Journal of Chromatography B, 989 (2015) 11–20
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Simultaneous determination of sixteen amide fungicides in vegetables and fruits by dispersive solid phase extraction and liquid chromatography–tandem mass spectrometry Yin-Liang Wu a,b,∗ , Ruo-Xia Chen a,b , Yong Zhu a,b , Jian Zhao a,b , Ting Yang a,b a b
Ningbo Academy of Agricultural Science, Ningbo 315040, PR China Laboratory of Quality and Safety Risk Assessment for Agricultural Products (Ningbo), Ministry of Agriculture, Ningbo 315040, PR China
a r t i c l e
i n f o
Article history: Received 18 November 2014 Accepted 26 February 2015 Available online 7 March 2015 Keywords: Amide fungicides DSPE Vegetables Fruits LC–MS/MS MWCNTs
a b s t r a c t A modified quick, easy, cheap, effective, rugged, and safe (QuEChERS) method using multi-walled carbon nanotubes (MWCNTs) as a reversed-dispersive solid phase extraction (r-dSPE) material combined with ultra-high liquid chromatography tandem mass spectrometry (UHPLC–MS/MS) was developed for the simultaneous determination of 16 novel amide fungicides in vegetables and fruits. After extraction with acetonitrile, a dSPE cleanup procedure, which was developed after the optimization of the type and amount of MWCNTs, the pH value of the extract, the extraction time for MWCNTs, and the type of eluent with MWCNTs material, was conducted. The determination of the target compounds was conducted in less than 7.0 min while the specificity is ensured through the MRM acquisition mode. The linearity of the analytical response across the studied range of concentrations (0.25–500 g/L) was excellent, obtaining correlation coefficients higher than 0.997. The samples were quantified with the matrix matched standard solutions. The average recoveries in cabbage, celery, strawberry, and grape at three spiked levels (0.01, 0.5, and 5.0 mg/kg) were ranged from 72.4 to 98.5% with all RSDs lower than 10%. The limits of detection were below 0.003 mg/kg and the limits of quantification did not exceed 0.01 mg/kg in all matrices. The method demonstrated to be suitable for the simultaneous determination of 16 novel amide fungicides in vegetables and fruits. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Since the first synthesis of carboxin by Schmeling and Kulkain in 1966, amide fungicides have been used for controlling plant diseases for more than 40 years [1]. Recently, many novel amide derivatives (Fig. 1) have been developed and commercialized as fungicides because of their high antifungal activities. In China, seven novel amide fungicides (boscalid, fenoxanil, mandipropamid, zoxamide, fluopicolide, fluopyram, and thifluzamide) have already been registered and widely used in vegetables and fruits [2]. Although these amide fungicides are moderately or low toxicity pesticides for mammals, several authorities around the world have established maximum residue limits (MRLs) in vegetables and fruits to protect consumers [2–4]. Therefore, there is a need for the development of a simple, rapid, specific, inexpensive, and
∗ Corresponding author at: Ningbo Academy of Agricultural Science, Ningbo 315040, PR China. Tel.: +86 574 87928060; fax: +86 574 87928062. E-mail address: wupaddyfi[email protected] (Y.-L. Wu). http://dx.doi.org/10.1016/j.jchromb.2015.02.038 1570-0232/© 2015 Elsevier B.V. All rights reserved.
sensitive method to detect the presence of these amide fungicides in vegetables and fruits. To determine these novel amide fungicides in agricultural products, some enzyme-linked immunosorbent assay (ELISA) [5,6], gas chromatography (GC) [7–9], liquid chromatography (LC) [10], gas chromatography–mass spectrometry (GC–MS) [11–14] and liquid chromatography–mass spectrometry (LC–MS and LC–MS/MS) methods [15–17] have been developed. However, each of ELISA methods can only determine one kind of the novel amide fungicides [5,6]. Moreover, the varieties of the novel amide fungicides are too few (the largest numbers of the novel amide fungicides are usually less than six) for many instrumental methods [11–14,16], which can simultaneous determine multiclass fungicides involved the novel amide fungicides in agricultural products with good sensitivity and accuracy. However, no development of the simultaneous determination method has been reported for those novel amide fungicides (boscalid, fenoxanil, mandipropamid, zoxamide, fluopicolide, fluopyram, and thifluzamide) with MRLs in fruits and vegetables. Recently, although Chen et al. have developed a LC–MS/MS method for determination of six novel amide fungicides
12
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Fig. 1. The chemical structures of 16 novel amide fungicides.
(silthiofam, boscalid, fluopicolide, mandipropamid, cyflufenamid, and mepanipyrim) in fruits and vegetables [19], it still cannot simultaneous determination of those amide fungicides with MRLs. Here, we develop and validate a simple and reliable confirmatory LC–MS/MS analytical method for the analysis of 16 novel amide fungicides including those amide fungicides with MRLs. Moreover, multi-walled carbon nanotubes (MWCNTs) as a dispersive solid phase extraction (dSPE) material has been firstly used for the simultaneous determination of these novel amide fungicides. The type and amount of MWCNTs, the pH value of the extract, the extraction time with the MWCNTs, and the type of eluent have been optimized in this study. After validation studies, the method is suitable for the routine determination of the 16 novel amide fungicides in fruits and vegetables. 2. Materials and methods
USA). Sodium hydroxide was purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Fenoxanil, boscalid, carpropamid, bixafen, silthiofam, thifluzamide (10 g/L in cyclohexane), fluopyram, tiadinil, ethaboxam (10 g/L in ACN), mandipropamid, fluopicolide, fenhexamid, cyflufenamid, zoxamide, furametpyr, and penthiopyrad were purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany). Five types of MWCNTs with similar lengths (10–30 m or 10–20 m) and different outer diameters (MWCNT01 50 nm) were purchased from Nanjing XF NANO Materials Tech Co. Ltd. (Nanjing, China). The specific surface areas (SSAs) of the MWCNTs were 500, 200, 110, 60, and 40 m2 /g for MWCNT01, MWCNT02, MWCNT03, MWCNT04, and MWCNT05, respectively. The water was purified with a Milli-Q reverse osmosis system (Millipore, Milford, Massachusetts, USA).
2.1. Materials and reagents
2.2. Standard solutions
Methanol (LC grade) and acetonitrile (ACN, LC grade) were obtained from Fisher Scientific (Fairlawn, USA). Formic acid (LC grade) was obtained from Tedia Company Inc. (Fairfield,
Individual stock solutions of 14 compounds (excluding ethaboxam and thifluzamide) were prepared in ACN at a concentration of 1000 g/ml. One mixed standard solution (10 g/ml) was
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
13
Table 1 LC–ESI–MS/MS parameters for 16 amide fungicides. Analyte
Precursor ion (m/z)
Daughter ion (m/z)
Dwell time (s)
Collision energy (eV)
Cone voltage (Volts)
Silthiofam Carpropamid Boscalid Fenoxanil Cyflufenamid Furametpyr Zoxamide Fenhexamid Fluopicolide Mandipropamid Ethaboxam Tiadinil Fluopyram Thifluzamide Bixafen Penthiopyrad
268.3 336.3 343.3 329.3 413.4 334.4 338.3 302.3 385.2 412.4 321.3 268.2 397.3 529.2 414.3 360.4
139.1, 252.0* 103.1* , 198.1 140.1* , 307.3 86.2, 302.3* 203.1* , 241.2 131.1* , 186.2 161.0, 189.1* 97.2* , 140.1 147.0, 175.1* 125.1, 328.3* 68.1, 183.2* 45.0, 101.1* 145.2* , 208.2 148.1, 489.1* 266.2, 394.3* 177.2, 276.2*
0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10
18, 12 40, 12 20, 18 22, 12 38, 24 22, 18 38, 22 34, 22 50, 24 38, 14 34, 22 26, 18 54, 20 40, 28 24, 16 34, 14
14 24 28 22 24 26 30 36 30 24 38 32 34 30 26 26
*
Ion for quantification.
prepared in ACN by diluting each stock solution with ACN for the 14compounds (excluding ethaboxam, and thifluzamide). Then, one spiking mixed standard solution (5 g/ml) was prepared by redissolving the residue of 0.5 ml of thifluzamide solution (10 g/ml in cyclohexane), which was evaporated to dryness under nitrogen at room temperature, with 0.5 ml of ethaboxam solution (10 g/ml in ACN) and 0.5 ml of the above mixed standard solution (10 g/ml). The other two spiking mixed standard solutions of the 16 compounds (0.1 and 0.5 g/ml) were prepared by diluting the mixed standard solution (5 g/ml) with ACN. Sixteen individual standard working solutions (1.0 g/ml for the 16 compounds) for MS–MS optimization were prepared by diluting each stock solution with 0.10% formic acid in water/ACN (80:20, v/v). Seven mixed working standard solutions (0.25, 1.0, 2.5, 10.0, 25.0, 100, and 500 g/L) were prepared by diluting the spiking mixed standard solution (5.0 g/ml) with 0.10% formic acid in water/ACN (80:20, v/v). 2.3. Chromatographic conditions A Waters Acquity UPLC instrument (Milford, Massachusetts, USA) was used for analysis. The separation was performed on an Acquity BEH C18 column (2.1 mm × 100 mm, 1.7 m) maintained at 35 ◦ C. The mobile phase consisted of solvent A (0.10% formic acid in water) and solvent B (ACN). The LC linear gradient program was set as follows for the time/percent solvent B: 0.0/20, 0.5/20, 4.0/80, 5.5/80, 5.6/20, and 7.0/20 (total run time = 7 min). The flow rate was 0.30 ml/min. The injection volume was 10 l in full loop injection mode. 2.4. Detection conditions Detection was performed using a Waters XevoTM TQ triplequadrupole mass spectrometer in positive electrospray ionization (ESI) mode. The operating parameters were capillary voltage, 1000 V; source temperature, 150 ◦ C; desolvation gas temperature, 500 ◦ C; cone gas flow, 50 L/h; desolvation gas flow, 1000 L/h. The detection was carried out in the multiple reaction monitoring (MRM) modes. Argon was used as the collision gas, and the collision cell pressure was 2.8 mbar. Additional parameters are shown in Table 1. 2.5. Sample preparation Five grams of homogenized samples were transferred into a 50 ml polypropylene centrifuge tube with screw caps. For recovery assays, an aliquot of standard solution (500 l) was added and
the sample was vortexed for 15 s and allowed to stand at room temperature for 60 min. Subsequently, 10 ml of ACN (9.5 ml of ACN for recovery assays) was added. The sample was homogenized for 1 min using a high-speed blender (Ultra-Tyrrax T25; IKA, Germany) and 2 g of NaCl were added. After vortexing intensively for 1 min, the tube was centrifuged for 3 min at 5000 rpm. Then, an aliquot of supernatant (1.0 ml) was transferred to a clean 50 ml volumetric tube and was diluted with approximately 4.0 ml of water to a final volume of 5 ml (the final concentration of ACN was approximately 20%). Then, the pH of the diluted extract (20% ACN) was adjusted to the range of 3–6 with acetic acid. Afterwards, 10 mg of MWCNTs were added to the diluted extract (5.0 ml). After 1.0 min of extraction with shaking, the suspension was centrifuged at 9,000 rpm for 3 min. The supernatant was discarded, and 10 ml of acetone was added to the tube. Then, the tube was vortexed for 1.0 min and was centrifuged at 9,000 rpm for 2 min. An aliquot of the supernatant (5.0 ml, equivalent to 0.25 g of sample) was evaporated to dryness in a water bath at 40 ◦ C under nitrogen and was reconstituted in 2.5 ml of 0.10% formic acid in water/ACN (80:20, v/v). The resulting solution was filtered through a 0.22 m filter, and 10 l of the filtrate was injected into the LC. 2.6. Matrix effects To evaluate the matrix effects, seven concentrations (0.25, 1.0, 2.5, 10, 25, 100, and 500 g/L) of the 16 fungicides were analyzed in the pure solvent and in the blank sample after the sample preparation procedure. The slope ratio was calculated by comparing the external matrix matched calibration slope of the 16 fungicides with the solvent external calibration slopes. 2.7. Method validation To evaluate the suitability of the method for the determination of the 16 fungicides in fruits and vegetables, validation was performed with a conventional validation procedure that includes the following parameters: specificity, linearity, limit of detection (LOD), limit of quantification (LOQ), accuracy and precision. To verify the absence of interfering substances around the retention time of the 16fungicides, 20 blank samples (cabbage, celery, strawberry, and grape) were analyzed. Calibration curves were constructed using working standard solutions and by plotting the peak area of the quantitative ion pair at concentrations of 0.25, 1.0, 2.5, 10, 25, 100, and 500 g/L for the 16 fungicides. The LODs and LOQs were defined as the concentrations of the 16compounds that produced chromatographic peaks (qualitative
14
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Fig. 2. The effects of different types of MWCNTs on the adsorption rates of the 16 amide fungicides in celery original extract.
