Fundamental study of a novel membrane filtration cleanup method for pesticide analysis in agricultural products

Food Control 64 (2016) 1e9

Contents lists available at ScienceDirect

Food Control journal homepage: www.elsevier.com/locate/foodcont

Fundamental study of a novel membrane filtration cleanup method for pesticide analysis in agricultural products Jangho Hong, Ayato Kawashima*, Minami Okamoto, Kana Kanetsuki, Takanori Makino, Noriaki Hamada Department of Environmental Science for Industry, Faculty of Agriculture, Ehime University, 3-5-7 Tarumi, Matsuyama, Ehime, 790-8566, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 October 2015 Received in revised form 30 November 2015 Accepted 9 December 2015 Available online 12 December 2015

The recovery and purification characteristics of membrane filtration for pesticide analysis of agricultural products were investigated. Eight different types of membranes classified by their molecular weight cut off (MWCO) and material were used. The results showed that the recovery and purification characteristics varied according to the eluting solvent used, as well as the membrane's MWCO and material. The recovery increased as the MWCO increased, and the purification increased when the eluting solvent contained water. A GK membrane with acetonitrile-water (1:1, v/v) was the most effective membrane filtration method among those tested. The pesticide analysis of spinach using the GK filtration method indicated that this method results in better purification than the modified QuEChERS method. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Membrane filtration method Pesticide analysis Purification characteristics Interference materials Spinach

1. Introduction Agrichemicals are widely used in agriculture to prevent the destruction of food crops by pests or unwanted plants and improve €zde, Dilek, Fatih, & Semih, plant quality (Bakirci & Hisil, 2011; Go 2014; Słowik-Borowiec, Szpyrka, & Walorczyk, 2015). Despite the wide range of benefits of using pesticides in agriculture, the incorrect application of these chemicals, such as the use of an inappropriate pesticide type on foodstuffs or their unwarranted use, can result in high and undesirable levels of these compounds in cs, Deyl, Miksik, the products that reach consumers (Cserh ati, Forga & Eckhardt, 2004). Therefore, analyzing pesticides in agricultural products is necessary. However, more than 800 pesticides belonging to more than 100 substance classes have been registered and used globally for decades, and the chemical and physical properties of these may vary considerably. Although a multi-class, multi-residue analytical method would be the most useful for regulatory pesticide monitoring, the diverse properties of pesticides complicate the development of such a universal method (Siweon, Sooyeon, Jinyoung, MeeKyung, & Jeonghan, 2015; Urairat, Steven, & Natchanun, 2011). The most used approach for pesticide

* Corresponding author. E-mail address: [email protected] (A. Kawashima). http://dx.doi.org/10.1016/j.foodcont.2015.12.003 0956-7135/© 2015 Elsevier Ltd. All rights reserved.

extraction from food samples is currently QuEChERS (Zeying et al., 2015). QuEChERS, which stands for quick, easy, cheap, effective, rugged, and safe, provides satisfactory results for a wide range of pesticides. Since it was first introduced, the QuEChERS method has been widely accepted by the scientific community (Anastassiades,  Mastovska, & Lehotay, 2003; Anastassiades, Lehotay, Stajnbaher, & Schenck, 2003; Daniela, Gian, Paola, Stefania, & Maria, 2012). This method and its modified variants are used primarily to analyze pesticide residues in fruits and vegetables (Raphaell et al., 2013; Słowik-Borowiec et al., 2015; Tibor, Esther, Zdenek, Ivan, & Adam, 2004). However, many agricultural product samples are known to be rich in pigments and fatty acids, and as a result, interference materials can be retained in the final extract, despite using a cleanup method. Matrix-induced signal enhancement can greatly affect analyte responses and performance by causing analyte retention and degradation. Furthermore, interference materials derived from agricultural products can become a burden to GC columns and detectors (Hyeyoung, Steven, & Lucía, 2012). To overcome this issue, we focused on developing a cleanup procedure specific to pesticide analysis. Membrane filtration is one of the efficient methods for removing interference materials (Juan, Javier, Francisco, & Carolina, 2009). The membrane's ability to control the permeation rate of a chemical species is key for the efficacy of this technique, and membranes also have additional advantages, such as flexibility, which can be combined with other separation

