Chromatographia

, Volume 73, Issue 3, pp 393–401

Combination of Extraction by Silylated Vessel-Dispersive Liquid–Liquid Microextraction as a High-Enrichment Factor Technique: Optimization and Application in Preconcentration of Some Triazole Pesticides from Aqueous Samples Followed by GC-FID Determination

Authors

    • Department of Analytical Chemistry, Faculty of ChemistryUniversity of Tabriz
  • Morteza Bahram
    • Department of Chemistry, Faculty of ScienceUrmia University
  • Farshad Jafary
    • Department of Analytical Chemistry, Faculty of ChemistryUniversity of Tabriz
  • Mehdi Bamorowat
    • GYAH Corporation
Full Short Communication

DOI: 10.1007/s10337-010-1895-0

Cite this article as:
Farajzadeh, M.A., Bahram, M., Jafary, F. et al. Chromatographia (2011) 73: 393. doi:10.1007/s10337-010-1895-0
  • 200 Views

Abstract

This study describes an extraction method based on silylated extraction vessel-dispersive liquid–liquid microextraction (SEV-DLLME) for preconcentration of some triazole pesticides (penconazole, hexaconazole, tebuconazole, diniconazole, triticonazole, and difenconazole) from aqueous samples. For this purpose, the interior surface of funnel-shaped extraction vessel is activated by concentrated NaOH and HCl solutions, silylated by trimethylchlorosilane (TMCS) and used in extraction of the analytes from a relatively high volume of aqueous sample. The adsorbed analytes are desorbed by methanol, which acts as a dispersive solvent in the following DLLME method. In the first step, the effects of different factors i.e., concentrations of NaOH, HCl, and silylated agent and their contact times were studied using central composite design (CCD) and response surface method. Extraction time, extraction solvent (chloroform) volume, dispersive solvent (methanol) volume, centrifugation rate and time, and salting-out effect in DLLME procedure were optimized in the same way using CCD, in the second step. High enrichment factors (EFs) (more than 1,000 in most cases) and low detection limits (at sub μg L−1 level) are attainable by using gas chromatography-flame ionization detection. The repeatability and reproducibility of the proposed method are good and the relative standard deviations (RSD %) for six repeated experiments (C = 100 μg L−1 of each pesticide) are less than 7.25%. Finally, the method was successfully applied in determination of analytes in some aqueous samples such as wastewater, well water, and some fruit juice samples.

Keywords

Dispersive liquid–liquid microextractionEnrichment factorGas chromatographySilylated extraction vesselTriazole pesticides

Introduction

Sample preparation is one of the most important steps in a chemical analysis. The objective of this challenging and critical step is to transfer the analyte into a form that is pre-purified, concentrated, and compatible with the analytical system [1]. Sample preparation typically takes 80% of the total analysis time. Sample preparation in analytical methods regularly employs liquid–liquid extraction (LLE) [2] and solid phase extraction (SPE) [3]. In contrast with relatively fast chromatographic analyses, conventional sample preparation approaches are still highly labor-intensive and time-consuming, consisting of many steps. For this reason, many new sample preparation techniques have been developed over the last decade.

Besides widespread conventional and automatic SPE and LLE techniques, newly developed sample preparation techniques including solid phase microextraction [4], liquid–liquid microextraction [5], pressurized liquid extraction [6], molecularly imprinted polymer [7], salting-out liquid–liquid extraction [8], stir bar sorptive extraction [9] and others. Recent investigations have focused on the development of methods to reduce the sample volume required, the analysis time, the cost and the solvent consumption. To eliminate organic solvent consumption a variety of novel materials are prepared through the chemical modification of silica gel with organic functionalities, and used in solid/liquid extractions [1017]. This application is said to be more environmental friendly.

In 2006, a novel microextraction technique named dispersive liquid–liquid microextraction (DLLME) was developed by Assadi et al. [18]. DLLME is a miniaturized liquid–liquid extraction that uses microliter volume of an extraction solvent. In DLLME, a water-immiscible extraction solvent dissolved in a water-miscible dispersive solvent is rapidly injected into an aqueous solution by syringe. A cloudy solution containing fine droplets of extraction solvent dispersed entirely in the aqueous phase is formed. The analytes in the sample are extracted into the fine droplets, which are further separated by centrifugation, and the enriched analytes in the sedimented phase are determined by either chromatographic or spectrometric methods. The advantages of the DLLME method are short extraction time, low cost, simplicity of operation and high enrichment factors. DLLME has been applied in preconcentration of different compounds such as antioxidants [19], polycyclic aromatic hydrocarbons [20], sulfonylurea herbicides [21], polychlorinated biphenyls [22], butyl and phenyltin compounds [23], lead [24], parabens [25], heterocyclic insecticides [26], and nitroaromatic compounds [27] in different real samples.