transitions) with signal-to-noise (S/N) greater than 3 and 10 for fruit and vegetable samples, respectively. The LOQs are estimated from the chromatogram corresponding to the lowest point used in the matrix-matched calibration. As no certified reference material was available, accuracy was determined as the mean recovery by using spiked blank samples. To this end, blank fruit and vegetable samples were spiked with the 16 fungicides at three different levels (0.01, 0.05, and 0.5 mg/kg). The repeatability was measured using the 15 spiked blank samples (n = 5 replicates per concentration level and analyzed in three independent analytical runs) and expressed as the coefficient of variation CVr . The within-laboratory reproducibility was measured using the 15 spiked blank samples (n = 5 replicates per concentration level and analyzed at three occasions with three different operators) and was expressed as the coefficient of variation CVR . 3. Results and discussion
current experiment, four identification points, one parent (1 point) and two transitions (each 1.5 points), were monitored. The optimal MS–MS conditions are described in Table 1. After optimization of MS parameters, the mobile phase composition was investigated under the chromatographic column (Acquity BEH C18 column, 2.1 mm × 100 mm, 1.7 m). The solution of water/ACN and 0.10% formic acid in water/ACN had been selected by Dong et al. [19] and Cui et al. [18] because they eluted more efficiently over time. After carefully investigation, it was found that there was not obvious difference between the responses of the 16amide fungicides using 0.10% formic acid in water/ACN and those of using water/ACN under the same gradient elution conditions. Moreover, it was found that only eight fungicides can be eluted from the column under the gradient conditions described at Section 2.3 using 0.10% formic acid in water/methanol as the mobile phase. So, 0.10% formic acid in water/ACN was selected as the mobile phase in the current study.
3.1. Optimization of instrumentation conditions
3.2. Optimization of sample preparation conditions
For some of established LC–MS/MS method, negative ion ESI and positive ion APCI had been used to determine some of amide fungicides with good sensitivity [11–19]. However, positive ion ESI is typically used for simultaneous determination of amide fungicides among most of established LC–MS/MS methods, because ESI source is easy to maintain and the [M + H]+ ion forms easily formed due to the imino group on them [19]. In the present study, the appropriate diagnostic ions with good sensitivity had been obtained for the 16 amide fungicides using IntelliStart software in ESI positive mode (Table 1). So, ESI positive mode had been chosen in this study. Identification points (IPs) for the confirmation and identification had been introduced by the European Commission Decision 2002/657/EC [20]. According to 2002/657/EC and LC–MS/MS, four identification points are required for most of compounds. In the
For most of established LC–MS/MS method, ACN had been used to extract amide fungicides in fruits and vegetables [15,18,19]. Thus, the extraction efficiency in this study was studied with ACN at different volume (10, 15, 20, 25, and 30 ml). All of amide fungicides were added to the vegetables and fruits (cabbage, celery, strawberry, and grape) and were extracted according the extraction method described in sample preparation section. It was found that the recoveries were above 95% for all of amide fungicides. Considering the environmental protection, which requires minimizing the use of undesirable organic solvents, 10 ml of ACN was chosen to extract amide fungicides from vegetables and fruits. Optimization of dSPE procedure is an important process to get good purification efficiency and full recovery. It usually involves the type and amount of MWCNTs, the pH value of the extract, shaking
Fig. 3. The effects of different types of MWCNTs on the adsorption rates of the 16 amide fungicides in diluted celery extract containing 50% of water.
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
15
Fig. 4. The effects of different types of MWCNTs on the adsorption rates of the 16 amide fungicides in diluted celery extract containing 80% of water.
extraction time with MWCNTs, and the type of eluting solvent and its volume. The recovery was calculated with a matrix-matched external standard method during optimization of sample preparation. All of experiments were performed in three replicates (n = 3). 3.2.1. Effects of type of MWCNTs A series of experiments were executed to investigate the influence of the type of MWCNTs. Firstly, the effects of the types of MWCNTs were investigated with 5 ml of the extract of celery and 50 mg of MWCNTs. The spiking concentration in the extract was 25 g/L and the shaking time was 2 min. The adsorption rate was acquired by using the formula: adsorption rate (%) = 100−(the concentration of the 16 amide fungicides in the celery extract after MWCNT adsorption/the initial concentration of the 16 amide fungicides in the celery extract) × 100. The results are shown in Fig. 2. It was found that the highest adsorption rates were observed for all of the 16 fungicides with MWCNT01, which have the largest SSAs. However, most of the adsorption rates were below 80% for the 16 fungicides when MWCNT01 had been used. In order to improve the adsorption efficiency of MWCNTs, the celery extract was diluted with water because the adsorption capacity of CNTs for hydrophobic compounds, such as the 16 amide fungicides, usually increases as the water concentration increases in mixtures of water and an organic solvent [21]. Significantly higher adsorption rate were obtained with the diluted celery extract containing 50% of water for these amide fungicides (Fig. 3). Surprisingly, all of the adsorption rates approached 100% except silthiofam when MWCNT01, MWCNT02, MWCNT03 and MWCNT04 had been used (Fig. 4) with the diluted extract containing 80% of water. Considering MWCNT01 has the strongest adsorption ability for these fungicides, MWCNT01 and the diluted celery extract with 80% of water were chosen to carry out further experiments. 3.2.2. Effects of amount of MWCNTs To explore the effects of the amount of MWCNT01 on the adsorption rates of the 16 fungicides, different amounts of MWCNT01 (5, 7.5, 10, 15, and 25 mg) were used. Other experimental conditions were the same as above. Results are shown in Fig. 5. It was found that the adsorption rates of most of fungicides were approached 100% at 10 mg. Moreover, the adsorption rates of sithiofam,
fluopyram, and thifluzamide were above 90% at 10 mg. To obtain the best results using the least amount of MWCNTs, 10 mg of MWCNTs were chosen for the present study. 3.2.3. Effects of pH The effects of the celery diluted extract (20% ACN) with different pH values (3, 4, 5, 6, 7, 8, and 9) and without pH adjustment (pH = 5.8) on the adsorption rate were investigated with 10 mg of MWCNT01. It was found that there were no obvious differences in the adsorption rates of the 16 amid fungicides when the pH of the diluted feed extract (20% ACN) varied from 3 to 6 (Fig. 6). However, there were slightly reductions in the adsorption rates of the 16 amide fungicides when pH values were greater than 7.0. For this phenomenon, it may be the reason that a decrease of hydrophobicity of the 16 amide fungicides had occurred when the pH values of diluted celery extract is greater than 7.0 because we had found that there were distinct correlations between the adsorption rate and hydrophobic properties for sulfonamides and resorcylic acid lactones [21,22]. In order to ensure the applicability of the method for all of samples, the diluted extract (20% ACN) had been required to adjust pH value to the range of 3–6 in the present study. 3.2.4. Effects of extraction time The effects on the recovery caused by different shaking extraction times (1, 2, 5, and 10 min) were investigated with 10 mg of MWCNT01 for the diluted celery extract (80% water). The adsorption on the MWCNTs was observed to be a fast process for the 16 fungicides because adsorption rates of most of fungicides were 100% when the extraction time was longer than 1 min. Moreover, no significant differences in the adsorption rates were observed for any of the 16 fungicides. The results observed for the 16 fungicides were similar with many other organic chemicals, such as resorcylic acid lactones [23], nitroaromatic compounds [24], organophosphorus pesticides [25], quinolone antibiotics [26], and sulfonamides [21]. Therefore, a 1 min extraction time was chosen for the subsequent experiments. 3.2.5. Effects of the type of eluent Due to the 16amide fungicides are easily dissolved in organic solvents, the effects of different solvents (ACN, methanol, acetone,
16
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Fig. 5. The effects of different amount of MWCNT01 on the adsorption rates of the 16 amide fungicides in the diluted celery extract containing 80% of water.
Fig. 6. The effects of different pH values on the adsorption rates of 16 amide fungicides in celery diluted extract.
Fig. 7. The effects of different solvents as eluents on the recoveries of the 16 amide fungicides.