2

J. Hong et al. / Food Control 64 (2016) 1e9

processes. For these reasons, membrane filtration is widely used in separation, purification, and concentration processes (Kai Yu and Tai-Shung, 2005; Kyu-hong, Kyung-guen, Ho-young, & Ick-tae, 1999; Robert, 1995, chap. 13). In particular, it has been developed mainly for gas separation, wastewater reclamation, and drinking water processing to remove pollutants and natural organic materials (Daeyoun & Sangyong, 2011; Patrick, Tulay, & Philip, 2013; Plakas, Karabelas, Wintgens, & Melin, 2006; Yeomin, Paul, Shane, & Eric, 2006). Although some studies investigating analysis of pesticides by membranes have been reported (Eriko et al., 2006; Juan et al., 2009; Konstantinos, Anastasios, Thomas, & Thomas, 2006; Lijun, Yelena, & Steven, 2014). However, there are few reports on the removal of interference materials by membrane filtration during pesticide analysis using GC/MS. In this study, we investigated the recovery and purification achievable by membrane filtration for pesticide analysis. We used 8 different types of membranes that were classified in terms of their MWCO and material. Additionally, spinach was used as the agricultural product because it is a representative leafy vegetable with a highly pigmented matrix. 2. Materials and methods 2.1. Chemicals and reagents All pesticide standards were high purity. Pesticide standard solution 31 (for GC analysis, containing 85 types of pesticides) was obtained from Kanto chemical (Tokyo, Japan) and used as the standard solution. It contains various types of pesticides such as hydrophilic, hydrophobic, low MW, high MW, carbamate, organophosphate, organochloride and so on. An internal standard solution (containing phenanthrene-d10, anthracene-d10, 9bromoanthracene) was obtained from Wako Pure chemical Industries (Tokyo, Japan) and used as a syringe spike which was added to the sample at the same concentration as used in the calibration for verifying validity of sample. Stock standard solutions of 200 ng mL
1 were prepared in acetone and stored in the dark at 4  C. All organic solvents were pesticide analysis grade and were obtained from Wako (Tokyo, Japan). InertSep PSA (500 mg/6 mL) and InertSep GC/NH2 (500 mg/500 mg/6 mL) cartridges were obtained from GL Sciences (Tokyo, Japan). Trisodium citrate dihydrate, disodium hydrogen citrate sesquihydrate, sodium chloride, and magnesium sulfate were obtained from Kanto Chemical (Tokyo, Japan).

were identified and quantified with GC (TRACE GC Ultra) coupled with MS (Polaris Q) obtained from Thermo Fisher Scientific (Waltham, MA, USA). Aliquots (2 mL) of the final extracts were injected into the GC system. The oven temperature program started at 50  C (hold for 1 min) and then increased at a rate of 30  C min
1 to 125  C and at a rate of 5  C min
1 to 200  C. Finally, the temperature was increased at a rate of 10  C min
1 to 300  C (hold for 11.5 min). Calibration standards in acetone were prepared at 50 ng mL
1, 100 ng mL
1, 200 ng mL
1, and 400 ng mL
1 for GC/MS measurement. Mass spectrometric detection was performed in electron ionization mode (EI, 70 eV). A DB-5MS capillary column (30 m  0.25 mm i.d., 0.25 mm) was obtained from Agilent Technologies (Santa Clara, CA, USA) and used. An untreated, non-polar fused silica capillary column (1.5 m  0.25 mm) from SigmaeAldrich (Saint Louis, MO, USA) was used as the guard column. 2.4. Method performance We used the calibration curve for determining the concentration of pesticides. The calibration curve was constructed for each pesticide using four different concentrations of the pesticide standard solution in acetone. The method linearity was in the range of 50e400 ng g
1. The resulting coefficients of regression (R2 value) exceeded 0.99 in all cases. The LOD was calculated three times of standard deviation, whereas the LOQ was equal to ten times of standard deviation. The LODs obtained for most of the pesticides were in the range of 1e10 ng g
1. 2.5. Determination of the purification by visual extract observation

Polyamide membranes (NFG with MWCO of 600e800 Da, NFW with MWCO of 300e500 Da, and NFX with MWCO of 150e300 Da) were obtained from Synder Filtration (CA, USA). A polyethersulfone membrane (NP030 with MWCO of 400e600 Da) was obtained from Daicen Membrane Systems (Tokyo, Japan). A polypiperazine amide membrane (XN45 with MWCO of 500 Da) was obtained from TriSep (Goleta, CA, USA). A composite polyethersulfone membrane (MPF36 with MWCO of 1000 Da) was obtained from Koch Membrane Systems (Wilmington, MA, USA), and thin-film composite membranes (GK with MWCO of 3000 Da and GH with MWCO of 2000 Da) were obtained from Lenntech (Rotterdam, Netherlands).

In this study, pesticide-free spinach obtained from a local market (Matsuyama, Ehime, Japan) was used. Sample preparation was based on a QuEChERS method (Steven et al., 2010). A chopped sample (10 g) was added to a 50-mL polypropylene centrifuge tube. Then, 10 mL of acetonitrile (MeCN) was added, the tubes were vigorously shaken for 1 min by hand. Subsequently, 1 g of trisodium citrate dihydrate, 0.5 g of disodium hydrogen citrate sesquihydrate, 1 g of sodium chloride, and 4 g of magnesium sulfate were added and shaken vigorously by hand for 1 min. The mixture was centrifuged at 3500 rpm for 10 min to separate the organic phase (MeCN) from the aqueous and solid phases. After 5 mL of supernatants were transferred to tubes, three types of mixed solvent were added each tubes respectively to yield a total volume of 12.5 mL to generate three types of the analyte solutions. In case of the methanol (MeOH) analyte solution, the MeCN was evaporated, and then, 12.5 mL of MeOH was added, in case of the MeCN analyte solution, 7.5 mL of MeCN was added and in case of the MeCN-water (1:1, v/v) analyte solution, 1.25 mL of MeCN and 6.25 mL of water were added. Before membrane filtration, 20 mL of mixed solvent (MeOH, MeCN, or MeCN-water (1:1, v/v)) was added to the membrane which was placed in the membrane-filtering apparatus for pre-washing (0.5 MPa, 30  C, and 400 rpm). Then, 10 mL of the analyte solution was added to the membrane-filtering apparatus for filtration. Finally, 5 mL of mixed solvent was added to rinse the membrane, and the total filtered solution reached a volume of 15 mL. At this point, the purification was determined visually according to the color of the extract.

