Article pubs.acs.org/JAFC
Electromembrane Surrounded Solid Phase Microextraction Followed by Injection Port Derivatization and Gas Chromatography−Flame Ionization Detector Analysis for Determination of Acidic Herbicides in Plant Tissue Maryam Rezazadeh,† Yadollah Yamini,*,† Shahram Seidi,‡ Elham Tahmasebi,† and Fatemeh Rezaei† †
Department of Chemistry, Tarbiat Modares University, P. O. Box 14115-175, Tehran, Iran Department of Analytical Chemistry, Faculty of Chemistry, K. N. Toosi University of Technology, Tehran 16315-1355, Iran
‡
ABSTRACT: Electromembrane surrounded solid phase microextraction (EM-SPME) of acidic herbicides was studied for the first time. In order to investigate the capability of this new microextraction technique to analyze acidic targets, chlorophenoxy acid (CPA) herbicides were quantified in plant tissue. 1-Octanol, was sustained in the pores of the wall of a hollow fiber and served as supported liquid membrane (SLM). Other EM-SPME related parameters, including extraction time, applied voltage, and pHs of the sample solution and the acceptor phase, were optimized using experimental design. A 20 min time frame was needed to reach the highest extraction efficiency of the analytes from a 24 mL alkaline sample solution across the organic liquid membrane and into the aqueous acceptor phase through a 50 V electrical field, and to their final adsorption on a carbonaceous anode. In addition to high sample cleanup, which made the proposed method appropriate for analysis of acidic compounds in a complicated media (plant tissue), 4.8% of 2-methyl-4-chlorophenoxyacetic acid (MCPA) and 0.6% of 2,4-dichlorophenoxyacetic acid (2,4-D) were adsorbed on the anode, resulting in suitable detection limits (less than 5 ng mL−1), and admissible repeatability and reproducibility (intra- and interassay precision were in the ranges of 5.2−8.5% and 8.8−12.0%, respectively). Linearity of the method was scrutinized within the ranges of 1.0−500.0 and 10.0−500.0 ng mL−1 for MCPA and 2,4-D, respectively, and coefficients of determination greater than 0.9958 were obtained. Optimal conditions of EM-SPME of the herbicides were employed for analysis of CPAs in whole wheat tissue. KEYWORDS: electromembrane surrounded solid phase microextraction, gas chromatography, acidic compounds, chlorophenoxy acid herbicides, plant tissue
■
INTRODUCTION Chlorophenoxy acid (CPA) herbicides are key regulators, preventing arable land from being overgrown by unwanted plants.1 They are widely used in agriculture to control the growth of broadleaf weeds in crops. Considering CPAs’ potential teratogenic and carcinogenic effects and large scale of application, their monitoring in plant tissues is necessary for protecting consumers’ health. Dirty and complicated samples, such as plant tissue, with solid particles and large amounts of interfering species could not be directly introduced into the analysis instrument. Thus, some sample preparation including sample cleanup, preconcentration of analytes of interest, and (in some cases) analytes derivatization is necessary prior to its analysis. Hollow fiber based extraction techniques which are known as green sample preparation methods could offer all of these purposes.2−9 Application of auxiliary energies to sample preparation techniques could be an alternative way to enhance the efficiency, reduce the operation time, and increase the selectivity. The outcome of recent research, carried out to achieve this aim, has been the introduction of an electrical potential stimulated liquid phase microextraction technique termed electromembrane extraction (EME). 2 In 2006, Pedersen-Bjergaard and Rasmussen presented EME based on the migration of ionized species in an electrical field.2 Utilizing electrical field, as the main driving force, makes EME a simple, © 2014 American Chemical Society
efficient, and rapid extraction method for analysis of ionizable compounds in complicated matrices. Some improvements have been done to EME setup to augment its advantages and overcome its drawbacks, comprising a new setup for exhaustive EME extraction,3,4 parallel EME under stagnant conditions using common batteries,5 simultaneous extraction of acidic and basic drugs at neutral sample pH,6,7 electric current controlled EME,10 development of pulsed electromembrane extraction (PEME),11−13 application of carbon nanotube reinforced hollow fiber to EME,14,15 and some installations for lab-on-achip purposes.16−19 Despite the benefits associated with EME, this microextraction technique suffers from some limitations. One of the most important disadvantages of EME is its incompatibility with gas chromatography (GC) instrument, due to its aqueous acceptor phase. There have been some attempts at coupling EME to GC by transportation of analytes into a final organic solvent. Two-phase EME is one of the solutions represented for this objective.20 However, two-phase extraction decreases the cleanup ability of the method, and it could not be performed in dirty matrices. Moreover, increasing the thickness of the organic phase, in this process, noticeably raises the Received: Revised: Accepted: Published: 3134
January 2, 2014 March 18, 2014 March 24, 2014 March 24, 2014 dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
Figure 1. Equipment used for the EM-SPME method, and mechanism of transportation across liquid−liquidliquid−solid boundaries.