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20 Table 2 The linearity and regression coefficients of standard curves of 16 amide fungicides by LC–MS/MS. Analyte
Linear equation
r2
Silthiofam Carpropamid Boscalid Fenoxanil Cyflufenamid Furametpyr Zoxamide Fenhexamid Fluopicolide Mandipropamid Ethaboxam Tiadinil Fluopyram Thifluzamide Bixafen Penthiopyrad
y = 8237.3x + 7889.1 y = 534.8x + 432.8 y = 2433.7x + 3559.3 y = 2016.3x + 3289.2 y = 1367.4x + 479.4 y = 2277.2x + 2433.3 y = 2653.1x + 1876.4 y = 1783.3x + 1107.2 y = 4127.8x + 3267.4 y = 3765.3x + 1877.3 y = 4578.2x + 2865.9 y = 1057.3x + 1077.42 y = 12779.1x + 7187.4 y = 202.6x + 133.1 y = 789.3x + 332.7 y = 7635.2x + 4339.0
0.9980 0.9987 0.9975 0.9979 0.9977 0.9982 0.9973 0.9982 0.9989 0.9990 0.9974 0.9979 0.9988 0.9984 0.9986 0.9983
and ethyl acetate) on elution recovery were carefully investigated at a spiking concentration of 50 g/L under the optimized MWCNT conditions (described above). For convenience, the volume of the three solvents was firstly fixed at 5 ml. The results are shown in
17
Fig. 7. From Fig. 7, it was found that acetone was the optimum solvent as eluent for elution of most of amide fungicides from MWCNTs. Moreover, the elution recoveries of most of amide fungicides were more than 80% when 5 ml of acetone were used except tiadinil (51.5%) and boscalid (73.9%). In order to improve elution efficiency for tiadinil and boscalid, 10 ml of acetone was further used to investigate. It was found that the elution recoveries of tiadinil and boscalid were reached 75.3 and 87.9%, respectively. These results of tiadinil and boscalid are within the regulations of document no. SANCO 12495-2011 [27]. So, 10 ml of acetone was finally selected as eluent in the present study. Under the above optimized conditions, the pre-treatment procedure for the MWCNTs was simple and fast. Compared to an SPE cartridge that costs at least $3 USD per cartridge, the cost of the MWCNTs required for one sample is approximately $0.10 USD. In addition, the total time required for the analysis of one sample was only approximately 25 min and less than the time (at least 50 min) of the method developed by Chen et al. [19]. 3.3. Matrix effects Co-elution of undetected matrix components may enhance or reduce the ion intensity of the analytes and affect the accuracy and
Table 3 LOD and LOQ (g/kg) obtained for 16 amide fungicides in vegetables and fruits by LC–MS/MS. Analyte
Silthiofam Carpropamid Boscalid Fenoxanil Cyflufenamid Furametpyr Zoxamide Fenhexamid Fluopicolide Mandipropamid Ethaboxam Tiadinil Fluopyram Thifluzamide Bixafen Penthiopyrad
Cabbage
Celery
Strawberry
Grape
LOD
LOQ
LOD
LOQ
LOD
LOQ
LOD
LOQ
0.2 0.5 0.4 0.2 0.8 0.3 0.4 1.5 0.1 0.2 0.1 2.5 0.1 1.9 0.3 0.4
0.8 1.6 1.5 0.8 2.6 1.1 1.3 5.0 0.3 0.5 0.3 8.3 0.2 6.2 1.0 1.2
0.3 0.4 0.4 0.3 0.7 0.4 0.5 1.9 0.1 0.1 0.1 1.9 0.1 2.5 0.3 0.3
0.9 1.5 1.4 0.9 2.4 1.4 1.5 6.3 0.2 0.4 0.4 6.2 0.2 8.3 1.1 1.1
0.2 0.5 0.5 0.3 1.0 0.4 0.4 1.5 0.1 0.2 0.1 1.9 0.1 1.9 0.3 0.4
0.7 1.6 1.6 0.9 3.2 1.4 1.5 5.0 0.3 0.5 0.3 6.2 0.2 6.2 1.0 1.3
0.2 0.5 0.4 0.3 0.9 0.5 0.4 1.5 0.1 0.2 0.1 2.5 0.1 1.9 0.3 0.4
0.7 1.6 1.5 0.9 3.1 1.5 1.4 5.0 0.3 0.5 0.3 8.3 0.2 6.2 1.0 1.2
Fig. 8. The representative MRM chromatograms of a working standard solution (0.25 g/L for the 16 amide fungicides, (a) a blank celery sample (b) and a spiked blank celery sample (0.01 mg/kg for the 16 amide fungicides, c).
18
Table 4 Mean recoveries and variation coefficients (CVr and CVR ) of 16 amide fungicides from celery and grape samples by LC–MS/MS. Analyte
Silthiofam
Carpropamid
Boscalid
Fenoxanil
Cyflufenamid
Zoxamide
Fenhexamid
Fluopicolide
Mandipropamid
Ethaboxam
Tiadinil
Fluopyram
Thifluzamide
Bixafen
Penthiopyrad
0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5
grape
Celery Mean recovery (%, n = 5) 87.4 (4.1) 83.9 (4.5) 88.1 (4.2) 92.7 (4.8) 94.5 (4.4) 91.4 (3.2) 83.7 (4.2) 85.1 (4.0) 80.4 (4.7) 91.6 (4.1) 93.2 (4.7) 94.0 (4.2) 96.7 (4.6) 94.2 (4.2) 93.7 (4.0) 92.4 (4.4) 95.1 (3.8) 90.9 (4.2) 94.5 (4.2) 93.2 (4.5) 96.1 (4.6) 93.9 (4.9) 90.1 (4.3) 93.5 (4.3) 95.1 (4.7) 93.7 (4.6) 96.2 (4.9) 91.3 (4.9) 88.9 (4.4) 93.2 (4.3) 93.7 (4.4) 90.3 (4.8) 87.4 (4.7) 80.1 (4.6) 73.3 (5.1) 78.1 (3.8) 93.7 (4.8) 93.2 (4.5) 94.6 (4.3) 90.6 (4.8) 93.4 (4.5) 91.3 (4.6) 88.4 (4.6) 89.2 (4.5) 87.7 (4.2) 96.7 (4.3) 94.6 (4.6) 98.1 (4.4)
85.7 (5.3) 82.2 (3.9) 89.3 (4.8) 93.2 (3.7) 92.8 (3.8) 94.6 (3.3) 86.4 (3.9) 87.2 (4.3) 86.9 (3.9) 94.9 (3.9) 93.1 (4.5) 92.5 (3.8) 94.3 (5.1) 95.9 (4.7) 98.1 (3.6) 92.8 (4.8) 91.0 (4.3) 94.5 (4.1) 94.8 (5.1) 92.7 (4.6) 92.0 (4.3) 96.1 (4.6) 93.9 (4.7) 95.2 (3.6) 88.4 (4.4) 92.2 (4.3) 94.1 (4.6) 87.8 (5.3) 90.7 (4.2) 92.9 (4.2) 89.1 (4.9) 86.9 (4.1) 91.3 (4.8) 76.4 (3.8) 81.3 (4.4) 70.8 (4.6) 91.9 (4.7) 93.6 (4.9) 93.8 (3.8) 93.2 (4.8) 92.4 (4.1) 90.1 (4.3) 89.9 (4.9) 90.1 (4.4) 91.4 (2.9) 94.7 (5.3) 95.9 (4.1) 94.7 (4.9)
88.3 (5.0) 83.9 (3.9) 84.2 (4.2) 90.9 (4.7) 95.1 (4.1) 96.3 (3.7) 88.1 (4.6) 81.8 (4.7) 83.2 (4.4) 96.1 (4.8) 94.2 (3.7) 92. 8(4.6) 92.8 (5.6) 93.7 (4.3) 95.1 (4.2) 93.7 (4.9) 92.9 (4.4) 88.9 (4.6) 94.3 (4.7) 93.8 (4.5) 91.6 (4.2) 93.7 (5.3) 94.3 (4.6) 92.9 (4.8) 89.5 (5.9) 93.7 (4.7) 93.2 (4.3) 91.6 (4.8) 92.2 (4.2) 90.8 (4.7) 93.2 (5.1) 90.1 (4.4) 86.9 (4.5) 77.2 (4.3) 72.4 (3.9) 76.9 (4.1) 90.6 (3.9) 92.2 (5.3) 94.7 (4.6) 93.2 (4.1) 87.6 (4.4) 89.2 (4.6) 86.9 (5.4) 90.2 (4.8) 92.3 (4.3) 97.8 (4.7) 96.4 (3.8) 98.5 (4.5)
CVr (%)
CVR (%)
Mean recovery (%, n = 5)
4.9 4.2 4.4 4.4 4.0 3.5 4.3 4.6 4.3 4.6 4.4 4.2 5.0 4.5 4.0 4.6 4.2 4.3 4.6 4.5 4.3 5.0 4.5 4.4 4.9 4.4 4.5 5.0 4.4 4.2 4.5 4.3 4.3 4.3 4.6 5.0 4.4 3.9 4.2 4.4 4.5 4.2 3.9 4.3 4.0 4.6 4.2 4.4
6.7 5.9 5.8 6.1 5.8 6.0 5.8 6.7 6.2 6.0 5.7 5.9 6.7 5.8 5.4 6.7 5.8 5.4 6.4 6.2 6.3 7.1 6.0 5.9 7.6 6.4 6.2 6.1 5.8 6.3 6.4 6.2 6.0 6.7 6.6 7.1 6.4 6.8 6.1 6.7 6.2 6.1 7.2 6.6 5.9 7.8 6.4 6.2
86.3 (4.5) 81.7 (4.1) 84.4 (3.8) 93.5 (4.2) 92.3 (3.9) 95.1 (4.7) 88.1 (4.4) 83.5 (4.7) 81.9 (4.1) 96.1 (3.9) 94.2 (4.6) 94.8 (4.1) 94.9 (5.3) 92.3 (4.6) 95.1 (3.8) 96.8 (4.9) 93.1 (4.3) 92.6 (4.2) 93.9 (4.4) 95.4 (4.6) 93.7 (4.8) 93.3 (4.6) 95.9 (4.1) 94.2 (4.5) 94.9 (4.8) 93.2 (4.4) 94.5 (4.3) 92.4 (5.4) 90.7 (4.7) 89.6 (4.8) 88.6 (5.0) 92.4 (5.2) 90.7 (4.5) 72.4 (4.3) 75.7 (4.6) 81.3 (4.0) 89.2 (5.2) 92.6 (4.8) 93.3 (4.7) 88.9 (5.0) 90.7 (4.7) 94.1 (4.6) 93.2 (4.9) 91.6 (4.3) 92.7 (4.4) 95.7 (4.4) 98.1 (4.6) 94.6 (4.1)
87.3 (4.3) 86.9 (4.8) 83.8 (4.1) 92.4 (4.6) 98.7 (4.8) 96.4 (4.6) 84.1 (3.9) 83.7 (4.4) 85.9 (3.6) 89.7 (4.2) 93.5 (4.6) 92.6 (4.9) 95.3 (4.4) 94.1 (4.8) 97.2 (4.6) 91.7 (5.6) 93.5 (4.7) 94.1 (4.9) 90.6 (4.9) 92.4 (4.3) 93.7 (4.2) 92.8 (4.6) 91.6 (5.0) 94.5 (4.7) 91.8 (4.9) 94.3 (4.6) 97.7 (4.7) 93.2 (4.8) 90.6 (4.3) 92.3 (4.4) 92.8 (4.7) 87.4 (4.5) 90.9 (4.5) 78.1 (4.3) 73.3 (4.6) 77.9 (4.3) 92.0 (4.4) 95.1 (4.7) 93.8 (4.3) 93.2 (4.7) 87.3 (4.4) 91.5 (4.7) 91.5 (5.0) 90.4 (4.8) 88.6 (4.5) 97.1 (3.6) 93.8 (4.3) 95.2 (4.5)
82.7 (5.1) 85.3 (4.3) 88.2 (4.7) 93.1 (3.9) 94.2 (4.5) 93.9 (3.4) 81.5 (4.6) 82.9 (4.8) 84.1 (4.1) 94.6 (4.8) 93.7 (4.1) 96.1 (3.9) 90.7 (4.1) 94.3 (4.6) 95.7 (4.2) 92.9 (4.5) 93.7 (4.3) 90.8 (4.1) 94.2 (4.6) 96.3 (4.4) 93.2 (4.6) 92.7 (4.3) 93.8 (4.6) 94.7 (4.1) 94.3 (5.1) 92.9 (4.7) 95.8 (4.5) 89.5 (5.4) 88.1 (4.9) 94.2 (4.1) 90.8 (4.8) 86.4 (5.2) 92.1 (4.6) 79.8 (4.4) 78.0 (4.7) 73.4 (4.5) 93.9 (4.8) 94.1 (4.3) 92.5 (4.6) 90.7 (5.2) 92.8 (4.6) 91.7 (4.3) 92.4 (4.7) 91.5 (4.3) 87.6 (3.3) 97.6 (5.2) 95.2 (4.5) 97.3 (4.2)
CVr (%)
CVR (%)
4.6 4.2 4.5 4.3 4.4 4.4 4.2 4.6 3.9 4.6 4.4 4.7 4.7 4.6 4.3 5.1 4.4 4.7 4.7 4.5 4.4 4.4 4.7 3.2 4.6 4.3 4.3 5.0 4.5 4.4 4.7 4.7 4.0 4.8 4.3 5.4 4.9 4.4 3.3 4.8 4.6 4.3 4.5 4.4 4.0 4.6 4.5 4.0
6.4 5.8 6.5 6.3 5.7 6.2 6.1 6.6 5.9 6.2 6.1 6.6 6.5 6.3 6.1 6.9 6.2 6.2 6.4 6.5 6.1 6.0 6.4 6.0 6.2 6.0 6.2 6.9 6.3 6.1 6.4 6.6 6.0 6.4 6.2 5.7 6.7 6.3 6.4 6.8 6.5 6.2 6.6 6.1 5.8 6.5 6.0 5.8
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Furametpyr
Fortified concentration (mg/kg)
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