2.3. Instruments and GC/MS analytical conditions

2.6. Determination of the recovery using spiked standard solutions

We used HP4750 Stirred Cell obtained from Sterlitech (Kent, WA, USA) as the membrane-filtering apparatus and a SepPak elution pump obtained from Waters (Millford, MA, USA) for the cleanup process. Model 5922 instrument was obtained from Kubota (Osaka, Japan) and used for centrifugation. The pesticides

Three types of 12.5 mL crude standard samples (MeOH, MeCN, and MeCN-water (1:1, v/v)) including 125 mL of pesticide standard solution were used as the pesticide-contaminated samples, and three types of 12.5 mL crude reference samples (MeOH, MeCN, and MeCN-water (1:1, v/v)) including 125 mL of acetone without

2.2. Membranes

J. Hong et al. / Food Control 64 (2016) 1e9

pesticide were used for calculating the recovery rates and matrix effects. For membrane filtration, 10 mL aliquot was examined as the analyte solution. After membrane filtration (refer to section 2.5) and dehydration by Na2SO4, the analytes were concentrated to approximately 50 mL using an evaporator (100 hPa and 40  C), followed by cleanup with a PSA cartridge. First, 10 mL of mixed solvent (MeCN-toluene (3:1, v/v)) was added into the PSA cartridge for pre-washing, and concentrated analyte was added, followed by 2 mL for of MeCN for dispersing of the analyte into PSA cartridge. Finally, 20 mL of mixed solvent (MeCN-toluene (3:1, v/v)) was eluted by an elution pump (0.033 mL s
1). The analytes were concentrated by a evaporator (100 hPa and 40  C). A recovery sample (from the crude standard sample) and a reference sample (from the crude reference sample) were adjusted by adding acetone to a total volume of 1 mL. Then, the reference sample was divided in half to produce a matrix sample (to determine the matrix effects) and a blank sample (to determine the recoveries). To determine recoveries, 50 mL of the syringe spike solution was spiked into the recovery sample to a final volume of 1 mL, and 25 mL of the syringe spike solution was spiked into the blank sample to a final volume of 500 mL for checking interference by spinach. For matrix effects determination, 25 mL of the syringe spike solution and 50 mL of the standard solution were spiked into the matrix sample to a final volume of 500 mL. Finally, GC/MS measurement was performed for to identify and quantify the pesticides. Fig. 1 shows the schematic diagram of analysis procedure. 2.7. Pesticide analysis in spinach by the membrane filtration and modified QuEChERS methods First of all, we prepared two 10 g of spinach samples which were from same spinach. One was used as crude standard sample which is pesticide-contaminated sample and the other was used as crude reference sample which is pesticide free sample for calculation of the recovery rates and matrix effects. QuEChERS extraction was the same as section 2.5, but, in this time 125 mL of pesticide standard solution was mixed with 10 g of spinach. 5 mL of supernatants were transferred to each of two tubes to generate a recovery sample and a reference sample. Then, the analyte solution was made by adding

Standard sample

3

mixed solvent (1.25 mL of MeCN and 6.25 mL of water) to a final volume of 12.5 mL. After membrane filtration (refer to section 2.5) and dehydration by Na2SO4, the analytes were concentrated to approximately 50 mL by an evaporator (100 hPa and 40  C), followed by cleanup with a GC/NH2 cartridge. The following method was the same as section 2.6. Briefly, after cleanup procedure, the analytes were measured by GC/MS. 3. Results and discussion 3.1. Preliminary study for finding out best parameters of membrane filtration We examined the purification achieved by membrane filtration by visually observing the spinach extracts. The pesticide recovery was determined by calculating the recovery rate (%) of the spiked a pesticide standard. 3.1.1. Effects of nanofiltration membranes material and eluting solvent on the purification and recovery Three different types of nanofiltration (NF) membranes (NFW, a polyamide membrane; NP030, a polyethersulfone membrane; and XN45, a polypiperazine amide membrane) with similar MWCO values (from 300 to 600 Da) but different materials were used for this experiment because the average molecular weight of pesticides is between 300 and 400 Da. MeCN, MeOH, and MeCN-water (1:1, v/ v) were used as the eluting solvent. The achieved purification was evaluated by visually observing the color of the extracts. Fig. 2 shows the colors of extracts achieved using the 3 different types of nanofiltration (NF) membranes with the 3 different solvents. In MeOH and MeCN-water (1:1, v/v), the extract colors obtained by NFW and XN45 were lighter than those obtained by NP030, while no substantial difference was observed for MeCN. In all membranes, the colors of the extracts with MeCN-water (1:1, v/v) were lighter than those with MeOH and MeCN. In other words, low purification was achieved using MeCN and MeOH compared with MeCN-water (1:1, v/v). The main reason underlying this difference is that MeCN and MeOH, which have relatively higher solubility could lead to membrane swelling (Patrick & Philip, 2013). In

Reference sample

(spinach 10 g)

(spinach 10 g) 125 μL of acetone

125 μL of standard solution

Extraction (QuEChERS)

Extraction (QuEChERS)

Membrane Filtration

Membrane Filtration

Cleanup

Cleanup

50 μL of syringe spike

25 μL of syringe spike

25 μL of syringe spike

50 μL of standard solution

Recovery sample

Blank sample

Matrix sample

(1 mL)

(500 μL)

(500 μL)

Fig. 1. Schematic diagram of analysis procedure.