were adsorbed on the solid adsorbent, which was also the anode. After derivatization to enhance the volatility of the CPAs, the carbonaceous sorbent was directly introduced into the GC-FID injection port. The goal of this work is to examine the possibility of employing EM-SPME as an effective and relatively fast extraction technique for analysis of both acidic and basic compounds in complicated matrices.
electrical resistance of the system, which reduces the extraction efficiency as a result of diminishing the strength of the electrical field and lengthening the extraction time.21 EME conjugated with dispersive liquid−liquid microextraction (EME-DLLME) is another route to transferring the analytes into an organic solvent.22,23 However, EME-DLLME is an off-line coupling of two extraction techniques and requires sufficient time for conducting both procedures. Rezazadeh et al. have introduced electromembrane surrounded solid phase microextraction (EM-SPME) as a simple and effective method to benefit from high extraction efficiency, sample cleanup, and fast kinetics related to EME technique as well as solid phase microextraction (SPME) compatibility with GC.24 EM-SPME setup was the same as that of EME, unless one of the electrodes, located in the lumen of the hollow fiber, was substituted with a conductive sorbent. EM-SPME was implemented for extraction of basic drugs utilizing a platinum electrode and a piece of pencil lead as anode and cathode, respectively.24 Thus, the basic analytes migrated in an electrical field from aqueous sample solution through a liquid membrane and into an aqueous acceptor phase, and then they were adsorbed on the solid sorbent, which acted as cathode. It was shown that EM-SPME has a great potential as a microextraction technique and could be established as a simple and inexpensive method for analysis of nonvolatile or ionizable compounds in complex matrices.24 The present study was focused on EM-SPME of acidic herbicides, to explore the capacity of the technique for extraction of this class of analytes by means of a sorbent as anode. Hence, the suggested method followed by GC combined with flame ionization detection (GC-FID) was exploited, for the first time, for extraction of CPA herbicides from plant tissue. In this work, an electrical field was applied to cause the CPAs, which existed in aqueous sample solution as anions, to migrate through the supported liquid membrane (SLM) into an aqueous phase inside the lumen of the hollow fiber. Using organic liquid membrane (SLM), not only the selectivity can be increased but also the sample cleanup highly improves and makes it possible to apply high voltages that increase extraction recoveries. Afterward, the target analytes
■
EXPERIMENTAL SECTION
EM-SPME Equipment. The equipment used for the extraction procedure is illustrated in Figure 1. A 24 mL vial with an internal diameter of 2.5 cm and a height of 5.5 cm was exploited. The platinum electrode employed in this work, with a diameter of 0.25 mm, was acquired from Pars Pelatine (Tehran, Iran). The electrodes were connected to a power supply model 8760T3 with a programmable voltage in the range of 0−600 V and a current output within the range of 0−500 mA from Paya Pajoohesh Pars (Tehran, Iran). During the extraction, the EM-SPME unit was stirred with a speed in the range of 100−1250 rpm via a heater−magnetic stirrer, model 3001 from Heidolph (Kelheim, Germany), using a 1.5 cm × 0.3 cm magnetic bar. A 40 kHz and 0.138 kW ultrasonic water bath with temperature control (Tecno-GazSpA, Parma, Italy) and a Sepand Teb Azma centrifuge (Tehran, Iran) were employed to carry out pretreatment steps for real samples. Chemicals and Materials. 2-Methyl-4-chlorophenoxyacetic acid (MCPA) and 2,4-dichlorophenoxyacetic acid (2,4-D), were purchased from Aldrich (Milwaukee, WI, USA). 1-Octanol, 1-hexanol, 1undecanol, 1-dodecanol, and tetrabutylammonium iodide (TBA), were supplied by Merck (Darmstadt, Germany). All of the chemicals used were of analytical reagent grade. The porous hollow fiber, utilized for supporting the liquid membrane, was a PPQ3/2 polypropylene hollow fiber from Membrana (Wuppertal, Germany), with inner diameter of 0.6 mm, wall thickness of 200 μm, and pore size of 0.2 μm. Ultrapure water was prepared by a Young Lin 370 series aquaMAX purification instrument (Kyounggi-do, Korea). The HB pencil lead fibers with the diameter of 0.3 mm (Owner, Seoul, Korea) were purchased from a local market. The fibers were conditioned via heating under N2 atmosphere from room temperature to 350 °C with a ramp of 3 °C min−1, held for 120 min at 350 °C, and then were allowed to cool to ambient temperature. Standard Solutions and Real Samples. A stock solution, containing 1 mg mL−1 of each analyte, was prepared in methanol and stored at −4 °C, protected from light. Working standard solutions were prepared by dilution of the stock solution with methanol. 3135
dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
The whole wheat plant was analyzed as the real sample. For analysis of the target herbicides in the plant tissue, wheat was harvested after 1 h, 2 days, and 6 days following field spraying. The whole wheat plants were first washed and completely blended. A 20 mL aliquot of a 100 mM NaOH solution was subsequently added to 100 g of the blended sample, and the sample was immersed in the ultrasonic water bath. After 2 min of sonication (at 25 ± 3 °C), the sample was centrifuged at 2500 rpm for 10 min to separate the phases. The final liquid phase was diluted (1:9) with pure water and, following pH adjustment (pH 9.0), was transferred into the EM-SPME vial for analysis. Gas Chromatography Conditions. Separation and detection of MCPA and 2,4-D were executed through an Agilent 7890A gas chromatograph (Palo Alto, CA, USA), equipped with a split−splitless injector, and a flame ionization detector. A 30 m HP-5 Agilent fusedsilica capillary column (0.32 mm, i.d.; 0.25 μm, film thickness) was exploited for separation of the target compounds. Helium (purity, 99.999%) was utilized as carrier gas at the constant flow rate of 2.0 mL min−1. The temperatures of injector and detector were set at 290 and 300 °C, respectively. The injection port was operated at the splitless mode. The oven temperature program was 100 °C for 2 min, increasing to 220 °C with a ramp of 9 °C min−1, then increasing to 280 °C at 100 °C min−1, and eventually holding at 280 °C for 3 min. EM-SPME Procedure. A 24 mL aliquot of the sample solution, containing the analytes in aqueous medium, was transferred into the sample vial. To impregnate the organic liquid membrane in the pores of the hollow fiber wall, a 2.8 cm piece of the hollow fiber was cut out and dipped into 1-octanol for 5 s, and subsequently the excess of the organic solvent was gently wiped away by blowing air with a Hamilton syringe. Pure water, as the acceptor phase, was introduced into the lumen of the hollow fiber by means of a microsyringe, and the lower end of the hollow fiber was mechanically sealed afterward. The pencil lead fiber (the anode) was placed in the lumen of the hollow fiber. After that, the hollow fiber, containing the anode, together with the SLM and the acceptor solution was directed into the sample solution. The platinum cathode was led directly into the sample solution. The electrodes were subsequently connected to the power supply, and the extraction unit was placed on a stirrer with stirring rate of 700 rpm. When the extraction was accomplished, the pencil lead was retracted into the SPME syringe needle. The tractable nature of the pencil lead facilitated its insertion into the SPME syringe. The carbonaceous fiber was well fitted into the syringe by some wearing on top of it, where it was in contact with the needle. Then, the pencil lead was immersed in a concentrated TBA solution, as derivatization reagent, and withdrawn very fast and immediately inserted into the GC injection port for thermal desorption of the analytes at 290 °C for 2 min. The contact time of the fiber and TBA solution should be long enough to form a stagnant thin layer of derivatization reagent around the fiber but it must be too short to prevent dissolving of analytes in TBA solution. Data Analysis and Statistical Methods. Response surface methodology (RSM) is effective for responses that are influenced by many factors as well as by their interactions and was originally described by Box and Wilson.25 Optimization of the parameters affecting the extraction efficiency of the analytes by PEME was performed by a face-centered central composite design. In all cases, design generation and statistical analyses were performed by the software package Design-Expert version 8.0.5 trial for Windows (StateEase, MN, USA).