reproducibility of the quantitative analysis when ESI-MS was used. So, the matrix effect was evaluated by comparing the ratios of the external calibration slopes for the matched matrix to the external calibration slopes for the solvent. The numbers of the amide fungicides, which the slope ratios were between 0.80–1.20 before purification with MWCNT01, were 11, 10, 12, and 14 for cabbage, celery, strawberry, and grape samples, respectively. Moreover, the ranges of slope ratios were 0.68–1.33, 0.59–1.42, 0.77–1.36, and 0.82–1.40 for the 16 amide fungicides in cabbage, celery, strawberry, and grape samples, respectively. After purification with MWCNT01, all of slope ratios were between 0.80–1.20 (0.90–1.02 for cabbage samples, 0.93–1.03 for celery samples, 0.92–1.04 for strawberry samples, and 0.94–1.02 for grape samples) in the above four kinds of samples. From the calculated matrix effect results, it was concluded that the matrix effect had decreased after purification with MWCNTs and there was a slight matrix effect for the 16 amide fungicides after purification. However, a strong matrix effect had been discovered for qualitative ions of two amide fungicides (carpropamid and thifluzamide). So, the matrix standard solutions were chosen for confirmation in this study. 3.4. Method validation 3.4.1. Linearity The calibration curves of working standard solutions were obtained by plotting the peak area of the quantitative ion pairs over the range from 0.25 to 500 g/L. The results are reported in Table 2. The correlation coefficients (r2 ) of the calibration curves were greater than 0.9973. To acquire more accurate quantification, single point calibration with the matrix matched standard solutions (1.0, 2.5, and 25 g/L for 0.01, 0.05 and 0.5 mg/kg spiking level, respectively) had been used in the recovery experiments according to the description of EURACHEM Guide [28]. Moreover, the standard curve point closest to the measured value of the “real” sample could be chosen for single point calibration when analyzing “real” samples. The MRM chromatograms of the standard solution are shown in Fig. 8a. 3.4.2. Specificity Fig. 8b and c indicated that there were no interfering peaks from endogenous compounds at the retention times of the 16 compounds. Therefore, the optimized pre-treatment procedure coupled to UHPLC–MS/MS provided a clean chromatogram without interferences. 3.4.3. LOD and LOQ In the present study, the LODs and LOQs are estimated from the chromatogram corresponding to the lowest point used in the matrix-matched calibration (0.25 g/L). Each matrix standard was injected three times. LOD, calculated as S/N ratio = 3, for the 16 amide fungicides was ranged from 0.1 to 2.5 g/kg in all matrices (Table 3), respectively. LOQ, calculated as S/N ratio = 10, for the 16 amide fungicides was ranged from 0.2 to 8.3 g/kg in all matrices (Table 3), respectively. 3.4.4. Accuracy and precision The recovery and reproducibility of the method were measured by spiking the blank samples with the 16 amide fungicides at three different concentrations (0.01, 0.05, and 0.5 mg/kg) on three separate occasions. The results are shown in Table 4 according to the quantitative method described in Section 3.4.1 using the matrix matched standard solutions, which had been prepared by spiking the analytes into the blank sample after the extraction/cleanup procedure. The average recoveries, repeatability, and reproducibility varied from 72.4 to 98.5%, from 3.2 to 5.4% (CVr), and from 5.4 to 7.8% (CVR ), respectively. These recoveries and CVs demonstrate that
19
this method can achieve a satisfactory accuracy and precision for the 16 amide fungicides residue analysis in vegetables and fruits. Fig. 8b and c shows the representative typical chromatograms of the 16 amide fungicides of the celery blanks and spiked samples. 3.5. Applications of the method To evaluate the effectiveness and applicability of this method for determining the 16 amide fungicides, real samples (45 grape samples, 14 celery samples, 10 strawberry samples, and 12 cabbage samples) had been obtained from several markets in Ningbo. Boscalid was detected in 4 grape samples ranged from 0.024 to 0.18 mg/kg. The presence of boscalid at these levels in grape does not pose a threat to the consumer since they are below the MRLs established by EU (2.0 mg/kg for grape) and China (3.0 mg/kg for strawberry and no rule for grape). No other amide fungicides were detected in these samples. 4. Conclusions In the present study, a method for the determination of 16 amide fungicides in vegetables and fruits was established using MWCNTs as a sorbent for dSPE coupled with LC–MS/MS. The pre-treatment procedure using MWCNTs was very simple and fast compared to previously reported methods. The current method was validated with spiked vegetables and fruit samples and good recoveries with good CVs were obtained. The LODs and LOQs were found to be sufficiently low to determine the residues of 16 amide fungicides samples. Moreover, the method was successfully applied to the routine analysis of amide fungicides in vegetable and fruit samples. Acknowledgements The Project in Agriculture of Ningbo, (No. 2013C11003, No. 2013NK33 and No. 2014C10058) and National Agricultural Products Risk Assessment Project (GJFP2014011) provided financial support for this work. References [1] G.A. Carter, J.L. Huppatz, R.L. Wain, Ann. Appl. Biol. 84 (1976) 33. [2] National Food Safety Standard-Maximum Residue Limits for Pesticides in Food, National Health and Family Planning Commission of the People’s Republic of China, Ministry of Agriculture of the People’s Republic of China, Beijing, 2014, GB 2763-2014. [3] European Union, European Union database on Pesticide on MRLs. (http://ec.europa.eu/sanco pesticides/public/?event=substance.selection). [4] United States Department of Agriculture Foreign Agricultural Service-Pesticide MRL Database. (http://www.mrldatabase.com/default.cfm?selectvetdrug=0). [5] F.C. Esteve-Turrillas, A. Abad-Fuentes, J.V. Mercader, Food Chem. 124 (2011) 1727. [6] J.V. Mercader, A. Abad-Fuentes, J. Agric. Food Chem. 57 (2009) 5129. [7] D.T. Likas, N.G. Tsiropoulos, G.E. Miliadis, J. Chromatogr. A 1150 (2007) 208. [8] V.T. Gajbhiye, Suman Gupta, Anshu Singh, R.K. Gupta, T.K. Das, Pestology 26 (2002) 54. [9] E.G. Amvrazi, A.T. Papadi-Psylloe, N.G. Tsiropoulos, Intern. J. Environ. Anal. Chem. 90 (2010) 245. [10] A. Melo, A. Aguiar, C. Mansilha, O. Pinho, I.M.P.L.V.O. Ferreira, Food Chem. 130 (2012) 1090. [11] L. Lagunas-Allué, J. Sanz-Asensio, M.T. Martínez-Soria, J. AOAC Int. 95 (2012) 1511. [12] L. Lagunas-Allué, J. Sanz-Asensio, M.T. Martínez-Soria, J. Chromatogr. A 1270 (2012) 62. [13] A. Angioni, L. Porcu, F. Dedola, F. Dedola, Pest Manag. Sci. 68 (2012) 543. [14] R.M. Gonzalez-Rodriguez, B. Cancho-Grande, J. Simal-Gandara, J. Chromatogr. A 1216 (2009) 6033. [15] F. Dong, X. Chen, X. Liu, J. Xu, Y. Li, W. Shan, Y. Zheng, J. Chromatogr. A 1262 (2012) 98. [16] C. Coscollà, V. Yusà, M.I. Beser, A. Pastor, J. Chromatogr. A 1216 (2009) 8817. [17] M. Hengel, B. Hung, J. Engebretson, T. Shibamoto, J. Agric. Food Chem. 51 (2003) 6635. [18] S.H. Cui, J.L. Qian, H. Duan, J.F. Liu, L.M. Lin, K.Y. Wang, Chin. J. Anal. Chem. 39 (2011) 1836.
20
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
[19] X. Chen, Z. Li, Z. Cao, J. Gong, X. Cao, M. Chen, Chin. J. Chromatogr. 31 (2013) 954. [20] European Commission, Decision 2002/657/EC (2002), Off. J. Eur. Commun. (2002) L221. [21] H. Peng, B. Pan, M. Wu, R. Liu, D. Zhang, D. Wu, B. Xing, J. Hazard. Mater. 342 (2012) 211–212. [22] X.L. Hou, Y.L. Wu, T. Yang, X.D. Du, J. Chromatogr. B 929 (2013) 107. [23] Y.F. Ying, Y.L. Wu, Y. Wen, T. Yang, X.Q. Xu, Y.Z. Wang, J. Chromatogr. A 1307 (2013) 41.
[24] X.E. Shen, X.Q. Shan, D.M. Dong, X.Y. Hua, G. Owens, J. Colloid Interface Sci. 330 (2009) 1. [25] M.Á. González-Curbelo, M. Asensio-Ramos, A.V. Herrera-Herrera, J. HernándezBorges, Anal. Bioanal. Chem. 404 (2012) 183. [26] A.V. Herrera-Herrera, L.M. Ravelo-Pérez, J. Hernández-Borges, M.M. Afonso, J. Palenzuela, M.A. Rodríguez-Delgado, J. Chromatogr. A 1218 (2011) 5352. [27] European Commission, Document No. SANCO/12495/2011, 2011. [28] EURACHEM Guide, the Fitness for Purpose of Analytical Methods, first ed, Teddington, Middlesex, 1998.