The aim of this work is to combine extraction by a silylated vessel with DLLME for the extraction of the selected triazole pesticides from different aqueous samples to achieve very high enrichment factors and lower detection limits. Figures of merit for the proposed approach are reported using gas chromatography-flame ionization detection. The effects of different parameters on the extraction are thoroughly discussed. In this work, the parameters affecting the proposed extraction procedure are optimized using central composite design (CCD). Central composite designs are response surface designs that can fit a full quadratic model [2832].

Experimental

Apparatus

A gas chromatograph (GC-15A, Shimadzu, Japan) with a split/splitless injector system, and a flame ionization detector (FID) was used for separation and determination of the selected triazole pesticides. Helium (99.999%, Gulf Cryo, United Arab Emirates) was used as the carrier gas at a constant linear velocity of 30 cm s−1. The injection port was held at 250 °C and used in the splitless mode with a purge time of 30 s. Separation was carried out on a PT-5 capillary column (30 m × 0.25 mm i.d., and film thickness 0.25 μm). The oven temperature was programmed as follows: initial temperature 100 °C (held 2 min), from 100 to 250 °C at a rate of 15 °C/min, and held at 250 °C for 5 min. The total time for one GC run was 19 min. The FID temperature was maintained at 250 °C. Hydrogen gas was generated with a hydrogen generator (OPGU-1500S, Shimadzu, Japan) for FID at a flow rate of 40 mL min−1. The flow rate of air for FID was 300 mL min−1.

For acceleration phases separation, a Tuttilingen D-7200 centrifuge (Hettich, Germany) was used. A rotary shaker (Tokyo Rikakikai, Japan) was utilized in conditioning extraction vessel and extraction of analytes from aqueous solutions. Home-made glass extraction vessels were the same as extraction funnels without stopcock. The volume of extracting vessels was 110 ± 5 mL.

Chemicals and Solutions

Chloroform, methanol, acetone, dichloromethane, carbon tetrachloride, toluene, trimethylchlorosilane, dimethylchlorooctadecylsilane, hydrochloric acid, sodium hydroxide, and sodium chloride were high purity grade from Merck (Darmstadt, Germany). Penconazole, hexaconazole, diniconazole, tebuconazole, triticonazole, and difenconazole were a gift from Gyah Corporation (Karadj, Iran). Deionized water (Ghazi Co., Tabriz, Iran) was used throughout the study.

Hydrochloric acid and sodium hydroxide solutions were prepared in deionized water at concentrations 0.1, 2.1, 5.0, 7.1, and 10 M. Trimethylchlorosilane solutions 0.2, 4.2, 10.0, 14.2, and 20% (v/v) were prepared in toluene.

Stock solutions of pesticides (except diniconazole) were prepared in methanol at a concentration of 1,000 mg L−1. Working standard solutions were prepared daily by appropriate dilutions with deionized water. Diniconazole solution 2,000 mg L−1 was prepared in chloroform and used as internal standard during optimization of experimental conditions.

Real Samples

A wastewater sample from a unit producing the selected pesticides (Karadj, Iran) and well water from a well located in the zone of that unit were used as samples. The samples were stored in polyethylene bottles and held in a refrigerator until analysis. Two apple juice samples and two grape juice samples were supplied from local supermarkets (Tabriz, Iran). They were analyzed immediately after opening. All samples were used without any filtration. Wastewater sample and fruit juice samples were diluted with deionized water at ratios of 1:5 and 1:10, respectively.

Procedures

Silylation of Internal Surface of Extraction Vessel

As mentioned above, an extraction vessel the same as an extraction funnel without stopcock was used in this study. Initially, activation of its interior surface was performed by sodium hydroxide and hydrochloric acid solutions. For this purpose, 20 mL sodium hydroxide solution (6 M) was transferred into the vessel and was shaken at a rate of 200 rpm for 2.5 h at room temperature. After that, the vessel was washed with deionized water and then 20 mL hydrochloric acid solution (8 M) was transferred into the vessel. The contact time of HCl solution with the extraction vessel was 2 h (shaking speed 200 rpm). The vessel was washed by deionized water and acetone, respectively, and dried at room temperature. Silylation agent, 10 mL trimethylchlorosilane 7% (v/v) in toluene, was added and shaken for 3 h at 200 rpm. Finally, the vessel was washed with toluene, acetone and deionized water, respectively.

