Food Chemistry 176 (2015) 197–204

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Application of magnetic solvent bar liquid-phase microextraction for determination of organophosphorus pesticides in fruit juice samples by gas chromatography mass spectrometry Lijie Wu a, Ying Song a, Mingzhu Hu a, Hanqi Zhang a, Aimin Yu a, Cui Yu b, Qiang Ma c, Ziming Wang a,⇑ a b c

College of Chemistry, Jilin University, 2699 Qianjin Street, Changchun 130012, China School of Chemical and Materials Engineering, Yanching Institute of Technology, Langfang 065201, China Chinese Academy of Inspection and Quarantine, Beijing 100123, China

a r t i c l e

i n f o

Article history: Received 8 August 2014 Received in revised form 9 December 2014 Accepted 13 December 2014 Available online 24 December 2014 Keywords: Magnetic solvent bar liquid-phase microextraction Organophosphorus pesticides Fruit juice Gas chromatography mass spectrometry

a b s t r a c t A simple, rapid and sensitive sample pretreatment technique, magnetic solvent bar liquid-phase microextraction (MSB-LPME) was developed for extracting organophosphorus pesticides from fruit juice. The analytes were extracted from the sample to the organic solvent immobilized in the fiber. The magnetic solvent bar not only can be used to stir the samples but also extract the analytes. After extraction, the analyte-adsorbed magnetic solvent bar can be readily isolated from the sample solution by a magnet, which could greatly simplify the operation and reduce the whole pretreatment time. The bar was eluted with methanol. The elute was evaporated to dryness and residue was dissolved in hexane. Several experimental parameters were investigated and optimized, including type of extraction solvent, number of magnetic solvent bar, extraction temperature, extraction time, salt concentration, stirring speed, pH and desorption conditions. The recoveries were in the range of 81.3–104.6%, and good reproducibilities were obtained with relative standard deviation below 6.1%. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Organophosphorus pesticides (OPPs) are widely used in China for agricultural activities because of their high effectively and relatively low price (Jaleel, Gopi, Manivannan, & Panneerselvam, 2008; Xiong & Hu, 2008). However, the overuse of OPPs may lead to the contamination in agricultural products and, hence, in derivate food commodities, i.e., wine, fruit juices and so on (Albero, Sánchez-Brunete, & Tadeo, 2003; Liu, Hashi, Song, & Lin, 2005; Schellin, Hauser, & Popp, 2004; Wong, Webster, Halverson, Hengel, Ngim & Ebeler, 2003). The presence of OPP residues in foods become a health hazard to humans because some of them have a high acute toxicity due to the prevention of neural impulse transmission by their inhibition of cholinesterase (Barata, Porte & Solayan, 2004; Fu, Liu, Hu, Zhao, Wang & Wang, 2009; Sogorb & Vilanova, 2002; Vidair, 2004). Fruit juice drinks are the favorite nutritional supplement. They are receiving considerable attention because they are comprised of highly abundant nutrition (e.g. vitamins and minerals) (Picó & ⇑ Corresponding author. Tel.: +86 431 85168399; fax: +86 431 85112355. E-mail address: [email protected] (Z. Wang). http://dx.doi.org/10.1016/j.foodchem.2014.12.055 0308-8146/Ó 2014 Elsevier Ltd. All rights reserved.

Kozmutza, 2007). In addition, children are the largest group for the consumption of fruit juice products (Cortés-Aguado, SánchezMorito, Arrebola, Garrido Frenich, & Martínez Vidal, 2008). So pesticide residues may be transferred from fruit into juice, being a significant route to human exposure (Romero-González, Frenich & Vidal, 2008). The European Union Directive on drinking water quality (98/83/EC) established a maximum allowed concentration of 0.1 ng mL1 for each individual pesticide and 0.5 ng mL1 for total pesticides in drinking water. Although many regulators (e.g. European Union) have not set maximum residue limits for pesticides in fruit juices till now, it is also of great importance to develop rapid, highly sensitive, and easily operated methods to monitor pesticide residues in fruit juices. A variety of analytical methods have been reported for the simultaneous determination of multiple pesticides in juice matrices. Liquid–liquid extraction (LLE) (Eva & Manuel, 2003; Kolbe & Andersson, 2006; Sannino, 2007) and solid phase extraction (SPE) (Albero, Sánchez-Brunete, & Tadeo, 2005; Topuz, Özhan & Alpertunga, 2005) are applied to the extraction of pesticides from juice samples. However, LLE is time-consuming and requires large volumes of organic toxic solvents. SPE uses much less solvent than LLE but the column needs pretreatment and