4

J. Hong et al. / Food Control 64 (2016) 1e9

Fig. 4. Colors of extracts after NF membrane filtration with different membrane MWCO values with MeCN-water (1:1, v/v). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2. Colors of extracts after NF membrane filtration with three different solvents. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

addition, dark-colored extracts became lighter when an eluting solvent containing water was used. Thus, the purification achieved by membrane filtration was increased by using a water-containing solvent (MeCN-water (1:1, v/v)). This is because of the hydrophobic effect, which is the tendency of nonpolar substances to aggregate in aqueous solution and exclude water molecules. NFW and XN45 resulted in particularly high purification because these membranes include amide bonds that could undergo shrinking in the presence

of water. Thus, NFW and XN45 might have experienced shrinkage when the water-containing solvent was used (Guo-Rong, Jiao-Na, & Cong-Ju, 2013; Juan et al., 2009). Based on these results, we can see that the purification can be adjusted by varying the solvent and membrane material used. The recoveries of pesticides are presented in Fig. 3, which shows that the pesticide recovery was higher in MeCN than in the other solvents tested for all membranes. In MeOH, the recovery was slightly lower for XN45 compared with the other membranes studied. In MeCN-water (1:1, v/v), the recovery was low for all membranes and was lowest for XN45. A watersolvent mixture, such as MeCN-water (1:1, v/v), could lead to shrinkage of amide-containing membranes and, possibly, low recovery. Specifically, XN45 is a polypiperazine amide membrane, and as a result, it could readily undergo shrinkage in a watercontaining solvent, leading to fouling effects (Fa, Da, & Shou, 2007; Guo-Rong et al., 2013; Juan et al., 2009). 3.1.2. Effect of nanofiltration membranes MWCO on purification and recovery Here, we examined the effect of MWCO on purification and recovery. Three different types of polyamide NF membranes (NFG

Fig. 3. Recovery of pesticides according to molecular weight by three membranes filtration with three different solvents.

J. Hong et al. / Food Control 64 (2016) 1e9

5

Fig. 5. Recovery of pesticides according to molecular weight by NF membrane filtration with different membrane MWCO values with MeCN-water (1:1, v/v).

with MWCO of 600e800 Da, NFW with MWCO of 300e500 Da, and NFX with MWCO of 150e300 Da) which differed in terms of their MWCO were investigated using MeCN-water (1:1, v/v) as the eluting solvent. Fig. 4 shows that dark-colored extracts became lighter as the membrane MWCO decreased. Thus, the purification improved as the membrane MWCO decreased and was remarkably good when the MWCO was below 500 Da. One possible reason for this behavior is MWCO of the membrane. Interference materials, such as chlorophyll, xanthophyll, pheophytin, and carotene with molecular weight exceeding 500 Da were removed by membranes. Another possible reason is the hydrophobic effect that occurs in water-containing solvent. In contrast, as shown in Fig. 5, the recovery decreased as the membrane MWCO decreased. Specifically, the recovery decreased remarkably when the MWCO was below 300 Da, which is a fairly low value because many pesticides have molecular weights over 300 Da and could not have passed through the membrane. However, some pesticides with molecular weights below 300 Da were not recovered, possibly because the membranes had experienced shrinkage and resisted swelling in the presence of the water-containing solvent, which may have caused the fouling effects. 3.1.3. Recovery and purification by ultrafiltration membranes The above results indicated that the purification was improved by using a water-containing solvent and that the pesticide recovery increased as the MWCO increased. Because the MWCO of ultrafiltration (UF) membranes is higher than that of nanofiltration (NF) membranes, the purification and recovery of UF membranes classified according to their MWCO values (from 3000 to 1000 Da) were tested. In this experiment, GK (MWCO of 3000 Da), GH (MWCO of 2000 Da), and MPF-36 (MWCO of 1000 Da) were used with MeCNwater (1:1, v/v) as the eluting solvent. Visual determination of the color of extracts revealed that the purification was not significantly different between UF membranes, as shown in Fig. 6. In other words, almost the same purification can achieved using any membrane with a MWCO exceeding 1000 Da. Fig. 7 shows that

Fig. 6. Purification by UF membrane filtration with MeCN-water (1:1, v/v).