Extraction of the Selected Pesticides from Standard and Real Samples

100 mL of standard solution (containing 0–2 mg of each pesticide, except diniconazole), real sample or diluted sample was transferred into the extraction vessel along with 30 g sodium chloride and it was shaken for 25 min at a rate of 200 rpm. After that, the solution was removed and 1.5 mL methanol was transferred into the vessel. After manually shaking for 30 s, 20 μL chloroform was added. The mixture was withdrawn into a 5-mL syringe and injected rapidly into a test tube (12 mL capacity with conic bottom) containing 5 mL sodium chloride solution 20% (w/v). The obtained turbid solution was centrifuged for 5 min at a rate of 6,000 rpm. The sedimented organic phase (10 ± 1 μL) was transferred into a 1-mL vial (conic bottom), 10 μL of the internal standard solution (diniconazole 2,000 mg L−1 in chloroform) was added, and 0.5 μL of the obtained solution was injected into GC. It is noted that diniconazole was used as an internal standard in optimization of experimental conditions. In real samples, no internal standard was used.

Statistical Software

Essential Regression and Experimental Design for Chemists and Engineers, EREGRESS, as MS Excel Add-In software [31], was used to design the experiments and to model and analyze the results.

Central Composite Design

Central composite design was used to model and optimize the extraction of mentioned triazole pesticides using the proposed method.

Six variables, namely, the concentrations of NaOH (F1), HCl (F2), silylation agent (F3) solutions, the contact times of NaOH (F4), HCl (F5) and silylation agent (F6) solutions in first step were studied using a five-levels CCD and four replicates of center point using a circumscribed central composite design.

The effect of other six parameters i.e., extraction time (F1), extracting solvent, (CHCl3) volume (F2), dispersive solvent (methanol) volume (F3), centrifuge rate (F4), centrifuge time (F5) and salting-out effect (F6) in dispersive liquid–liquid microextraction step were optimized in the same way.

For each of the studied variables, high (coded value: +1) and low (coded value: −1) and middle points were selected as presented in Table 1. Forty-eight experiments were designed using EREGRESS software to build a full quadratic model including linear, quadratic and cross-terms, which can be expressed as Eq. 1. Replicates of the central points (n = 4) were performed to estimate the experimental error.
Table 1

The variables and values used for central composite design for silylation and DLLME steps

 

Variable name

Coded factor levels

−1 (low)

−0.42

0

+0.42

+1 (high)

Silylation step

 F1

NaOH concentration (M)

0.1

2.1

5

7.1

10

 F2

HCl concentration (M)

0.1

2.1

5

7.1

10

 F3

Silylation agent concentration (%, v/v)

0.2

4.2

10

14.2

20

 F4

NaOH contact time (h)

0.1

1.55

2.55

3.55

5

 F5

HCl contact time (h)

0.1

1.55

2.55

3.55

5

 F6

Silylation time (h)

0.1

1.55

2.55

3.55

5

DLLME step

 F1

Extraction time (min)

1

18

30

43

60

 F2

CHCl3 volume (μL)

10

25

35

45

60

 F3

Methanol volume (mL)

0.5

0.9

1.5

1.9

2.5

 F4

Centrifuge rate (rpm)

1,000

2,000

3,500

5,000

6,000

 F5

Centrifuge time (min)

1

3

6

9

11

 F6

Salt amount (%, w/v)

1

7

15

22

30

$$ \begin{aligned} {\text{Response}} & = b0 + b1*F1 + b2*F2 + b3*F3 + b4*F4 + b5*F5 + b6*F6 + b7*F1*F1 \\ & \quad + b8*F2*F2 + b9*F3*F3 + b10*F4*F4 + b11*F5*F5 + b12*F6*F6 + b13*F1*F2 \\ & \quad + b14*F1*F3 + b15*F1*F4 + b16*F1*F5 + b17*F1*F6 + b18*F2*F3 + \\ & \quad b19*F2*F4 + b20*F2*F5 + b21*F2*F6 + b22*F3*F4 + b23*F3*F5 + b24*F3*F6 \\ & \quad + b25*F4*F5 + b26*F4*F6 + b27*F5*F6 \\ \end{aligned} $$
(1)

Within Eq. 1, F1F6 are the variable parameters, and b0b27 are the coefficient values obtained through MLR using EREGRESS. The response surface plots were obtained through a statistical process that described the design and the modeled CCD data. Response surface methodologies graphically illustrate relationships between parameters and responses and are the way to obtain an exact optimum [29, 3337].