Contents lists available at ScienceDirect
Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb
Simultaneous determination of sixteen amide fungicides in vegetables and fruits by dispersive solid phase extraction and liquid chromatography–tandem mass spectrometry Yin-Liang Wu a,b,∗ , Ruo-Xia Chen a,b , Yong Zhu a,b , Jian Zhao a,b , Ting Yang a,b a b
Ningbo Academy of Agricultural Science, Ningbo 315040, PR China Laboratory of Quality and Safety Risk Assessment for Agricultural Products (Ningbo), Ministry of Agriculture, Ningbo 315040, PR China
a r t i c l e
i n f o
Article history: Received 18 November 2014 Accepted 26 February 2015 Available online 7 March 2015 Keywords: Amide fungicides DSPE Vegetables Fruits LC–MS/MS MWCNTs
a b s t r a c t A modified quick, easy, cheap, effective, rugged, and safe (QuEChERS) method using multi-walled carbon nanotubes (MWCNTs) as a reversed-dispersive solid phase extraction (r-dSPE) material combined with ultra-high liquid chromatography tandem mass spectrometry (UHPLC–MS/MS) was developed for the simultaneous determination of 16 novel amide fungicides in vegetables and fruits. After extraction with acetonitrile, a dSPE cleanup procedure, which was developed after the optimization of the type and amount of MWCNTs, the pH value of the extract, the extraction time for MWCNTs, and the type of eluent with MWCNTs material, was conducted. The determination of the target compounds was conducted in less than 7.0 min while the specificity is ensured through the MRM acquisition mode. The linearity of the analytical response across the studied range of concentrations (0.25–500 g/L) was excellent, obtaining correlation coefficients higher than 0.997. The samples were quantified with the matrix matched standard solutions. The average recoveries in cabbage, celery, strawberry, and grape at three spiked levels (0.01, 0.5, and 5.0 mg/kg) were ranged from 72.4 to 98.5% with all RSDs lower than 10%. The limits of detection were below 0.003 mg/kg and the limits of quantification did not exceed 0.01 mg/kg in all matrices. The method demonstrated to be suitable for the simultaneous determination of 16 novel amide fungicides in vegetables and fruits. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Since the first synthesis of carboxin by Schmeling and Kulkain in 1966, amide fungicides have been used for controlling plant diseases for more than 40 years [1]. Recently, many novel amide derivatives (Fig. 1) have been developed and commercialized as fungicides because of their high antifungal activities. In China, seven novel amide fungicides (boscalid, fenoxanil, mandipropamid, zoxamide, fluopicolide, fluopyram, and thifluzamide) have already been registered and widely used in vegetables and fruits [2]. Although these amide fungicides are moderately or low toxicity pesticides for mammals, several authorities around the world have established maximum residue limits (MRLs) in vegetables and fruits to protect consumers [2–4]. Therefore, there is a need for the development of a simple, rapid, specific, inexpensive, and
∗ Corresponding author at: Ningbo Academy of Agricultural Science, Ningbo 315040, PR China. Tel.: +86 574 87928060; fax: +86 574 87928062. E-mail address: wupaddyfi[email protected] (Y.-L. Wu). http://dx.doi.org/10.1016/j.jchromb.2015.02.038 1570-0232/© 2015 Elsevier B.V. All rights reserved.
sensitive method to detect the presence of these amide fungicides in vegetables and fruits. To determine these novel amide fungicides in agricultural products, some enzyme-linked immunosorbent assay (ELISA) [5,6], gas chromatography (GC) [7–9], liquid chromatography (LC) [10], gas chromatography–mass spectrometry (GC–MS) [11–14] and liquid chromatography–mass spectrometry (LC–MS and LC–MS/MS) methods [15–17] have been developed. However, each of ELISA methods can only determine one kind of the novel amide fungicides [5,6]. Moreover, the varieties of the novel amide fungicides are too few (the largest numbers of the novel amide fungicides are usually less than six) for many instrumental methods [11–14,16], which can simultaneous determine multiclass fungicides involved the novel amide fungicides in agricultural products with good sensitivity and accuracy. However, no development of the simultaneous determination method has been reported for those novel amide fungicides (boscalid, fenoxanil, mandipropamid, zoxamide, fluopicolide, fluopyram, and thifluzamide) with MRLs in fruits and vegetables. Recently, although Chen et al. have developed a LC–MS/MS method for determination of six novel amide fungicides
12
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Fig. 1. The chemical structures of 16 novel amide fungicides.
(silthiofam, boscalid, fluopicolide, mandipropamid, cyflufenamid, and mepanipyrim) in fruits and vegetables [19], it still cannot simultaneous determination of those amide fungicides with MRLs. Here, we develop and validate a simple and reliable confirmatory LC–MS/MS analytical method for the analysis of 16 novel amide fungicides including those amide fungicides with MRLs. Moreover, multi-walled carbon nanotubes (MWCNTs) as a dispersive solid phase extraction (dSPE) material has been firstly used for the simultaneous determination of these novel amide fungicides. The type and amount of MWCNTs, the pH value of the extract, the extraction time with the MWCNTs, and the type of eluent have been optimized in this study. After validation studies, the method is suitable for the routine determination of the 16 novel amide fungicides in fruits and vegetables. 2. Materials and methods
USA). Sodium hydroxide was purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Fenoxanil, boscalid, carpropamid, bixafen, silthiofam, thifluzamide (10 g/L in cyclohexane), fluopyram, tiadinil, ethaboxam (10 g/L in ACN), mandipropamid, fluopicolide, fenhexamid, cyflufenamid, zoxamide, furametpyr, and penthiopyrad were purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany). Five types of MWCNTs with similar lengths (10–30 m or 10–20 m) and different outer diameters (MWCNT01 50 nm) were purchased from Nanjing XF NANO Materials Tech Co. Ltd. (Nanjing, China). The specific surface areas (SSAs) of the MWCNTs were 500, 200, 110, 60, and 40 m2 /g for MWCNT01, MWCNT02, MWCNT03, MWCNT04, and MWCNT05, respectively. The water was purified with a Milli-Q reverse osmosis system (Millipore, Milford, Massachusetts, USA).
2.1. Materials and reagents
2.2. Standard solutions
Methanol (LC grade) and acetonitrile (ACN, LC grade) were obtained from Fisher Scientific (Fairlawn, USA). Formic acid (LC grade) was obtained from Tedia Company Inc. (Fairfield,
Individual stock solutions of 14 compounds (excluding ethaboxam and thifluzamide) were prepared in ACN at a concentration of 1000 g/ml. One mixed standard solution (10 g/ml) was
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
13
Table 1 LC–ESI–MS/MS parameters for 16 amide fungicides. Analyte
Precursor ion (m/z)
Daughter ion (m/z)
Dwell time (s)
Collision energy (eV)
Cone voltage (Volts)
Silthiofam Carpropamid Boscalid Fenoxanil Cyflufenamid Furametpyr Zoxamide Fenhexamid Fluopicolide Mandipropamid Ethaboxam Tiadinil Fluopyram Thifluzamide Bixafen Penthiopyrad
268.3 336.3 343.3 329.3 413.4 334.4 338.3 302.3 385.2 412.4 321.3 268.2 397.3 529.2 414.3 360.4
139.1, 252.0* 103.1* , 198.1 140.1* , 307.3 86.2, 302.3* 203.1* , 241.2 131.1* , 186.2 161.0, 189.1* 97.2* , 140.1 147.0, 175.1* 125.1, 328.3* 68.1, 183.2* 45.0, 101.1* 145.2* , 208.2 148.1, 489.1* 266.2, 394.3* 177.2, 276.2*
0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10 0.10, 0.10
18, 12 40, 12 20, 18 22, 12 38, 24 22, 18 38, 22 34, 22 50, 24 38, 14 34, 22 26, 18 54, 20 40, 28 24, 16 34, 14
14 24 28 22 24 26 30 36 30 24 38 32 34 30 26 26
*
Ion for quantification.
prepared in ACN by diluting each stock solution with ACN for the 14compounds (excluding ethaboxam, and thifluzamide). Then, one spiking mixed standard solution (5 g/ml) was prepared by redissolving the residue of 0.5 ml of thifluzamide solution (10 g/ml in cyclohexane), which was evaporated to dryness under nitrogen at room temperature, with 0.5 ml of ethaboxam solution (10 g/ml in ACN) and 0.5 ml of the above mixed standard solution (10 g/ml). The other two spiking mixed standard solutions of the 16 compounds (0.1 and 0.5 g/ml) were prepared by diluting the mixed standard solution (5 g/ml) with ACN. Sixteen individual standard working solutions (1.0 g/ml for the 16 compounds) for MS–MS optimization were prepared by diluting each stock solution with 0.10% formic acid in water/ACN (80:20, v/v). Seven mixed working standard solutions (0.25, 1.0, 2.5, 10.0, 25.0, 100, and 500 g/L) were prepared by diluting the spiking mixed standard solution (5.0 g/ml) with 0.10% formic acid in water/ACN (80:20, v/v). 2.3. Chromatographic conditions A Waters Acquity UPLC instrument (Milford, Massachusetts, USA) was used for analysis. The separation was performed on an Acquity BEH C18 column (2.1 mm × 100 mm, 1.7 m) maintained at 35 ◦ C. The mobile phase consisted of solvent A (0.10% formic acid in water) and solvent B (ACN). The LC linear gradient program was set as follows for the time/percent solvent B: 0.0/20, 0.5/20, 4.0/80, 5.5/80, 5.6/20, and 7.0/20 (total run time = 7 min). The flow rate was 0.30 ml/min. The injection volume was 10 l in full loop injection mode. 2.4. Detection conditions Detection was performed using a Waters XevoTM TQ triplequadrupole mass spectrometer in positive electrospray ionization (ESI) mode. The operating parameters were capillary voltage, 1000 V; source temperature, 150 ◦ C; desolvation gas temperature, 500 ◦ C; cone gas flow, 50 L/h; desolvation gas flow, 1000 L/h. The detection was carried out in the multiple reaction monitoring (MRM) modes. Argon was used as the collision gas, and the collision cell pressure was 2.8 mbar. Additional parameters are shown in Table 1. 2.5. Sample preparation Five grams of homogenized samples were transferred into a 50 ml polypropylene centrifuge tube with screw caps. For recovery assays, an aliquot of standard solution (500 l) was added and
the sample was vortexed for 15 s and allowed to stand at room temperature for 60 min. Subsequently, 10 ml of ACN (9.5 ml of ACN for recovery assays) was added. The sample was homogenized for 1 min using a high-speed blender (Ultra-Tyrrax T25; IKA, Germany) and 2 g of NaCl were added. After vortexing intensively for 1 min, the tube was centrifuged for 3 min at 5000 rpm. Then, an aliquot of supernatant (1.0 ml) was transferred to a clean 50 ml volumetric tube and was diluted with approximately 4.0 ml of water to a final volume of 5 ml (the final concentration of ACN was approximately 20%). Then, the pH of the diluted extract (20% ACN) was adjusted to the range of 3–6 with acetic acid. Afterwards, 10 mg of MWCNTs were added to the diluted extract (5.0 ml). After 1.0 min of extraction with shaking, the suspension was centrifuged at 9,000 rpm for 3 min. The supernatant was discarded, and 10 ml of acetone was added to the tube. Then, the tube was vortexed for 1.0 min and was centrifuged at 9,000 rpm for 2 min. An aliquot of the supernatant (5.0 ml, equivalent to 0.25 g of sample) was evaporated to dryness in a water bath at 40 ◦ C under nitrogen and was reconstituted in 2.5 ml of 0.10% formic acid in water/ACN (80:20, v/v). The resulting solution was filtered through a 0.22 m filter, and 10 l of the filtrate was injected into the LC. 2.6. Matrix effects To evaluate the matrix effects, seven concentrations (0.25, 1.0, 2.5, 10, 25, 100, and 500 g/L) of the 16 fungicides were analyzed in the pure solvent and in the blank sample after the sample preparation procedure. The slope ratio was calculated by comparing the external matrix matched calibration slope of the 16 fungicides with the solvent external calibration slopes. 2.7. Method validation To evaluate the suitability of the method for the determination of the 16 fungicides in fruits and vegetables, validation was performed with a conventional validation procedure that includes the following parameters: specificity, linearity, limit of detection (LOD), limit of quantification (LOQ), accuracy and precision. To verify the absence of interfering substances around the retention time of the 16fungicides, 20 blank samples (cabbage, celery, strawberry, and grape) were analyzed. Calibration curves were constructed using working standard solutions and by plotting the peak area of the quantitative ion pair at concentrations of 0.25, 1.0, 2.5, 10, 25, 100, and 500 g/L for the 16 fungicides. The LODs and LOQs were defined as the concentrations of the 16compounds that produced chromatographic peaks (qualitative
14
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Fig. 2. The effects of different types of MWCNTs on the adsorption rates of the 16 amide fungicides in celery original extract.