correlation between retention time and molecular weight of pesticides. Fig. 8 shows the recovery obtained using the ultrafiltration (UF) membranes. The percentage of recovered pesticides (ranging from 70% to 120%) using the GK filtration method was 73%, which was higher than those obtained via GH filtration (54%) and MPF-36 filtration (62%). The main underlying reason is that the GK membrane has a comparatively high MWCO. Fig. 9 shows that the recovery of pesticides was related to their water solubility; however, the recoveries obtained by GK filtration did not differ significantly relative to water solubility compared with other membranes. In contrast, pesticides that are highly water soluble, or hydrophilic pesticides, were highly recovered compared with hydrophobic pesticides by GH and MPF-36 membranes, which have relatively low MWCO values. The reason underlying this result might have been that hydrophilic pesticides can pass through these membranes easier than hydrophobic pesticides in water-containing solvents because of surface interactions and the solubility parameter during mass transfer (Patrick et al., 2013; Robert, 1995). These results indicated that the recovery attainable using GK was higher than that of other UF membranes, although the purification was almost equivalent. Furthermore, GK membrane filtration was completed with relatively fast volumetric flow rate (0.94 mL min
1) compared with GH membrane filtration (0.83 mL min
1) and MPF36 filtration (0.33 mL min
1) Therefore, we chose the GK membrane for the pesticide analysis via membrane filtration method. 3.2. Comparison study with best parameters According to the results above, the GK filtration method, which consisted of QuEChERS extraction, GK filtration, GC/NH2 cleanup, and GC/MS measurement was the most effective method tested. Thus, we used this method to analyze pesticides in an agricultural product. Pesticide-free spinach was used as a representative agricultural product because it is a typical leafy vegetable with a highly pigmented matrix. The sample was generated by adding the pesticide standard solution into the pesticide-free spinach. After the analyte was extracted and cleaned up, it was measured by GC/ MS. This experiment was designed to determine the efficacy of membrane filtration compared with a modified QuEChERS method which consisted of QuEChERS extraction, GC/NH2 cleanup, and GC/

Fig. 7. Correlation between retention time and molecular weight of pesticides.

6

J. Hong et al. / Food Control 64 (2016) 1e9

Fig. 8. Recovery of pesticides according to molecular weight by UF membrane filtration with MeCN-water (1:1, v/v).

MS measurement, which is one of commonly used pesticide analysis methods. 3.2.1. Recovery The recovery (%) is defined by following equation.

Interference materials are known to be problematic in pesticide analysis using GC/MS. Matrix-induced signal enhancement occurs in GC when active surfaces in the system cause analyte retention and/or degradation (Erney, Gillespie, & Gilvydis, 1993). Table 1 shows the matrix effects (%) and RSD. In GK filtration, the matrix effects were 79.4e207.8%, with RSD values (n ¼ 3) below 20%, and

Recovery ð%Þ ¼ ðpeak area of analytes in recovery sample=peak area of analytes in matrix sampleÞ  100

The recoveries of 89 pesticides (from a pesticide standard solution) by the GK filtration and modified QuEChERS methods are shown in Table 1. The recoveries of most pesticides by GK filtration were 70.5e95.2%, although 15 pesticides were recovered below 70% and the relative standard deviation (RSD) values (n ¼ 3) were below 20%, except for tetradifon. Meanwhile, the overall recoveries obtained by the modified QuEChERS method were 85.5e114.7%, and the RSD values (n ¼ 3) were below 20%, except for tecnazene. In some cases, the GK filtration method was unable to meet the regulatory requirements for pesticide analysis because of retention of the pesticides on the membrane (Hana, Sapozhnikova, & Lehotay, 2014). The possible reason may have been the hydrophobic effect or the influence of the solubility parameter. However, more than 80% of the 89 pesticides tested were acceptably recovered by GK filtration according to an acceptable range of recoveries from 70 to 120%. 3.2.2. Matrix effects To evaluation matrix effect, a standard solution of 200 mg kg
1 were prepared. The matrix effects (%) are defined by following equation.

(1)

those of the modified QuEChERS method were 75.0e231.1%, with RSD values (n ¼ 3) below 20%, except for quintozene. The percentage of pesticides suffering from matrix effects for tested 89 pesticides in the GK filtration method (26%) was lower than in the modified QuEChERS method (29%). 3.2.3. Chromatographic interference materials Another way to compare the effects of interference materials (matrix co-extracts) between the two methods is by examining the chromatograms. Fig. 10 shows the GC/MS total ion chromatograms for blank spinach samples using the modified QuEChERS and GK filtration methods. Some prominent peaks from the spinach sample appear in both chromatograms. However, the peaks representing fatty acid derivatives obtained by the GK filtration method were fewer and less intense than those from the modified QuEChERS method. Furthermore, phytol (RT:19.55), tocopherol (RT:28.60) and phylloquinone (RT:30.03) were removed by GK filtration. This might be the result of the hydrophobic effect; surface interactions, such as van der Waal's forces during mass transfer; and the solubility parameter (Shiao-Shing, James, Luke, & Charles, 2004). Although the recovery achieved by membrane filtration was slightly lower than that of the modified QuEChERS method, the

Matrix effects ð%Þ ¼ ðpeak area of analytes in matrix sample=peak area of analytes in standard solutionÞ  100

Fig. 9. Recovery of pesticides according to water solubility by UF membrane filtration with MeCN-water (1:1, v/v).

