solutions were found to be as follows:

that the extraction recoveries of the analytes were dramatically increased when the dosages of CG-MWNTs and chitosan were changed from 10 to 20 mg and from 2 to 4 mg, respectively. Obviously, increasing the amounts of CG-MWNTs and chitosan could significantly increase the extraction efficiency. The reason for this phenomenon was that a higher dosage of chitosan modified CG-MWNTs provides a larger number of active sites and a larger available surface area for the sorption of the target compounds. Nevertheless, a visible reduction in the extraction capacity was observed upon increasing the doses from 20 to 50 mg for CG-MWNTs and from 4 to 10 mg for chitosan. This was due to an overly strong adsorption of the five tetracyclines on chitosan-modified CG-MWNTs, which ultimately increased the difficulty of the elution process. Additionally, the maximum adsorption capacity of chitosanmodified CG-MWNTs toward tetracyclines obviously increased with an increase in the chitosan-modified CG-MWNTs dosage from 24 (20 + 4) mg to 60 (50 + 10) mg, which further resulted in a greater difficulty in eluting the analytes due to the increase in the adsorption capacity between the analytes and nanomaterials (see the Supporting Information). As shown in Figure 3B,C, the effects of using only CG-MWNTs or only chitosan (12, 24, 36, 48, and 60 mg) on the extraction efficiencies of tetracyclines were also investigated. It was obvious that the results of the use of only CG-MWNTs (Figure 3B) or only chitosan (Figure 3C) were not better than when using both CG-MWNTs and chitosan, as shown in Figure 3A. Therefore, the chitosan modification technique improved the extraction efficiency for tetracyclines. On the basis of the 2650

DOI: 10.1021/acs.jafc.6b00748 J. Agric. Food Chem. 2016, 64, 2647−2654

Article

Journal of Agricultural and Food Chemistry Table 1. Precision, Contents, and Mean Recoveries of the Investigated Compounds precision (RSD %) interday (n = 6)

intraday (n = 3)

content (μg/kg)

analyte

peak area

retention time

peak area

retention time

acacia honey

orange blossom honey

pure milk

mean recovery (%)

mean RSD (%)

MEa (%)

tetracycline oxytetracycline chlortetracycline metacycline doxycycline

0.70 0.73 0.78 0.53 1.11

0.25 0.25 0.05 0.10 0.13

5.09 7.33 6.89 6.40 5.78

0.24 0.25 0.10 3.70 0.12

trb 6.19 tr 3.06 tr

tr 7.91 tr 6.8 tr

tr tr tr tr tr

91.2 93.6 88.9 101.4 81.5

5.89 5.42 7.71 4.76 6.12

95.9 109.8 97.7 113.6 96.7

a

ME (%) was calculated by comparing the slope of the standard addition plot with the slope of the standard calibration plot. btrace, could not be quantitate.

Figure 4. Total ion chromatograms (TICs) of the spiked samples (containing 2 μg/mL standard solution) by miniaturized SPE: (A) acacia honey; (B) orange blossom honey; (C) pure milk. Extracted ion chromatograms (EICs) of tetracycline antibiotics: 1, tetracycline; 2, oxytetracycline; 3, chlortetracycline; 4, metacycline; 5, doxycycline.

antibiotics, which increased the difficulty of elution. The result showed that the maximum binding amount of tetracycline antibiotics is highest in the case of hydroxyl G-MWNTs, followed by carboxyl G-MWNTs and G-MWNTs (see the Supporting Information), which further increased the elution time of the target analytes due to the increase in the chemical affinity between the tested solutes and MWNTs. Therefore, the G-MWNTs sorbent was selected as the best sorbent for subsequent experiments. Method Validation. Analytical Performance. Under the optimum experimental conditions, the proposed procedure was validated by evaluation of the following performance parameters: linearity, coefficient of determination (r2), limit of detection (LOD), limit of quantitation (LOQ), precision, decision limit (CCα), and detection capability (CCβ). The calibration curves were studied by plotting the peak areas (y) against the standard solution concentrations (x, μg/ mL) at six different concentration levels in the ranges of 0.016− 0.81 μg/mL for tetracycline, 0.010−0.52 μg/mL for oxytetracycline, 0.018−0.89 μg/mL for chlortetracycline, 0.011− 0.56 μg/mL for metacycline, and 0.014−0.69 μg/mL for doxycycline in pure solvents as well as in acacia honey, orange blossom honey, and pure milk matrices. The coefficients of determination (r2) of the calibration curves for all of the