transitions) with signal-to-noise (S/N) greater than 3 and 10 for fruit and vegetable samples, respectively. The LOQs are estimated from the chromatogram corresponding to the lowest point used in the matrix-matched calibration. As no certified reference material was available, accuracy was determined as the mean recovery by using spiked blank samples. To this end, blank fruit and vegetable samples were spiked with the 16 fungicides at three different levels (0.01, 0.05, and 0.5 mg/kg). The repeatability was measured using the 15 spiked blank samples (n = 5 replicates per concentration level and analyzed in three independent analytical runs) and expressed as the coefficient of variation CVr . The within-laboratory reproducibility was measured using the 15 spiked blank samples (n = 5 replicates per concentration level and analyzed at three occasions with three different operators) and was expressed as the coefficient of variation CVR . 3. Results and discussion
current experiment, four identification points, one parent (1 point) and two transitions (each 1.5 points), were monitored. The optimal MS–MS conditions are described in Table 1. After optimization of MS parameters, the mobile phase composition was investigated under the chromatographic column (Acquity BEH C18 column, 2.1 mm × 100 mm, 1.7 m). The solution of water/ACN and 0.10% formic acid in water/ACN had been selected by Dong et al. [19] and Cui et al. [18] because they eluted more efficiently over time. After carefully investigation, it was found that there was not obvious difference between the responses of the 16amide fungicides using 0.10% formic acid in water/ACN and those of using water/ACN under the same gradient elution conditions. Moreover, it was found that only eight fungicides can be eluted from the column under the gradient conditions described at Section 2.3 using 0.10% formic acid in water/methanol as the mobile phase. So, 0.10% formic acid in water/ACN was selected as the mobile phase in the current study.
3.1. Optimization of instrumentation conditions
3.2. Optimization of sample preparation conditions
For some of established LC–MS/MS method, negative ion ESI and positive ion APCI had been used to determine some of amide fungicides with good sensitivity [11–19]. However, positive ion ESI is typically used for simultaneous determination of amide fungicides among most of established LC–MS/MS methods, because ESI source is easy to maintain and the [M + H]+ ion forms easily formed due to the imino group on them [19]. In the present study, the appropriate diagnostic ions with good sensitivity had been obtained for the 16 amide fungicides using IntelliStart software in ESI positive mode (Table 1). So, ESI positive mode had been chosen in this study. Identification points (IPs) for the confirmation and identification had been introduced by the European Commission Decision 2002/657/EC [20]. According to 2002/657/EC and LC–MS/MS, four identification points are required for most of compounds. In the
For most of established LC–MS/MS method, ACN had been used to extract amide fungicides in fruits and vegetables [15,18,19]. Thus, the extraction efficiency in this study was studied with ACN at different volume (10, 15, 20, 25, and 30 ml). All of amide fungicides were added to the vegetables and fruits (cabbage, celery, strawberry, and grape) and were extracted according the extraction method described in sample preparation section. It was found that the recoveries were above 95% for all of amide fungicides. Considering the environmental protection, which requires minimizing the use of undesirable organic solvents, 10 ml of ACN was chosen to extract amide fungicides from vegetables and fruits. Optimization of dSPE procedure is an important process to get good purification efficiency and full recovery. It usually involves the type and amount of MWCNTs, the pH value of the extract, shaking
Fig. 3. The effects of different types of MWCNTs on the adsorption rates of the 16 amide fungicides in diluted celery extract containing 50% of water.
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
15
Fig. 4. The effects of different types of MWCNTs on the adsorption rates of the 16 amide fungicides in diluted celery extract containing 80% of water.
extraction time with MWCNTs, and the type of eluting solvent and its volume. The recovery was calculated with a matrix-matched external standard method during optimization of sample preparation. All of experiments were performed in three replicates (n = 3). 3.2.1. Effects of type of MWCNTs A series of experiments were executed to investigate the influence of the type of MWCNTs. Firstly, the effects of the types of MWCNTs were investigated with 5 ml of the extract of celery and 50 mg of MWCNTs. The spiking concentration in the extract was 25 g/L and the shaking time was 2 min. The adsorption rate was acquired by using the formula: adsorption rate (%) = 100−(the concentration of the 16 amide fungicides in the celery extract after MWCNT adsorption/the initial concentration of the 16 amide fungicides in the celery extract) × 100. The results are shown in Fig. 2. It was found that the highest adsorption rates were observed for all of the 16 fungicides with MWCNT01, which have the largest SSAs. However, most of the adsorption rates were below 80% for the 16 fungicides when MWCNT01 had been used. In order to improve the adsorption efficiency of MWCNTs, the celery extract was diluted with water because the adsorption capacity of CNTs for hydrophobic compounds, such as the 16 amide fungicides, usually increases as the water concentration increases in mixtures of water and an organic solvent [21]. Significantly higher adsorption rate were obtained with the diluted celery extract containing 50% of water for these amide fungicides (Fig. 3). Surprisingly, all of the adsorption rates approached 100% except silthiofam when MWCNT01, MWCNT02, MWCNT03 and MWCNT04 had been used (Fig. 4) with the diluted extract containing 80% of water. Considering MWCNT01 has the strongest adsorption ability for these fungicides, MWCNT01 and the diluted celery extract with 80% of water were chosen to carry out further experiments. 3.2.2. Effects of amount of MWCNTs To explore the effects of the amount of MWCNT01 on the adsorption rates of the 16 fungicides, different amounts of MWCNT01 (5, 7.5, 10, 15, and 25 mg) were used. Other experimental conditions were the same as above. Results are shown in Fig. 5. It was found that the adsorption rates of most of fungicides were approached 100% at 10 mg. Moreover, the adsorption rates of sithiofam,
fluopyram, and thifluzamide were above 90% at 10 mg. To obtain the best results using the least amount of MWCNTs, 10 mg of MWCNTs were chosen for the present study. 3.2.3. Effects of pH The effects of the celery diluted extract (20% ACN) with different pH values (3, 4, 5, 6, 7, 8, and 9) and without pH adjustment (pH = 5.8) on the adsorption rate were investigated with 10 mg of MWCNT01. It was found that there were no obvious differences in the adsorption rates of the 16 amid fungicides when the pH of the diluted feed extract (20% ACN) varied from 3 to 6 (Fig. 6). However, there were slightly reductions in the adsorption rates of the 16 amide fungicides when pH values were greater than 7.0. For this phenomenon, it may be the reason that a decrease of hydrophobicity of the 16 amide fungicides had occurred when the pH values of diluted celery extract is greater than 7.0 because we had found that there were distinct correlations between the adsorption rate and hydrophobic properties for sulfonamides and resorcylic acid lactones [21,22]. In order to ensure the applicability of the method for all of samples, the diluted extract (20% ACN) had been required to adjust pH value to the range of 3–6 in the present study. 3.2.4. Effects of extraction time The effects on the recovery caused by different shaking extraction times (1, 2, 5, and 10 min) were investigated with 10 mg of MWCNT01 for the diluted celery extract (80% water). The adsorption on the MWCNTs was observed to be a fast process for the 16 fungicides because adsorption rates of most of fungicides were 100% when the extraction time was longer than 1 min. Moreover, no significant differences in the adsorption rates were observed for any of the 16 fungicides. The results observed for the 16 fungicides were similar with many other organic chemicals, such as resorcylic acid lactones [23], nitroaromatic compounds [24], organophosphorus pesticides [25], quinolone antibiotics [26], and sulfonamides [21]. Therefore, a 1 min extraction time was chosen for the subsequent experiments. 3.2.5. Effects of the type of eluent Due to the 16amide fungicides are easily dissolved in organic solvents, the effects of different solvents (ACN, methanol, acetone,
16
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Fig. 5. The effects of different amount of MWCNT01 on the adsorption rates of the 16 amide fungicides in the diluted celery extract containing 80% of water.
Fig. 6. The effects of different pH values on the adsorption rates of 16 amide fungicides in celery diluted extract.
Fig. 7. The effects of different solvents as eluents on the recoveries of the 16 amide fungicides.