(2)

J. Hong et al. / Food Control 64 (2016) 1e9

7

Table 1 Recoveries and matrix effects of pesticides from spinach samples using the GK filtration and modified QuEChERS methods (n ¼ 3). Pesticides

XMC Tecnazene Propachlor Propoxur Benfluralin Monocrotophos Dicloran Carbofuran Quintozene Clomazone Simazine Atrazine Propyzamide Cyanophos Triallate Isazofos Iprobenfos Benoxacor Bromobutide Acetochlor Phosphamidon Chlorpyrifos-methyl Vinclozolin Metalaxyl Ametryn Prometryn Propanil Fenpropimorph Ethofumesate Chlorthal-dimethyl Triadimefon Quinoclamine Nitrothal-isopropyl Bromacil Bromophos-methyl Diphenamid Fthalide Allethrin-1 Allethrin-2 Allethrin-3 Allethrin-4 Dimethametryn Dimepiperate Methidathion Tetrachlorvinphos Alpha-Endosulfan(I) Fenothiocarb Napropamide Fenamiphos Flutriafol Profenofos DEF Oxadiazon E-Metominostrobin Isoprothiolane Isoxathion_oxon Buprofezin Flamprop-methyl Bupirimate Oxyfluofen Z-Metominostrobin Azaconazole Isoxathion Imazamethabenz-methylester Ethion Beta-Endosulfan(II) Fluacrypyrim Oxadixyl Benalaxyl Carfentrazone-ethyl Trifloxystrobin Quinoxyfen

M.W

179 261 212 209 355 223 207 221 295 240 202 216 256 243 305 314 288 260 312 270 300 323 286 279 227 241 218 303 286 332 294 208 295 261 366 239 272 302 302 302 302 255 263 302 366 407 253 271 303 301 374 315 345 284 290 257 305 336 316 362 284 300 313 288 385 407 426 278 325 412 408 308

GK filtration method

Modified QuEChERS methods

Recovery (%)

RSD (%)

Matrix effect (%)

RSD (%)

Recovery (%)

RSD (%)

Matrix effect (%)

RSD (%)

94 47 86 87 65 95 86 90 35 87 92 89 86 91 65 87 87 84 86 84 90 71 79 87 80 78 91 72 87 76 76 89 76 91 58 88 64 N.D N.D 83 79 76 80 86 88 69 89 90 88 92 81 73 77 93 88 89 79 89 89 62 92 89 81 91 72 89 90 93 91 89 85 52

5 13 3 3 5 4 7 6 16 2 5 7 5 5 5 4 8 4 7 5 6 6 4 5 8 5 4 1 4 5 1 1 8 4 6 5 5 N.D N.D 12 19 6 10 10 9 19 6 7 8 8 6 14 9 7 5 6 3 5 7 8 7 7 7 5 9 7 8 6 7 11 10 11

207 101 125 137 104 175 170 125 88 132 134 114 121 137 107 107 114 107 98 101 120 93 93 116 121 104 153 99 97 89 104 160 107 116 95 104 110 N.D N.D 92 102 108 94 98 112 81 156 111 102 118 104 103 89 100 97 99 93 96 102 93 101 100 87 142 97 79 103 103 114 105 100 91

5 11 4 7 5 3 5 3 19 4 7 6 3 6 2 1 2 1 1 1 1 2 2 3 3 3 3 3 2 2 3 2 1 4 1 3 2 N.D N.D 11 5 1 4 4 6 8 4 4 5 3 4 7 4 4 4 4 4 3 3 3 4 6 4 2 4 7 4 5 2 5 6 8

94 97 94 94 94 91 94 98 89 98 97 95 94 95 94 94 93 99 101 102 96 99 100 98 97 97 98 88 103 100 90 97 97 102 98 97 94 N.D N.D 97 104 97 94 96 100 91 95 96 91 95 94 86 98 95 97 98 96 96 98 90 98 97 97 90 93 98 99 95 97 99 96 95

4 7 5 5 6 7 6 10 2 6 2 6 7 7 5 8 10 5 10 5 12 7 6 7 7 7 5 7 6 5 7 9 9 6 6 6 5 N.D N.D 13 21 9 11 7 6 9 4 8 12 6 6 12 3 7 6 6 6 5 6 9 8 8 6 12 10 5 11 7 6 4 11 6

219 108 139 143 111 194 185 120 107 136 142 122 122 155 106 109 115 102 96 92 119 92 92 113 122 103 148 103 97 84 104 159 107 116 93 100 106 N.D N.D 103 100 109 93 101 123 83 162 114 113 114 111 104 83 103 96 93 85 90 91 86 100 91 81 125 101 75 111 99 111 101 100 89

5 8 9 8 12 7 4 12 9 13 8 11 11 13 15 14 12 13 15 10 14 15 12 11 14 13 7 11 9 13 12 8 12 14 14 14 16 N.D N.D 10 17 12 6 7 3 10 8 10 0 4 8 6 7 4 6 11 8 9 8 4 6 5 8 39 7 6 7 1 6 7 6 3 (continued on next page)