198

L. Wu et al. / Food Chemistry 176 (2015) 197–204

can be relatively expensive and labor intensive. Nowadays, to overcome those problems, several microextraction techniques by reducing organic solvent consumption and simplifying the sample preparation techniques have been reported as alternatives to conventional sample preparation procedures, such as solid phase microextraction (SPME) (Cortés-Aguado et al., 2008; Simplı´cioa & Boas, 1999), stir bar sorptive extraction (SBSE) (Farajzadeh, Djozan, Nouri, Bamorowat, & Shalamzari, 2010; Zuin, Schellin, Montero, Yariwake, Augustod & Popp, 2006) and liquid-phase microextraction (LPME) (Lambropoulou & Albanis, 2007; Pintoa, Sontag, Bernardino, & Noronha, 2010). However, SPME and SBSE are time-consuming and the coated fibers or bars are generally expensive and easily destroyed. LPME is a solvent miniaturized procedure of LLE, which has a good preconcentration ability. Hollow fiber liquid phase microextraction (HF-LPME) provides mechanical stability and protection to the organic solvent because the use of hollow fiber membrane, which is simple, effective, low cost, uses microliters of organic solvents and provides excellent sample clean-up ability, obtaining very clean extracts. Hollow fiber liquid phase microextraction (HF-LPME) is an LPME-based technique and has more advantages than LPME (Barahona, Gjelstad, Pedersen-Bjergaard &, Rasmussen, 2010; Bedendo, Jardim, & Carasek, 2012; Bolaños, Romero-González, Frenich, & Vidal, 2008). First, HF-LPME uses microliters of organic solvent which is placed in a hollow fiber, and then the analytes in the aqueous sample can be extracted into the organic solvent; second, the hollow fiber also plays a role as an excellent filter due to the large molecules cannot permeate through the pore in the hollow fiber; third, HF-LPME device is simple cost-efficient and easy operation. Thus, HF-LPME has attracts increasing attentions. So far, HF-LPME devices have highly flexible formats. The most common format is the hollow fiber fixed to a microsyringe during extraction so that the final extract may be withdrawn into the syringe and analyzed directly (Ouyang & Pawliszyn, 2006; Zhao, Zhu, & Lee, 2002; Zhu, Zhu, & Lee, 2001). Pedersen-Bjergaard and Rasmussen (1999) first reported the U-shaped HF-LPME (Pedersen-Bjergaard & Rasmussen, 1999). Jiang and Lee (2004) proposed solvent bar microextraction (SBME). The organic extracting solvent (1-octanol) was confined within the hollow fiber membrane (sealed at both ends) that was placed in a stirred aqueous sample solution. Due to the free movement of the solvent bar in the sample solution, rapid extraction equilibrium can be achieved. Hultgren, Larsson, Nilsson, and Jönsson (2009) fixed the fiber on a 4 cm metal rod from it end for keeping it at a fixed position in the stirring sample. Yu et al. proposed dual solvent-stir bars microextraction (DSSBME), a stainless-steel wire was used to fix the hollow fibers and the device could stir by itself (Yu, Liu, Lan, & Hu, 2008). Although the devices of HF-LPME are developed rapidly, few of devices can be used as both the stirring bar of microextraction, and extractor of the analytes. This paper aims to present a new extraction alternative that provides a simple and easy method to extract analytes from the complex matrices. A magnetic solvent bar liquid-phase microextraction (MSB-LPME) was first applied for the extraction of OPPs in juice samples. The MSB-LPME devices cheaply manufactured and easily assembled. The stainless-steel wire was inserted into the hollow of the hollow fiber, and the stainless-steel wire not only used as the magnet stirrer, but also achieved magnetic separation which was isolated from the sample matrix easily with an external magnetic field. The operation for the sample treatment was simple and the problem of solvent bar floating up on the samples was solved. Several experimental conditions were studied and optimized. The performances of developed method were evaluated.