The statistical significance of the models prediction was evaluated by the analysis of variance and least squares techniques.

Results and Discussion

In this study, combination of extraction by silylated extraction vessel with DLLME as a new preconcentration technique was applied for enrichment of several triazole pesticides from aqueous samples. Gas chromatography along with flame ionization detection was used for quantitation of the analytes after preconcentration. It was necessary to investigate the effect of all parameters that can probably influence the performance of the both extraction steps. Here, optimizations were performed via central composite design method. Effective parameters such as dispersive and extracting solvents volumes, sample size, pH, etc., were studied and the optimum conditions were obtained.

Selection of Dispersive and Extraction Solvents in DLLME Technique

In a previous work [38], the authors proposed a preconcentration procedure using stir bar sorptive extraction followed by the DLLME technique to enrich and analyze the same triazole pesticides selected in this study. Therefore, chloroform as extraction solvent and methanol as dispersive solvent for the DLLME step were selected on the basis of the results obtained from the former work. It is noted that in the previous work, octadecylsilane powder coated on a stir bar using a binder (polyvinyl chloride) was used as a sorbent. Due to the limited surface of the stir bar, recoveries and enrichment factors of analytes were relatively low. In the present study, it is expected that high enrichment factors and low limits of detection to be attainable. In addition, simplicity, and rapidity of the extraction method are other advantages compared to the stir bar sorptive extraction.

Experimental Design

Three antagonist criteria namely enrichment factor (EF), recovery (R) and sedimented phase volume (SPV) were investigated as responses in order to optimize six independent variables, namely the concentrations of NaOH (F1), HCl (F2), silylation agent (F3) solutions and the contact times of NaOH (F4), HCl (F5) and silylation agent (F6) solutions in first step and extraction time (F1), extracting solvent (CHCl3) volume (F2), dispersive solvent (methanol) volume (F3), centrifuge rate (F4), centrifuge time (F5) and salting-out effect (F6) in DLLME step (second step). Table 1 present the levels of coded and actual experimental variables that were tested.

The aims of the CCD strategy were: (1) to maximize the enrichment factor (EF) and recovery (R) and gives the optimum conditions to achieve suitable amount of sedimented phase volume (SPV); (2) to determine which variables have a higher impact on extraction recovery and enrichment factor; (3) to give an insight on the robustness of the method close to the optimum conditions and (4) eventually show interactions between the variables.

As stated, a central composite experiment design allows estimation of a full quadratic model such as Eq. 1. These kinds of designs are easy to construct since they are based on multilevel factorials that have been augmented with the center point and 2n (n is the number of studied variables) extra star points. The repeatability of the method can be assessed using the optimum obtained through a CCD.

Central composite designs are of three types. Circumscribed designs consist of cube points at the corners of a unit cube that is the product of the intervals [−1, 1], star points along the axes at outside of the cube, and center points at the origin. Inscribed designs are as described above, but scaled so that the star points take the values −1 and +1, and the cube points lie in the interior of the cube. Faced designs have the star points on the faces of the cube. Faced designs have three levels per factor, in contrast with the other types, which have five levels per factor.

A full quadratic model including all terms as in Eq. 1 was used to construct the primary model and then the insignificant terms (p > 0.05) are eliminated from the model through ‘backward elimination’ process. The main characteristics of reduced models are obtained (not shown). Since R always decreases when a regression variable is eliminated from a regression model, in statistical modeling the adjusted R, and R for prediction which takes the number of regression variables into account, are usually selected [31, 37]. R, adjusted R, and R for prediction together are very convenient to get a quick impression of the overall fit of the model and the predictive power based on one data point removed. In a good model, these parameters should not be too different from each other. However, for small data sets, it is very likely that every data point is influential. In these cases, a high value for prediction R cannot be expected [31].