8

J. Hong et al. / Food Control 64 (2016) 1e9

Table 1 (continued ) Pesticides

Propargite Norflurazon Diclofop_methyl Hexazinone Pyridaphenthion Bromopropylate Piperophos Methoxychlor Phosmet Phenothrin-1 Phenothrin-2 Tetradifon Pyrazophos Fenbuconazole Flumioxazin Flumiclorac-pentyl Tolfenpyrad

M.W

351 304 341 252 340 428 354 346 317 350 350 356 373 337 354 424 384

GK filtration method

Modified QuEChERS methods

Recovery (%)

RSD (%)

Matrix effect (%)

RSD (%)

Recovery (%)

RSD (%)

Matrix effect (%)

RSD (%)

69 93 68 94 78 78 92 68 87 79 78 63 87 91 91 70 68

10 6 5 8 7 8 9 11 8 9 15 27 13 5 15 11 14

102 166 112 134 104 107 119 92 133 111 113 85 153 208 148 170 193

2 3 1 5 3 3 4 1 3 3 5 15 10 7 8 7 7

89 97 101 103 99 94 99 99 96 96 115 94 99 94 105 98 86

4 4 6 13 10 6 20 8 7 11 20 14 10 8 20 17 16

111 160 119 160 107 108 137 87 124 128 102 91 162 207 167 193 231

3 0 8 4 6 6 11 1 3 6 8 11 2 7 19 16 17

Fig. 10. GC/MS chromatograms of blank spinach samples obtained with the modified QuEChERS and GK filtration methods.

examination of the matrix effects and chromatograms revealed that the GK filtration method is more effective in the removal of interference materials than the modified QuEChERS method. 4. Conclusions In this study, we investigated a novel membrane filtration cleanup method for pesticide analysis. The results showed that the recovery and purification varied according to the eluting solvent and the membrane MWCO and material of the membrane. These are therefore important parameters that determine the membrane performance. The recovery of pesticides increased as the membrane MWCO increased, and the purification improved when an eluting solvent containing water was used. The main underlying reason might have been the hydrophobic effect, which is the

tendency of nonpolar substances to aggregate in aqueous solution and exclude water molecules. Another possible reason may have been surface interactions, such as van der Waal's forces during mass transfer, and the solubility parameter. The results revealed that the GK membrane was the most effective membrane tested. Most of the interference materials were removed or decreased by GK filtration method while these interference materials were retained by the modified QuEChERS method. Although the use of membrane filtration for pesticide analysis still needs to be improved, this technique has potential to be an effective cleanup method. Furthermore, this study provides fundamental data regarding the use of membrane filtration in different types of pesticide analyses. In the future, various membrane filtration parameters, such as flux, temperature, membrane MWCO and material, and eluting solvent will be experimentally investigated.

J. Hong et al. / Food Control 64 (2016) 1e9

References  Anastassiades, M., Lehotay, S. J., Stajnbaher, D., & Schenck, F. J. (2003). Fast and easy multiresidue method employing acetonitrile extraction/partitioning and “dispersive solid-phase extraction” for the determination of pesticide residues in produce. Journal of AOAC International, 86, 412e431. Anastassiades, M., Mastovska, K., & Lehotay, S. J. (2003). Evaluation of analyte protectants to improve gas chromatographic analysis of pesticides. Journal of Chromatography A, 1015, 163e184. Bakirci, G. T., & Hisil, Y. (2011). Fast and simple extraction of pesticide residues in selected fruits and vegetables using tetrafluoroethane and toluene followed by ultrahigh-performance liquid chromatography/tandem mass spectrometry. Food Chemistry, 135, 1901e1913. ti, T., Forga cs, E., Deyl, Z., Miksik, I., & Eckhardt, E. (2004). Chromatographic Cserha determination of herbicide residues in various matrices. Biomedical Chromatography, 18, 350e359. Daeyoun, O., & Sangyong, N. (2011). Developmental trend of polyimide membranes for gas separation. Membrane Journal, 21, 307e320. Daniela, M., Gian, L. F., Paola, B., Stefania, S., & Maria, C. A. (2012). Detection of pesticides in crops: a modified QuEChERS approach. Food Control, 25, 270e273. Eriko, H., Hiroko, K., Takashi, S., Akira, S., Satoru, T., & Takashi, K. (2006). Simultaneous determination of pesticides in agricultural products by LC/MS/MS using clean-up with ultrafiltration. Food Hygiene and Safety Science, 47, 137e145. Erney, D. R., Gillespie, A. M., & Gilvydis, D. M. (1993). Explanation of the matrixinduced chromatographic response enhancement of organophosphorus pesticides during open tubular column gas chromatography with splitless or hot oncolumn injection and flame photometric detection. Journal of Chromatography A, 638, 57e63. Fa, J. Y., Da, L. Y., & Shou, H. Z. (2007). Synthesis of polypiperazine-amide thin-film membrane on PPESK hollow fiber UF membrane. Chinese Chemical Letters, 18, 966e968. € (2014). Pesticide residues in fruits € zde, T. B., Dilek, B. Y. A., Fatih, B., & Semih, O. Go and vegetables from the Aegeanregion, Turkey. Food Chemistry, 160, 379e392. Guo-Rong, X., Jiao-Na, W., & Cong-Ju, L. (2013). Strategies for improving the performance of the polyamide thin film composite (PA-TFC) reverse osmosis (RO) membranes: surface modifications and nanoparticles incorporations. Desalination, 328, 83e100. Hana, L., Sapozhnikovab, Y., & Lehotay, S. J. (2014). Streamlined sample cleanup using combined dispersive solid-phase extraction and in-vial filtration for analysis of pesticides and environmental pollutants in shrimp. Analytica Chimica Acta, 827, 40e46. Hyeyoung, K., Steven, J. L., & Lucía, G. A. (2012). Variability of matrix effects in liquid and gas chromatographyemass spectrometry analysis of pesticide residues after QuEChERS sample preparation of different food crops. Journal of Chromatography A, 1270, 235e245. Juan, L. A., Javier, B., Francisco, J. R., & Carolina, G. (2009). Removal of phenyl-urea herbicides in natural waters by UF membranes: permeate flux, analysis of resistances and rejection coefficients. Separation and Purification Technology, 65, 322e330. Kai Yu, W., & Tai-Shung, C. (2005). The characterization of flat composite nanofiltration membranes and their applications in the separation of Cephalexin. Journal of Membrane Science, 247, 37e50. Konstantinos, V. P., Anastasios, J. K., Thomas, W., & Thomas, M. (2006). A study of