2. Experimental 2.1. Chemicals and reagents Eight OPPs including phorate, diazinon, tebupirimfos, tolclofosmethyl, pirimiphos-methyl, fenthion, fenamiphos, and sulprofos were purchased from National Institute of Metrology (Beijing China), and the purity of OPPs is P98%. Stock solutions for the OPPs were prepared in acetone at 1000 lg mL1 and stored at 4 °C. Working standard solutions were prepared daily by diluting the stock solution with acetone. Analytical reagent grade ethyl acetate, chloroform, hexane, 1-octanol and toluene were obtained from Beijing Chemical Factory (Beijing, China). Pure water was obtained with a Milli-Q water system (Millipore, Billerica, MA, USA). Q3/2 Accurel PP hydrophobic polypropylene hollow fiber membrane (600 lm inner diameter, 200 lm wall thickness and 0.2 lm pore size) was purchased from Membrana GmbH (Wuppertal, Germany). The stainless-steel wire (505 lm outer diameter) was just fit the hollow fiber membrane. 2.2. Instruments and apparatus The analytical solution was analyzed using GC–MS QP 2010 (Shimadzu, Kyoto, Japan). Chromatographic separation was conducted with a DB-5MS capillary column (30 m  0.25 mm I.D., film thickness of 0.25 lm, J & W Scientific, Folsom, CA, USA). Helium (purity P 99.999%) was used as carrier gas at a constant flow of 1.0 mL min1. The temperature program was set initially at 70 °C for 3 min; ramp to 170 °C at a rate of 15 °C min1, held for 9 min; and then ramp to 200 °C at a rate of 3 °C min1, held for 1 min; finally raised to 230 °C at a rate of 10 °C min1, held for 2 min. Injector temperature was maintained at 280 °C, and the injection volume was 1.0 lL in a 5:1 split ratio. Mass spectrometric parameters: electron impact ionization mode with an ionizing energy of 70 eV, injector temperature 280 °C interface temperatures 250 °C, ion source temperature 200 °C. The mass spectrometer was operated in the selected ion monitoring (SIM) mode for quantitative analysis and the characteristic ions are given in Table 1. Full-scan MS data were acquired in the range of m/z 50– 900 to obtain the fragmentation spectra of the analytes. 2.3. Sample preparation The fruit juice samples, including lemon juice, apple juice, peach juice and orange juice were purchased from local supermarket (Changchun, China). The fresh spiked samples containing OPPs were prepared by spiking the mixed working standard solutions into juice samples and shaking for 10 min. Except for the experiments mentioned in Section 3.2.3, which were performed with all four samples, all other experiments were performed with lemon juice sample. All experiments were performed in triplicate. 2.4. MSB-LPME procedure The magnetic solvent bar (MSB) was designed (Fig. 1(A)), and it contained the hollow fiber and stainless-steel wire. The hollow fiber and stainless-steel wire were both manually and carefully cut into segments of 1.2 cm length. These segments were ultrasonically cleaned in an ultrasonic bath to remove impurities, and dried in the air. In order to prepare the extraction unit, the stainless steel wire was inserted into the hollow of the hollow fiber. Then the resulting fiber piece was immersed in 1-octanol for 1 min in order to impregnate pores of the fiber wall. In order to remove the extra

199

L. Wu et al. / Food Chemistry 176 (2015) 197–204 Table 1 Qualitative and quantitative data of OPPs.a

a b *

Analytes

Retention time (min)

Main fragment ion (m/ z)

Regression equationb A = (a ± SDa) + (b ± SDb)c

Linear range (lg L1)

r

LOD (lg L1)

LOQ (lg L1)

Matrix effect (%)

Phorate Diazinon Tebupirimfos Tolclofos-methyl Pirimiphos-methyl Fenthion Fenamiphos Sulprofos

15.673 18.973 20.761 22.943 24.750 26.323 31.900 35.094

121*, 304*, 318*, 265*, 305*, 278*, 303*, 322*,

A = (100.53 ± 21.91)+(1188.74 ± 4.45)c A = (30.13 ± 8.78)+(421.84 ± 0.89)c A = (166.29 ± 43.13)+(830.26 ± 4.37)c A = (233.58 ± 62.75)+(4864.30 ± 12.73)c A = (101.03 ± 23.84)+(630.70 ± 2.42)c A = (219.57 ± 30.34)+(2286.08 ± 6.16)c A = (166.90 ± 21.43)+(1144.10 ± 4.35)c A = (156.31 ± 24.55)+(1236.62 ± 4.98)c

0.125–12.5 0.250–25.0 0.250–25.0 0.125–12.5 0.250–25.0 0.125–12.5 0.125–12.5 0.125–12.5

0.9994 0.9993 0.9992 0.9998 0.9996 0.9996 0.9987 0.9985

0.033 0.096 0.078 0.018 0.051 0.027 0.030 0.057

0.11 0.32 0.26 0.06 0.17 0.09 0.10 0.19

2.1 1.9 1.5 3.4 8.6 7.3 3.6 5.5

75, 97, 260 137, 152, 179 261, 234, 152 267, 250, 125 290,276, 125 125, 109, 169 154, 288, 260 140, 156, 113

The experimental results were obtained with spiked lemon juice samples. A, peak area of analyte; c, concentration of analyte in lg L1; a, intercept; b, slope; SDa and SDb, standard deviations of intercept and slope, respectively. The ion for quantitative analysis.