From the obtained results there is not too differences between R values which revealed that the experimental data shows a good fit with the second-order polynomial equations. This allows us to further use the response surface as a predictive tool to obtain responses over the whole parameter uncertainty range.

From the obtained results for silylation step, the following results could be obtained:
  • For all of components modeling, all terms, namely, the concentrations of NaOH (F1), HCl (F2), silylation agent (F3) solutions and the contact time of NaOH (F4), the contact time of HCl (F5) and the silylation time (F6) have significant effects on the response (peak area) but the effect of contact time of HCl (F5) is very low.

  • The significant interactions variables for silylation step are found as F2*F1, F2*F5, F2*F6, F1*F3, F1*F5, and F3*F6.

  • Among the interaction terms, the interaction of F3 and F6 is very significant. For silylation with enough contact time of silylation agent, very low concentration of agent will be necessary. On the other hand, for high concentration of silylation agent the contact time of silylation agent will be ineffective.

  • In addition, it was concluded that the contact time of NaOH (F4) is significant for components triticonazole, tebuconazole, difenconazole (1) and difenconazole (2) while have little effect in components hexaconazole and penconazole modeling. This little effect also completely depends to concentration of NaOH. It is noted that difenconazole has cis and trans isomers and gives split peaks in standard and sample solutions. The preferred approach for reporting of these isomeric compounds is to integrate the peaks separately [difenconazole (1) and difenconazole (2)].

Furthermore, from the constructed models for DLLME step, the following results were concluded:
  • There are significant interactions between centrifugation rate and time.

  • Also, the extraction time have interaction with the time of centrifugation.

  • Another important interaction was found as salting-out effect and extraction time. The results showed that in order to reach the high enrichment factor as well as high recovery and sedimented phase volume high concentrations of salt should be used.

  • The volume of extraction solvent has antagonist effect; by increasing the volume of chloroform, the enrichment factor is decreased while the recovery is increased.

  • The dispersive solvent (methanol) volume has contrary effect with extraction solvent (chloroform) volume.

Response Surface and Selection of Optimum Conditions

The obtained regression models were used to calculate the surface for each response variable separately.

The selection of optimum conditions of the proposed method is possible from the response surface plots (not shown). Such conditions can be chosen based on the responses that give the highest peak area for each of investigated component. The conditions that meet this requirement are as follows: NaOH concentration, 5–7 M; HCl concentration, >6 M; silylation agent concentration, 6–8% (v/v); NaOH contact time, 2–3 h; HCl contact time, 2 h and silylation time, 2–4 h.

The response surface for optimization of DLLME procedure (not shown) revealed that increasing the volume of extracting solvent, CHCl3 (F2), has antagonist effect. That is increasing increases recovery while decreasing the enrichment factor.

The optimum conditions of the proposed method was chosen based on the responses that meet the following criteria: recovery near to 100%, sedimented phase volume 10 μL and enrichment factor as high as possible, the conditions that meet these requirements are as follows: extraction time, 20–30 min; CHCl3 volume, 20 μL; methanol volume, 1–1.5 mL; centrifuge rate, 5,000–6,000 rpm; extraction time, 5 min and salt concentration, 30% (w/v).

Optimization of Other Parameters

Other parameters such as extraction vessel volume, silylation agent kind, and aqueous sample size were studied before application of the proposed method in real samples analysis. To study the effect of extraction vessel volume (sample size), three vessels with different volumes (capacities 100, 250, and 500 mL) were selected, and after silylation, were used in sample preparation. The obtained results show that by increasing extraction vessel volume and sample volume, the analytical signals and enrichment factors of analytes increase. By using the 500-mL extraction vessel, high enrichment factors up to 16,000 are attainable. It is noted that working with high volume of sample such as 500 mL is difficult. However, when the sample is dilute and there is no limitation in providing high volume of sample, the use of bigger extraction vessel is advisable. In this study, the 100-mL extraction vessel was selected for further experiments. To study the effect of sample size, different volumes of standard solutions ranging 10–100 mL (containing 100 μg L−1 of each pesticide) were used. The highest recoveries are obtained in low solution volume (10 mL) whereas the highest enrichment factors and the reasonable recoveries are achievable in solution volume 50 mL. Therefore, 50 mL as sample size was selected in the following experiments.