9

selected herbicides retention by nanofiltration membranesdthe role of organic fouling. Journal of Membrane Science, 284, 291e300. Kyu-hong, A., Kyung-guen, S., Ho-young, C., & Ick-tae, Y. (1999). Removal of ions in nickel electroplating rinse water using low-pressure nanofiltration. Desalination, 122, 77e84. Lijun, H., Yelena, S. Y., & Steven, J. L. (2014). Streamlined sample cleanup using combined dispersive solid-phase extraction and in-vial filtration for analysis of pesticides and environmental pollutants in shrimp. Analytica Chimica Acta, 827, 40e46. Patrick, S., & Philip, L. (2013). Characterisation of organic solvent nanofiltration membranes in multi-component mixtures: phenomena-based modelling and membrane modelling maps. Journal of Membrane Science, 445, 183e199. Patrick, S., Tulay, K., & Philip, L. (2013). Characterisation of organic solvent nanofiltration membranes in multi- componentmixtures: membrane rejection maps and membrane selectivity maps for conceptual process design. Journal of Membrane Science, 429, 103e120. Plakas, K. V., Karabelas, A. J., Wintgens, T., & Melin, T. (2006). A study of selected herbicides retention by nanofiltration membranesdthe role of organic fouling. Journal of Membrane Science, 284, 291e300. Raphaella, P. C., Fabiano, A. S. O., Fernando, D. M., Gilsara, S., Wesley, R. D. S., & Renata, P. L. (2013). Development and method validation for determination of 128 pesticides in bananas by modified QuEChERS and UHPLCeMS/MS analysis. Food Control, 33, 413e423. Robert, S. (1995). Membrane separation technology: principles and applications. Amsterdam: Elsevier. Shiao-Shing, C., James, S. T., Luke, A. M., & Charles, D. N. (2004). Influence of molecular weight, molecular size, flux, and recovery for aromatic pesticide removal by nanofiltration membranes. Desalination, 160, 103e111. Siweon, C., Sooyeon, K., Jinyoung, S., MeeKyung, K., & Jeonghan, K. (2015). Development and verification for analysis of pesticides in eggs and egg products using QuEChERS and LCeMS/MS. Food Chemistry, 173, 1236e1242. Słowik-Borowiec, M., Szpyrka, E., & Walorczyk, S. (2015). Gas chromatographic determination of pestcide residues in white mustard. Food Chemistry, 173, 997e1005. Steven, J. L., Kyungae, S., Hyeyoung, K., Urairat, K., Wusheng, F., Katerina, M., et al. (2010). Comparison of QuEChERS sample preparation methods for the analysis of pesticide residues in fruits and vegetables. Journal of Chromatography A, 1217, 2548e2560. Tibor, C., Esther, F., Zdenek, D., Ivan, M., & Adam, E. (2004). Chromatographic determination of herbicide residues in various matrices. Biomedical Chromatography, 18, 350e359. Urairat, K., Steven, J. L., & Natchanun, L. (2011). Fast, low-pressure gas chromatography triple quadrupole tandem mass spectrometry for analysis of 150 pesticide residues in fruits and vegetables. Journal of Chromatography A, 1218, 7039e7050. Yeomin, Y., Paul, W., Shane, A. S., & Eric, C. W. (2006). Nanofiltration and ultrafiltration of endocrine disrupting compounds, pharmaceuticals and personal care products. Journal of Membrane Science, 270, 88e100. Zeying, H., Lu, W., Yi, P., Ming, L., Wenwen, W., & Xiaowei, L. (2015). Multiresidue analysis of over 200 pesticides in cereals using a QuEChERS and gas chromatographyetandem mass spectrometry-based method. Food Chemistry, 169, 372e380.