Fig. 1. Magnetic solvent bar (A) and extraction procedure of OPPs from juice sample (B).

1-octanol from the surface of the fiber, the fiber was rinsed with water. Each segment was used only once to decrease the memory effect. The extraction procedure of OPPs from juice sample was depicted in Fig. 1(B). Six MSBs were placed into a 20 mL screw top vial containing 10 mL of homogenized juice sample. In the optimized procedure, the vial was carefully closed and put into a water bath on the magnetic stirrer for 25 min at 600 rpm. The extraction was performed at 30 °C. After the extraction stage, with the help of an external magnet, the MSBs were separated rapidly from the sample solution. Then the bars were eluted with 400 lL of methanol in an ultrasonic bath for 3 min. The eluate was separated from the MSBs also by a magnet. The eluate obtained was evaporated to dryness under a nitrogen stream and the residue was dissolved in 100 lL of hexane. The resulting solution was filtrated through 0.22 lm PTFE filter membrane, and then directly analyzed by GC–MS.

2.5. Box–Behnken design Box–Behnken design (BBD) is a class of rotatable or nearly rotatable second-order design based on three-level incomplete factorial design (Leitão & Esteves da Silva, 2008; Martendal, Budziak, & Carasek, 2007). Experimental design, data analysis, and quadratic model building were conducted using the Design Expert software (Trial Version 7.1.3, Stat-Ease Inc., Minneapolis, MN, USA).

A three-factor, three-level Box–Behnken design (BBD) with extraction time (X1), number of MSB (X2) and extraction temperature (X3) as the independent variables was employed for the study. Experiments were randomized, which obtained the maximum effects of unexplained variability in the observed responses, due to extraneous factors. Accordingly, the low, middle and high levels of each variable were designated as 1, 0, and +1, respectively. The actual design experiment was shown in Table S-1 of the Supporting Material. In order to predict the optimal point, a second-order model was fitted to correlate relationship between independent variables and responses. For three significant factors, the equation is:

Y ¼ b0 þ b1 X 1 þ b2 X 2 þ b3 X 3 þ b12 X 1 X 2 þ b13 X 1 X 3 þ b23 X 2 X 3 þ b11 X 21 þ b22 X 22 þ b33 X 23 where Y is estimate response, b0 is model constant, X1, X2 and X3 are independent variables; b1, b2 and b3 are linear coefficients; b12, b13 are b23 are interaction coefficients between the three factors (X1, X2 and X3); b11, b22 and b33 are quadratic coefficients. 3. Results and discussion 3.1. Optimization of MSB-LPME conditions In the MSB-LPME procedure, some parameters that may affect the extraction efficiency including type of extraction solvent,

200

L. Wu et al. / Food Chemistry 176 (2015) 197–204

number of MSB, extraction temperature, extraction time, salt concentration, stirring speed, pH of the sample solution and desorption conditions were investigated to determine the most favorable conditions. All experiments were performed in triplicate and the concentration of OPPs in the spiked samples was 1 lg L1 for phorate, tolclofos-methyl, fenthion, fenamiphos, and sulprofos; 2 lg L1 for diazinon, tebupirimfos, and pirimiphos-methyl. The recoveries of the analytes were used to evaluate the extraction efficiency. 3.1.1. Effect of extraction solvent In MSB-LPME, appropriate selection of the organic solvent is of great importance to obtain efficient extraction. Ideally, the solvent must be compatible with the fiber so that it can easily immobilize within the pores of the fiber. This is important since the extraction occurs on the surface of the solvent immobilized in the pores. The solvent must be immiscible with water in order to reduce solvent loss during the extraction. Furthermore, the solvent should have low saturated vapor pressure and highly soluble for the target analyte. On the basis of these considerations, 1-octanol, chloroform, ethyl acetate, toluene and hexane were investigated in preliminary experiments. The experimental results are shown in Fig. S-1 of the Supporting Material. It can be seen that the highest extraction efficiency for all analytes was obtained with 1-octanol. Therefore, 1-octanol was selected as the extraction solvent. 3.1.2. Effect of stirring speed The stirring of the sample solution plays an important role in LPME, as it can enhance the mass transfer rate of analytes and reduce the time required for equilibrium partitioning. The experimental results in Fig. S-2 of the Supporting Material showed that the recoveries of most analytes increase significantly when the stirring speed increases from 200 to 600 rpm, and then decreased. The main reasons may be that too high a stirring rate results in some solvent loss and produces air bubbles that affect extraction efficiency and reproducibility. Therefore, an optimum stirring speed of 600 rpm was selected. 3.1.3. Effect of salt concentration The salting-out effect has been widely used to enhance the extraction efficiency of polar compounds in extraction and microextraction techniques (Schellin et al., 2004; Lord & Pawliszyn, 2000; Shen & Lee, 2002). The effect of salt addition on the extraction efficiency was examined with NaCl concentration in the range of 0–20% (m/v). As shown in Fig. S-3 of the Supporting Material, the extraction efficiency rose with the increase in salt concentration until a maximum at 10% of NaCl, and after that it decreased. It is possible because the proper concentrations of NaCl can decrease the solubility of analytes and enhance their partitioning onto the fiber and organic phase. But the excess of the ionic strength affect diffusion of the analytes into the organic phase because of the electrostatic interaction of salt ions and analytes in solution. Considering the overall responses of the target compounds, 10% NaCl was selected in further experiments.