To study the effect of silylation agent kind, three experiments were carried out:

(1) Trimethylchlorosilane (TMCS) was used as the silylation agent as in the above experiments, (2) dimethylchlorooctadecylsilane (DMCODS) was used instead of TMCS, and (3) after silylation by DMCODS, end-capping was performed by TMCS.

Three above-silylated vessels were used in extraction of the analytes from aqueous solutions. The obtained results showed that the vessel, which is silylated by DMCODS, could extract the analytes much better as compared to the others. However, when considering the cost of DMCODS and the high concentration of the silylation agent (10 mL 8%, w/v or v/v in toluene), TMCS was used in this study instead of DMCODS.

Analytical Features of the Proposed Method

To assess the analytical characteristics of the method, some quantitative parameters including dynamic ranges of calibration graphs, correlation coefficients, limits of detection (LOD), limits of quantification (LOQ), relative standard deviations (RSD), enrichment factors (EF) and recoveries (R) are calculated. The calibration graphs are linear in the concentration range of 0.5–20,000 μg L−1 for penconazole, hexaconazole, and tebuconazole, 5–20,000 μg L−1 for triticonazole and difenconazole (2) and 2–20,000 μg L−1 for difenconazole (1). In all cases correlation coefficients of calibration graphs are higher than 0.99. The other analytical parameters of the method are summarized in Table 2.
Table 2

Analytical features of the proposed method

Analyte

LODa (μg L−1)

LOQb (μg L−1)

RSDc (%)

RSDd (%)

EF ± SDe

R (%) ± SDf

Penconazole

0.11

0.37

4.42

4.80

2738 ± 17

55 ± 0.34

Hexaconazole

0.09

0.30

5.26

5.26

1580 ± 14

32 ± 0.28

Tebuconazole

0.14

0.47

3.14

3.43

482 ± 7

10 ± 0.14

Triticonazole

0.78

2.60

3.69

2.51

974 ± 11

19 ± 0.22

Difenconazole (1)

0.69

2.30

6.53

4.03

1048 ± 24

21 ± 0.48

Difenconazole (2)

1.04

3.47

7.25

4.97

1142 ± 30

23 ± 0.60

aLimit of detection, S/N = 3

bLimit of quantification, S/N = 10

cRelative standard deviation for different extraction vessels, n = 6, C = 100 μg L−1, each pesticide (reproducibility)

dRelative standard deviation for different experiments performed using one extraction vessel, n = 6, C = 100 μg L−1, each pesticide (repeatability)

eMean enrichment factor ± standard deviation (n = 3)

fMean recovery ± standard deviation (n = 3)

Low LODs and LOQs, high enrichment factors and good repeatability and reproducibility are main points of the proposed method. Enrichment factors are higher than 1,000 (except for tebuconazole and triticonazole) which lead to LODs for GC-FID are obtained at sub μg L−1 range [except difenconazole (2)]. Relative standard deviations for one extraction vessel in extraction of analytes from six similar solutions (C = 100 μg L−1, each pesticide) are less than 5.26% and those for six different extraction vessels are less than 7.25% which indicate that the repeatability and reproducibility of the method are good. Accuracy of the method was investigated using added-found method. In this case, relative recoveries (in comparison with recoveries obtained from application of the proposed method on standard solutions at the same concentrations added to samples) were obtained in three concentrations (50, 100, and 500 μg L−1, each analyte). They were in the range 86–103% in the cases of grape and apple juice samples and 95–102% in well water and wastewater samples.

Applications

In order to evaluate performance of the method in preconcentration of the selected pesticides, some aqueous samples including one wastewater from an unit which produces the studied pesticides, one well water (in the zone of the pesticides producing unit), two apple juice samples and two grape juice samples were chosen and the proposed method was performed on them. In wastewater and well water samples, all the studied analytes were present in the concentrations higher then LODs of the method. In one of the grape juice samples, three of the studied pesticides (penconazole, hexaconazole, and tebuconazole) were in the concentrations higher than LODs. The obtained concentrations for the pesticides in real samples are summarized in Table 3.
Table 3

The obtained concentrations of the studied pesticides in real samples

Pesticide

Concentration (mg L−1) found in

Wastewater

Well water

Grape juice (1)

Grape juice (2)

Apple juice (1)

Apple juice (2)

Penconazole

7.78

0.12

0.007

n.d.

n.d.

n.d.

Hexaconazole

4.53

0.25

0.097

n.d.

n.d.

n.d.