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20 Table 2 The linearity and regression coefficients of standard curves of 16 amide fungicides by LC–MS/MS. Analyte
Linear equation
r2
Silthiofam Carpropamid Boscalid Fenoxanil Cyflufenamid Furametpyr Zoxamide Fenhexamid Fluopicolide Mandipropamid Ethaboxam Tiadinil Fluopyram Thifluzamide Bixafen Penthiopyrad
y = 8237.3x + 7889.1 y = 534.8x + 432.8 y = 2433.7x + 3559.3 y = 2016.3x + 3289.2 y = 1367.4x + 479.4 y = 2277.2x + 2433.3 y = 2653.1x + 1876.4 y = 1783.3x + 1107.2 y = 4127.8x + 3267.4 y = 3765.3x + 1877.3 y = 4578.2x + 2865.9 y = 1057.3x + 1077.42 y = 12779.1x + 7187.4 y = 202.6x + 133.1 y = 789.3x + 332.7 y = 7635.2x + 4339.0
0.9980 0.9987 0.9975 0.9979 0.9977 0.9982 0.9973 0.9982 0.9989 0.9990 0.9974 0.9979 0.9988 0.9984 0.9986 0.9983
and ethyl acetate) on elution recovery were carefully investigated at a spiking concentration of 50 g/L under the optimized MWCNT conditions (described above). For convenience, the volume of the three solvents was firstly fixed at 5 ml. The results are shown in
17
Fig. 7. From Fig. 7, it was found that acetone was the optimum solvent as eluent for elution of most of amide fungicides from MWCNTs. Moreover, the elution recoveries of most of amide fungicides were more than 80% when 5 ml of acetone were used except tiadinil (51.5%) and boscalid (73.9%). In order to improve elution efficiency for tiadinil and boscalid, 10 ml of acetone was further used to investigate. It was found that the elution recoveries of tiadinil and boscalid were reached 75.3 and 87.9%, respectively. These results of tiadinil and boscalid are within the regulations of document no. SANCO 12495-2011 [27]. So, 10 ml of acetone was finally selected as eluent in the present study. Under the above optimized conditions, the pre-treatment procedure for the MWCNTs was simple and fast. Compared to an SPE cartridge that costs at least $3 USD per cartridge, the cost of the MWCNTs required for one sample is approximately $0.10 USD. In addition, the total time required for the analysis of one sample was only approximately 25 min and less than the time (at least 50 min) of the method developed by Chen et al. [19]. 3.3. Matrix effects Co-elution of undetected matrix components may enhance or reduce the ion intensity of the analytes and affect the accuracy and
Table 3 LOD and LOQ (g/kg) obtained for 16 amide fungicides in vegetables and fruits by LC–MS/MS. Analyte
Silthiofam Carpropamid Boscalid Fenoxanil Cyflufenamid Furametpyr Zoxamide Fenhexamid Fluopicolide Mandipropamid Ethaboxam Tiadinil Fluopyram Thifluzamide Bixafen Penthiopyrad
Cabbage
Celery
Strawberry
Grape
LOD
LOQ
LOD
LOQ
LOD
LOQ
LOD
LOQ
0.2 0.5 0.4 0.2 0.8 0.3 0.4 1.5 0.1 0.2 0.1 2.5 0.1 1.9 0.3 0.4
0.8 1.6 1.5 0.8 2.6 1.1 1.3 5.0 0.3 0.5 0.3 8.3 0.2 6.2 1.0 1.2
0.3 0.4 0.4 0.3 0.7 0.4 0.5 1.9 0.1 0.1 0.1 1.9 0.1 2.5 0.3 0.3
0.9 1.5 1.4 0.9 2.4 1.4 1.5 6.3 0.2 0.4 0.4 6.2 0.2 8.3 1.1 1.1
0.2 0.5 0.5 0.3 1.0 0.4 0.4 1.5 0.1 0.2 0.1 1.9 0.1 1.9 0.3 0.4
0.7 1.6 1.6 0.9 3.2 1.4 1.5 5.0 0.3 0.5 0.3 6.2 0.2 6.2 1.0 1.3
0.2 0.5 0.4 0.3 0.9 0.5 0.4 1.5 0.1 0.2 0.1 2.5 0.1 1.9 0.3 0.4
0.7 1.6 1.5 0.9 3.1 1.5 1.4 5.0 0.3 0.5 0.3 8.3 0.2 6.2 1.0 1.2
Fig. 8. The representative MRM chromatograms of a working standard solution (0.25 g/L for the 16 amide fungicides, (a) a blank celery sample (b) and a spiked blank celery sample (0.01 mg/kg for the 16 amide fungicides, c).
18
Table 4 Mean recoveries and variation coefficients (CVr and CVR ) of 16 amide fungicides from celery and grape samples by LC–MS/MS. Analyte
Silthiofam
Carpropamid
Boscalid
Fenoxanil
Cyflufenamid
Zoxamide
Fenhexamid
Fluopicolide
Mandipropamid
Ethaboxam
Tiadinil
Fluopyram
Thifluzamide
Bixafen
Penthiopyrad
0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5 0.01 0.05 0.5
grape
Celery Mean recovery (%, n = 5) 87.4 (4.1) 83.9 (4.5) 88.1 (4.2) 92.7 (4.8) 94.5 (4.4) 91.4 (3.2) 83.7 (4.2) 85.1 (4.0) 80.4 (4.7) 91.6 (4.1) 93.2 (4.7) 94.0 (4.2) 96.7 (4.6) 94.2 (4.2) 93.7 (4.0) 92.4 (4.4) 95.1 (3.8) 90.9 (4.2) 94.5 (4.2) 93.2 (4.5) 96.1 (4.6) 93.9 (4.9) 90.1 (4.3) 93.5 (4.3) 95.1 (4.7) 93.7 (4.6) 96.2 (4.9) 91.3 (4.9) 88.9 (4.4) 93.2 (4.3) 93.7 (4.4) 90.3 (4.8) 87.4 (4.7) 80.1 (4.6) 73.3 (5.1) 78.1 (3.8) 93.7 (4.8) 93.2 (4.5) 94.6 (4.3) 90.6 (4.8) 93.4 (4.5) 91.3 (4.6) 88.4 (4.6) 89.2 (4.5) 87.7 (4.2) 96.7 (4.3) 94.6 (4.6) 98.1 (4.4)
85.7 (5.3) 82.2 (3.9) 89.3 (4.8) 93.2 (3.7) 92.8 (3.8) 94.6 (3.3) 86.4 (3.9) 87.2 (4.3) 86.9 (3.9) 94.9 (3.9) 93.1 (4.5) 92.5 (3.8) 94.3 (5.1) 95.9 (4.7) 98.1 (3.6) 92.8 (4.8) 91.0 (4.3) 94.5 (4.1) 94.8 (5.1) 92.7 (4.6) 92.0 (4.3) 96.1 (4.6) 93.9 (4.7) 95.2 (3.6) 88.4 (4.4) 92.2 (4.3) 94.1 (4.6) 87.8 (5.3) 90.7 (4.2) 92.9 (4.2) 89.1 (4.9) 86.9 (4.1) 91.3 (4.8) 76.4 (3.8) 81.3 (4.4) 70.8 (4.6) 91.9 (4.7) 93.6 (4.9) 93.8 (3.8) 93.2 (4.8) 92.4 (4.1) 90.1 (4.3) 89.9 (4.9) 90.1 (4.4) 91.4 (2.9) 94.7 (5.3) 95.9 (4.1) 94.7 (4.9)
88.3 (5.0) 83.9 (3.9) 84.2 (4.2) 90.9 (4.7) 95.1 (4.1) 96.3 (3.7) 88.1 (4.6) 81.8 (4.7) 83.2 (4.4) 96.1 (4.8) 94.2 (3.7) 92. 8(4.6) 92.8 (5.6) 93.7 (4.3) 95.1 (4.2) 93.7 (4.9) 92.9 (4.4) 88.9 (4.6) 94.3 (4.7) 93.8 (4.5) 91.6 (4.2) 93.7 (5.3) 94.3 (4.6) 92.9 (4.8) 89.5 (5.9) 93.7 (4.7) 93.2 (4.3) 91.6 (4.8) 92.2 (4.2) 90.8 (4.7) 93.2 (5.1) 90.1 (4.4) 86.9 (4.5) 77.2 (4.3) 72.4 (3.9) 76.9 (4.1) 90.6 (3.9) 92.2 (5.3) 94.7 (4.6) 93.2 (4.1) 87.6 (4.4) 89.2 (4.6) 86.9 (5.4) 90.2 (4.8) 92.3 (4.3) 97.8 (4.7) 96.4 (3.8) 98.5 (4.5)
CVr (%)
CVR (%)
Mean recovery (%, n = 5)
4.9 4.2 4.4 4.4 4.0 3.5 4.3 4.6 4.3 4.6 4.4 4.2 5.0 4.5 4.0 4.6 4.2 4.3 4.6 4.5 4.3 5.0 4.5 4.4 4.9 4.4 4.5 5.0 4.4 4.2 4.5 4.3 4.3 4.3 4.6 5.0 4.4 3.9 4.2 4.4 4.5 4.2 3.9 4.3 4.0 4.6 4.2 4.4
6.7 5.9 5.8 6.1 5.8 6.0 5.8 6.7 6.2 6.0 5.7 5.9 6.7 5.8 5.4 6.7 5.8 5.4 6.4 6.2 6.3 7.1 6.0 5.9 7.6 6.4 6.2 6.1 5.8 6.3 6.4 6.2 6.0 6.7 6.6 7.1 6.4 6.8 6.1 6.7 6.2 6.1 7.2 6.6 5.9 7.8 6.4 6.2
86.3 (4.5) 81.7 (4.1) 84.4 (3.8) 93.5 (4.2) 92.3 (3.9) 95.1 (4.7) 88.1 (4.4) 83.5 (4.7) 81.9 (4.1) 96.1 (3.9) 94.2 (4.6) 94.8 (4.1) 94.9 (5.3) 92.3 (4.6) 95.1 (3.8) 96.8 (4.9) 93.1 (4.3) 92.6 (4.2) 93.9 (4.4) 95.4 (4.6) 93.7 (4.8) 93.3 (4.6) 95.9 (4.1) 94.2 (4.5) 94.9 (4.8) 93.2 (4.4) 94.5 (4.3) 92.4 (5.4) 90.7 (4.7) 89.6 (4.8) 88.6 (5.0) 92.4 (5.2) 90.7 (4.5) 72.4 (4.3) 75.7 (4.6) 81.3 (4.0) 89.2 (5.2) 92.6 (4.8) 93.3 (4.7) 88.9 (5.0) 90.7 (4.7) 94.1 (4.6) 93.2 (4.9) 91.6 (4.3) 92.7 (4.4) 95.7 (4.4) 98.1 (4.6) 94.6 (4.1)
87.3 (4.3) 86.9 (4.8) 83.8 (4.1) 92.4 (4.6) 98.7 (4.8) 96.4 (4.6) 84.1 (3.9) 83.7 (4.4) 85.9 (3.6) 89.7 (4.2) 93.5 (4.6) 92.6 (4.9) 95.3 (4.4) 94.1 (4.8) 97.2 (4.6) 91.7 (5.6) 93.5 (4.7) 94.1 (4.9) 90.6 (4.9) 92.4 (4.3) 93.7 (4.2) 92.8 (4.6) 91.6 (5.0) 94.5 (4.7) 91.8 (4.9) 94.3 (4.6) 97.7 (4.7) 93.2 (4.8) 90.6 (4.3) 92.3 (4.4) 92.8 (4.7) 87.4 (4.5) 90.9 (4.5) 78.1 (4.3) 73.3 (4.6) 77.9 (4.3) 92.0 (4.4) 95.1 (4.7) 93.8 (4.3) 93.2 (4.7) 87.3 (4.4) 91.5 (4.7) 91.5 (5.0) 90.4 (4.8) 88.6 (4.5) 97.1 (3.6) 93.8 (4.3) 95.2 (4.5)
82.7 (5.1) 85.3 (4.3) 88.2 (4.7) 93.1 (3.9) 94.2 (4.5) 93.9 (3.4) 81.5 (4.6) 82.9 (4.8) 84.1 (4.1) 94.6 (4.8) 93.7 (4.1) 96.1 (3.9) 90.7 (4.1) 94.3 (4.6) 95.7 (4.2) 92.9 (4.5) 93.7 (4.3) 90.8 (4.1) 94.2 (4.6) 96.3 (4.4) 93.2 (4.6) 92.7 (4.3) 93.8 (4.6) 94.7 (4.1) 94.3 (5.1) 92.9 (4.7) 95.8 (4.5) 89.5 (5.4) 88.1 (4.9) 94.2 (4.1) 90.8 (4.8) 86.4 (5.2) 92.1 (4.6) 79.8 (4.4) 78.0 (4.7) 73.4 (4.5) 93.9 (4.8) 94.1 (4.3) 92.5 (4.6) 90.7 (5.2) 92.8 (4.6) 91.7 (4.3) 92.4 (4.7) 91.5 (4.3) 87.6 (3.3) 97.6 (5.2) 95.2 (4.5) 97.3 (4.2)
CVr (%)
CVR (%)
4.6 4.2 4.5 4.3 4.4 4.4 4.2 4.6 3.9 4.6 4.4 4.7 4.7 4.6 4.3 5.1 4.4 4.7 4.7 4.5 4.4 4.4 4.7 3.2 4.6 4.3 4.3 5.0 4.5 4.4 4.7 4.7 4.0 4.8 4.3 5.4 4.9 4.4 3.3 4.8 4.6 4.3 4.5 4.4 4.0 4.6 4.5 4.0
6.4 5.8 6.5 6.3 5.7 6.2 6.1 6.6 5.9 6.2 6.1 6.6 6.5 6.3 6.1 6.9 6.2 6.2 6.4 6.5 6.1 6.0 6.4 6.0 6.2 6.0 6.2 6.9 6.3 6.1 6.4 6.6 6.0 6.4 6.2 5.7 6.7 6.3 6.4 6.8 6.5 6.2 6.6 6.1 5.8 6.5 6.0 5.8
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
Furametpyr
Fortified concentration (mg/kg)
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