3.1.4. Effect of pH The effect of the pH of the sample solution on OPPs extraction efficiency was also investigated in the range from pH 2 to 13. As expected, due to the neutrality of the analytes, no effect of sample pH was observed. Thus, pH of samples did not need to be adjusted. 3.1.5. Desorption conditions After the extraction was completed, 400 lL of organic solvents including n-hexane, acetone, methanol and ethyl acetate were used to elute the analytes from the MSBs. The results showed that the best desorption performance was obtained with methanol. Therefore, methanol was chosen to be the optimal solvent for extraction of OPPs. The effect of desorption time was also investigated. After extraction, the analyte-enriched MSB was ultrasonicated for 1–10 min. The result indicated that 3 min was sufficient for desorption completely. When the time was too long, the analytes would be lost. Thus, 3 min was chosen as the appropriate desorption time. 3.1.6. Effect of number of MSB, extraction time and extraction temperature Three significant variables that could be potentially affected on the OPP recoveries, such as extraction time, number of MSB and extraction temperature were chosen with BBD. A 17-run BBD was used to optimize the parameters and to study the possible interaction between the parameters. The experimental design is shown in Table S-1 of the Supporting Material. The summary of the analysis of variance (ANOVA) is shown in Table 2. The ANOVA of regression model demonstrates that the model is highly significant. The goodness of the model can be checked by the determination coefficient (R2). The value of R2 (0.9756–0.9853) indicates good relation between the experimental and predicted values of the response. The lack-of-fit measures the failure of the model to represent data in the experimental domain at points which are not included in the regression (Khajeh, 2011). The P value and lack of fit value of the model is 0.0805 (not significant), respectively. The non-significant value of lack-of-fit (>0.05) revealed that the quadratic model is statistically significant for the response. These two values confirm that the model fitness is good. Three response surfaces obtained based on the results in the BBD are illustrated in Fig. 2. Phorate was chosen as a model analyte for the OPPs. The effect of different parameters on the phorate recovery was shown in Fig. 2. Each figure shows the 3D response surface based on two variables and quality of the extraction method at center level of the third variable. From Fig. 2, it can be concluded that the number of MSB has the most significant effect on the OPP recoveries. With increase of the number of MSB, the response first increases and then decreases. The reason may be that too many bars cannot make desorption complete. Extraction time is another important factor to be considered to obtain efficient extraction. MSB-LPME is an equilibrium extraction, rather than an exhaustive extraction. A prolonged

Table 2 Parameters for the BBD.

Phorate Diazinon Tebupirimfos Chlorphrifos-methyl Pirimiphos-methyl Fenthion Fenamiphos Sulprofos

b0

b1

b2

b3

b12

b13

b23

b11

b22

b33

P-value of the mode

Lack of fit value

R2

89.58 82.55 86.90 102.08 93.18 97.68 85.58 86.02

4.04 4.88 6.24 4.25 3.78 4.06 4.96 4.51

20.03 19.99 20.03 20.21 20.58 19.51 19.46 19.73

2.71 3.21 2.76 2.86 2.90 3.13 2.18 2.84

1.85 2.82 1.75 1.87 2.17 2.12 1.77 1.67

0.67 0.23 0.53 0.42 0.37 0.85 1.75 0.45

0.0500 0.15 0.45 0.25 0.68 1.05 0.32 0.28

13.13 13.17 14.96 13.07 13.43 13.70 12.95 13.46

29.90 27.45 27.09 19.59 29.48 30.25 29.62 28.34

12.38 12.70 13.96 12.09 11.93 12.03 12.14 12.46