Diniconazole

0.56

0.08

n.d.

n.d.

n.d.

n.d.

Tebuconazole

1.97

0.43

0.005

n.d.

n.d.

n.d.

Triticonazole

0.53

0.17

n.d.

n.d.

n.d.

n.d.

Difenconazole (1)

0.86

0.09

n.d.

n.d.

n.d.

n.d.

Difenconazole (2)

1.08

0.11

n.d.

n.d.

n.d.

n.d.

n.d. not detected

Comparison of the Proposed Method with Other Methods

Table 4 lists relative standard deviation, limit of detection, limit of quantification and enrichment factor for the reported methods and those of the proposed method applied for quantification of analytes. The presented method has high enrichment factors compared to the others. LOQs of the method are less than those of all reported methods used in determination of the studied pesticides. The repeatability of the proposed method is better than or comparable with other methods.
Table 4

Comparison of the proposed method with some other methods used in analytes determination

Pesticide

Sample

RSD (%)a

LODb

LOQc

EFd

Method

Reference

Penconazole

Soya grain

2.2

 

100 μg L−1

QuEChERS-LC–MS/MSe

[39]

Hexaconazole

12.1

 

50 μg L−1

 

Tebuconazole

5.4

100 μg L−1

 

Triticonazole

10.8

100 μg L−1

 

Difenconazole

6.3

 

Penconazole

Wine

60 μg mL−1

190

SPE-DLLME-GC-ECDf

[40]

Diniconazole

40 μg mL−1

188

 

Difenconazole

250 μg mL−1

254

 

Penconazole

Wine

20 ng mL−1

190

SPE-DLLME-GC-MSg

[40]

Diniconazole

30 ng mL−1

188

 

Difenconazole

100 ng mL−1

254

 

Tebuconazole

Water

7.7

8.3 ng L−1

SBSE-TD-GC-MSh

[41]

Difenconazole

 

9

17 ng L−1

 

Tebuconazole

Ground water

0.04 μg L−1

LLE-SPE-GC-IT-MSi

[42]

Hexaconazole

Crops

0.0005 ng kg−1

0.005 ng kg−1

SPE-GC-ECDj

[43]

Tebuconazole

 

0.004 ng kg−1

0.01 ng kg−1

 

Penconazole

Aqueous samples

4.8

0.11 μg L−1

0.37 μg L−1

2,738

SEV-DLLME-GC-FIDk

This method

Hexaconazole

5.26

0.09 μg L−1

0.30 μg L−1

1,580

Tebuconazole

3.43

0.14 μg L−1

0.47 μg L−1

482

Triticonazole

2.51

0.78 μg L−1

2.60 μg L−1

974

Difenconazole (1)

4.03

0.69 μg L−1

2.30 μg L−1

1,048

Difenconazole (2)

4.97

1.04 μg L−1

3.47 μg L−1

1,142

aRelative standard deviation

bLimit of detection

cLimit of quantitation

dEnrichment factor

eQuick easy, cheap, effective, rugged and safe-liquid chromatography–tandem mass spectrometry

fSolid phase extraction-dispersive liquid–liquid microextraction-gas chromatography- electron capture detector

gSolid phase extraction-dispersive liquid–liquid microextraction-gas chromatography- mass spectrometry

hStir bar sorptive extraction-thermal desorption–gas chromatography–mass spectrometry

iLiquid–liquid extraction–solid phase extraction–gas chromatography–ion trap-mass spectrometry

jSolid phase extraction–gas chromatography–electron capture detector

kSilylated extraction vessel-dispersive liquid–liquid microextraction–gas chromatography–flame ionization detector

Conclusion

In this study, an extraction vessel was designed on the basis of silylation of interior surface of an extraction funnel-shaped vessel. After adsorption of some triazole pesticides on the interior surface of vessel, they desorbed by methanol. Methanol acts as a dispersive solvent in the following DLLME procedure. By combination of two extraction methods, a high-enrichment factor technique was obtained. The method is sensitive, cheep, relatively fast and repeatable. LODs are obtained at sub μg L−1 levels in most cases by the proposed SEV-DLLME-GC-FID method. Finally, this method was successfully used in quantification of the selected pesticides in different aqueous samples.

Acknowledgments

We would like to thank University of Tabriz for supporting this work under the Research Grant Contract No. D/27/3707.

Copyright information

© Springer-Verlag 2011