reproducibility of the quantitative analysis when ESI-MS was used. So, the matrix effect was evaluated by comparing the ratios of the external calibration slopes for the matched matrix to the external calibration slopes for the solvent. The numbers of the amide fungicides, which the slope ratios were between 0.80–1.20 before purification with MWCNT01, were 11, 10, 12, and 14 for cabbage, celery, strawberry, and grape samples, respectively. Moreover, the ranges of slope ratios were 0.68–1.33, 0.59–1.42, 0.77–1.36, and 0.82–1.40 for the 16 amide fungicides in cabbage, celery, strawberry, and grape samples, respectively. After purification with MWCNT01, all of slope ratios were between 0.80–1.20 (0.90–1.02 for cabbage samples, 0.93–1.03 for celery samples, 0.92–1.04 for strawberry samples, and 0.94–1.02 for grape samples) in the above four kinds of samples. From the calculated matrix effect results, it was concluded that the matrix effect had decreased after purification with MWCNTs and there was a slight matrix effect for the 16 amide fungicides after purification. However, a strong matrix effect had been discovered for qualitative ions of two amide fungicides (carpropamid and thifluzamide). So, the matrix standard solutions were chosen for confirmation in this study. 3.4. Method validation 3.4.1. Linearity The calibration curves of working standard solutions were obtained by plotting the peak area of the quantitative ion pairs over the range from 0.25 to 500 g/L. The results are reported in Table 2. The correlation coefficients (r2 ) of the calibration curves were greater than 0.9973. To acquire more accurate quantification, single point calibration with the matrix matched standard solutions (1.0, 2.5, and 25 g/L for 0.01, 0.05 and 0.5 mg/kg spiking level, respectively) had been used in the recovery experiments according to the description of EURACHEM Guide [28]. Moreover, the standard curve point closest to the measured value of the “real” sample could be chosen for single point calibration when analyzing “real” samples. The MRM chromatograms of the standard solution are shown in Fig. 8a. 3.4.2. Specificity Fig. 8b and c indicated that there were no interfering peaks from endogenous compounds at the retention times of the 16 compounds. Therefore, the optimized pre-treatment procedure coupled to UHPLC–MS/MS provided a clean chromatogram without interferences. 3.4.3. LOD and LOQ In the present study, the LODs and LOQs are estimated from the chromatogram corresponding to the lowest point used in the matrix-matched calibration (0.25 g/L). Each matrix standard was injected three times. LOD, calculated as S/N ratio = 3, for the 16 amide fungicides was ranged from 0.1 to 2.5 g/kg in all matrices (Table 3), respectively. LOQ, calculated as S/N ratio = 10, for the 16 amide fungicides was ranged from 0.2 to 8.3 g/kg in all matrices (Table 3), respectively. 3.4.4. Accuracy and precision The recovery and reproducibility of the method were measured by spiking the blank samples with the 16 amide fungicides at three different concentrations (0.01, 0.05, and 0.5 mg/kg) on three separate occasions. The results are shown in Table 4 according to the quantitative method described in Section 3.4.1 using the matrix matched standard solutions, which had been prepared by spiking the analytes into the blank sample after the extraction/cleanup procedure. The average recoveries, repeatability, and reproducibility varied from 72.4 to 98.5%, from 3.2 to 5.4% (CVr), and from 5.4 to 7.8% (CVR ), respectively. These recoveries and CVs demonstrate that
19
this method can achieve a satisfactory accuracy and precision for the 16 amide fungicides residue analysis in vegetables and fruits. Fig. 8b and c shows the representative typical chromatograms of the 16 amide fungicides of the celery blanks and spiked samples. 3.5. Applications of the method To evaluate the effectiveness and applicability of this method for determining the 16 amide fungicides, real samples (45 grape samples, 14 celery samples, 10 strawberry samples, and 12 cabbage samples) had been obtained from several markets in Ningbo. Boscalid was detected in 4 grape samples ranged from 0.024 to 0.18 mg/kg. The presence of boscalid at these levels in grape does not pose a threat to the consumer since they are below the MRLs established by EU (2.0 mg/kg for grape) and China (3.0 mg/kg for strawberry and no rule for grape). No other amide fungicides were detected in these samples. 4. Conclusions In the present study, a method for the determination of 16 amide fungicides in vegetables and fruits was established using MWCNTs as a sorbent for dSPE coupled with LC–MS/MS. The pre-treatment procedure using MWCNTs was very simple and fast compared to previously reported methods. The current method was validated with spiked vegetables and fruit samples and good recoveries with good CVs were obtained. The LODs and LOQs were found to be sufficiently low to determine the residues of 16 amide fungicides samples. Moreover, the method was successfully applied to the routine analysis of amide fungicides in vegetable and fruit samples. Acknowledgements The Project in Agriculture of Ningbo, (No. 2013C11003, No. 2013NK33 and No. 2014C10058) and National Agricultural Products Risk Assessment Project (GJFP2014011) provided financial support for this work. References [1] G.A. Carter, J.L. Huppatz, R.L. Wain, Ann. Appl. Biol. 84 (1976) 33. [2] National Food Safety Standard-Maximum Residue Limits for Pesticides in Food, National Health and Family Planning Commission of the People’s Republic of China, Ministry of Agriculture of the People’s Republic of China, Beijing, 2014, GB 2763-2014. [3] European Union, European Union database on Pesticide on MRLs. (http://ec.europa.eu/sanco pesticides/public/?event=substance.selection). [4] United States Department of Agriculture Foreign Agricultural Service-Pesticide MRL Database. (http://www.mrldatabase.com/default.cfm?selectvetdrug=0). [5] F.C. Esteve-Turrillas, A. Abad-Fuentes, J.V. Mercader, Food Chem. 124 (2011) 1727. [6] J.V. Mercader, A. Abad-Fuentes, J. Agric. Food Chem. 57 (2009) 5129. [7] D.T. Likas, N.G. Tsiropoulos, G.E. Miliadis, J. Chromatogr. A 1150 (2007) 208. [8] V.T. Gajbhiye, Suman Gupta, Anshu Singh, R.K. Gupta, T.K. Das, Pestology 26 (2002) 54. [9] E.G. Amvrazi, A.T. Papadi-Psylloe, N.G. Tsiropoulos, Intern. J. Environ. Anal. Chem. 90 (2010) 245. [10] A. Melo, A. Aguiar, C. Mansilha, O. Pinho, I.M.P.L.V.O. Ferreira, Food Chem. 130 (2012) 1090. [11] L. Lagunas-Allué, J. Sanz-Asensio, M.T. Martínez-Soria, J. AOAC Int. 95 (2012) 1511. [12] L. Lagunas-Allué, J. Sanz-Asensio, M.T. Martínez-Soria, J. Chromatogr. A 1270 (2012) 62. [13] A. Angioni, L. Porcu, F. Dedola, F. Dedola, Pest Manag. Sci. 68 (2012) 543. [14] R.M. Gonzalez-Rodriguez, B. Cancho-Grande, J. Simal-Gandara, J. Chromatogr. A 1216 (2009) 6033. [15] F. Dong, X. Chen, X. Liu, J. Xu, Y. Li, W. Shan, Y. Zheng, J. Chromatogr. A 1262 (2012) 98. [16] C. Coscollà, V. Yusà, M.I. Beser, A. Pastor, J. Chromatogr. A 1216 (2009) 8817. [17] M. Hengel, B. Hung, J. Engebretson, T. Shibamoto, J. Agric. Food Chem. 51 (2003) 6635. [18] S.H. Cui, J.L. Qian, H. Duan, J.F. Liu, L.M. Lin, K.Y. Wang, Chin. J. Anal. Chem. 39 (2011) 1836.
20
Y.-L. Wu et al. / J. Chromatogr. B 989 (2015) 11–20
[19] X. Chen, Z. Li, Z. Cao, J. Gong, X. Cao, M. Chen, Chin. J. Chromatogr. 31 (2013) 954. [20] European Commission, Decision 2002/657/EC (2002), Off. J. Eur. Commun. (2002) L221. [21] H. Peng, B. Pan, M. Wu, R. Liu, D. Zhang, D. Wu, B. Xing, J. Hazard. Mater. 342 (2012) 211–212. [22] X.L. Hou, Y.L. Wu, T. Yang, X.D. Du, J. Chromatogr. B 929 (2013) 107. [23] Y.F. Ying, Y.L. Wu, Y. Wen, T. Yang, X.Q. Xu, Y.Z. Wang, J. Chromatogr. A 1307 (2013) 41.
[24] X.E. Shen, X.Q. Shan, D.M. Dong, X.Y. Hua, G. Owens, J. Colloid Interface Sci. 330 (2009) 1. [25] M.Á. González-Curbelo, M. Asensio-Ramos, A.V. Herrera-Herrera, J. HernándezBorges, Anal. Bioanal. Chem. 404 (2012) 183. [26] A.V. Herrera-Herrera, L.M. Ravelo-Pérez, J. Hernández-Borges, M.M. Afonso, J. Palenzuela, M.A. Rodríguez-Delgado, J. Chromatogr. A 1218 (2011) 5352. [27] European Commission, Document No. SANCO/12495/2011, 2011. [28] EURACHEM Guide, the Fitness for Purpose of Analytical Methods, first ed, Teddington, Middlesex, 1998.