acetonitrile/acetic acid (7:3, v/v), 48.32 mN/m; acetonitrile/ acetic acid (8:2, v/v), 42.49 mN/m; acetonitrile/acetic acid (9:1, v/v), 35.39 mN/m; and acetonitrile, 27.54 mN/m; which also proved the increase in the difficulty of elution as a result of increasing of surface tension. On the basis of the above discussion, acetonitrile/acetic acid (8:2, v/v) was selected as the optimal elution solvent in the following work. Choice of CNTs. Efficient, functional CNTs are an extremely vital factor for this analytical approach because they significantly influence the extraction efficiency. In this work, five types of CNTs, including CG-MWNTs, G-MWNTs, HG-MWNTs, CMWNTs, and MWNTs, were investigated, and their extraction efficiencies for the target tetracyclines were evaluated. The results seen in Figure 3E show that compared with other CNTs, the mixture based on the modification of G-MWNTs as a sorbent was the most effective for the adsorption of all of the target analytes with the extraction efficiencies ranging from 28to 48-fold. It is likely that G-MWNTs compared with MWNTs have a dramatically larger hydrophobic surface and a unique structure with the inner tube cavity as well as stronger interactions between chitosan and the target compounds. Furthermore, the functionalization of G-MWNTs with carboxyl and hydroxyl groups increased the electrostatic interaction and hydrogen bonding between the sorbent and tetracycline 2651

DOI: 10.1021/acs.jafc.6b00748 J. Agric. Food Chem. 2016, 64, 2647−2654

Article

Journal of Agricultural and Food Chemistry

Comparison of the Proposed Method with Other Conventional Methods Reported in the Literature. To further evaluate the selectivity and efficiency of the proposed miniaturized SPE, a comparison of other published extraction methods for the analysis of tetracycline antibiotics from different samples was performed in terms of the sorbent type, sorbent amount, extraction method, elution solvent, instrumental technique, extraction efficiency, extraction time, analysis time, and recoveries. As seen from the references (see the Supporting Information), the analysis time of the five tetracycline antibiotics obtained by the present method was much less than that obtained by other sample extraction methods, including solid-phase extraction (MWNTs),24 ultrasound-assisted extraction (Oasis HLB),31 ultrasound-assisted magnetic solid-phase extraction (Bond Elut Plexa),32 magnetic dispersion extraction (MIMM),33 and molecularly imprinted solid phase extraction (MIPs),34 followed by CE, HPLC, or LC/MS-MS analysis. Additionally, the extraction time for the target analytes in this method was also shorter than that in other methods. Moreover, compared to traditional absorbents, chitosan-modified G-MWNTs have a larger specific surface area and better adsorption capacity, so excellent enrichment factors were achieved, ranging from 28 to 48. Moreover, this method required lower consumption of solid phase sorbent and a smaller volume of elution solvent than its counterparts,31−34 which significantly saves in operating costs. In addition, the recoveries of the miniaturized SPE-UHPLC-Q-TOF-MS method are comparable with and in some cases better than those of the other methods. On the basis of the comparison, miniaturized SPE using chitosan-modified G-MWNTs as the sorbent provided a relatively low-cost and rapid analysis of tetracycline antibiotics in honey and milk samples. In the present study, a simple, rapid, accurate, selective, and effective miniaturized SPE analytical method coupled with UHPLC-Q-TOF/MS has been proposed for the effective preconcentration and simultaneous determination of five tetracycline antibiotics in foodstuffs. Compared with traditional methods, such as ultrasound and reflux, miniaturized SPE possessed notable advantages, including simpler manipulation, smaller amounts of sample and sorbent, lower consumption of organic solvents, and shorter analysis time. In addition, chitosan-modified G-MWNTs as a solid sorbent showed a relatively strong adsorption and enrichment capacity toward tetracyclines. The results indicated that the developed miniaturized SPE-UHPLC-Q-TOF/MS method could be utilized as an advantageous alternative procedure for the analysis of antibiotics in food samples.

tetracycline antibiotics were >0.992, thus confirming the linearity of the method (see the Supporting Information). The LOD was achieved as the lowest concentration of the tetracycline antibiotics based on a signal-to-noise (S/N) ratio of 3, and the values ranged from 0.61 to 10.34 μg/kg. The LOQs, defined as the lowest point on the standard curve resulting in a signal-to-noise ratio of 10, are in the range of 2.02−34.46 μg/ kg. To evaluate the precision of the method, six replicate injections were carried out for the matrix-matched solutions at concentrations of 2 μg/mL tetracycline antibiotics on one day and three successive days, respectively. The intra- and interday repeatabilities based on the peak areas and retention times were calculated as the relative standard deviation (RSD). As shown in Table 1, the RSD values of the peak areas and retention times for inter- and intraday variations were less than 1.11% (n = 6) and 7.33% (n = 3), respectively, for all of the tested compounds, which indicated that the proposed method gave good precision for the analysis of the target analytes. In addition, the repeatability of the proposed method, evaluated by three parallel miniaturized SPE extractions of a sample solution, was