of SLM composition, compositions of donor and acceptor phases, extraction time, applied voltage, and stirring rate. It is obvious that stirring of the sample solution possesses a positive impact upon the kinetics and efficiency of extraction via enhancing the mass transfer of the analytes and decreasing the thickness of the double layer around the SLM. The influence of the stirring rate on extractability was investigated up to 1250 rpm. A stirring rate of 700 rpm was selected, because of creation of an intense vortex in the sample solution and bubble formation around the hollow fiber at higher rates. The SLM composition was optimized separately, and other EM-SPME variables, comprising extraction time, applied voltage, and pHs of sample solution and acceptor phase, were evaluated by means of experimental design methodology. Derivatization Procedure and GC Analysis. Direct injection port derivatization with ion-pair reagents, known as tetraalkylammonium salt pyrolysis derivatization method, is a rapid and simple alternative to conventional derivatization techniques for aliphatic and aromatic acids.26,27 The product of the reaction of an acidic analyte with tetraalkylammonium salt is a carboxylate ion pair, which is derivatized to its corresponding volatile alkyl ester in a high-temperature GC injection port. RCOOH + N(Bu)4 + X− ↔ RCOO−N(Bu)4 + + HX
(1)
RCOO−N(Bu)4 ↔ RCOO Bu + N(Bu)3
(2)
Two different methods were tested to form the carboxylate ion pair. In the first technique, TBA was added to the acceptor phase of EM-SPME and once the extraction was completed, the pencil lead was directly introduced into the GC injection port. The second method of carboxylate ion pair production was through the immersion of pencil lead in a TBA solution following accomplishment of EM-SPME. Comparison of the attained results demonstrated that good derivatization performance was gained by immersion of the SPME fiber in TBA solution. The main point of the injection port derivatization technique is that both analyte and derivatization reagent should simultaneously exist in the injection port during the process. When TBA was added into the acceptor solution, its adsorption on the surface of the pencil lead was difficult since both of them have positive charges. The charge of TBA is positive in the neutral acceptor phase while the positive charges of pencil lead is attributed to the connection of the fiber to the positive pole of the power supply during the extraction procedure. Therefore, the derivatization efficiency decreases due to reduction of the TBA concentration in the injection port. However, immersion of the fiber into a concentrated TBA solution occurred after completion of the extraction and disconnection of the electrical field. In this way, a thin stagnant layer of TBA solution was formed on the surface of the fiber by means of the porous nature of the pencil lead. When the fiber together with the TBA solution layer is introduced into the GC injection port, the TBA solvent (methanol) is evaporated and thus a considerable amount of TBA is available. Hence, desorption of both TBA and analytes lead to formation of derivatized analytes. The TBA concentration was assessed thereafter, to find the best derivatization conditions. The concentration of the derivatization reagent in methanol was varied in the range of 30−400 mM. After EM-SPME of the analytes, the carbonaceous fiber was immersed in 30 μL of TBA solution and instantly inserted into the GC injection port. It was noticed that
■
RESULTS AND DISCUSSION In order to figure out how the impressive parameters affected the final signals of the analytes, a number of experiments were designed. Two different influential parameters in the final chromatographic signals included EM-SPME related variables and those linked to GC analysis (i.e., derivatization process and desorption conditions). GC instrument associated parameters were optimized in terms of derivatization procedure, TBA concentration, injection port temperature, carrier gas flow rate, and desorption time. EM-SPME affected parameters consisted 3136
dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
Figure 2. (A) Optimization of membrane composition, (B) response surface of the square root of the normalized peak area vs pH of donor phase and pH of acceptor phase, (C) response surface of the square root of the normalized peak area vs pH of the acceptor phase and applied voltage, and (D) response surface of the square root of the normalized peak area vs extraction time and applied voltage.
extraction time are directly affected by the composition of organic liquid membrane. Analytes should have a proper affinity for SLM so that adequate selectivity and favorable extraction efficiency could be acquired. In fact, SLM provides the major electrical resistance for the system and the range of applied voltage could be extended via raising the electrical resistance of SLM. Accordingly, system stability and extractability may improve utilizing organic solvents with high electrical resistance such as SLM. Nevertheless, the organic phase should possess satisfactory conductivity to enable ion transportation across the phases. Regarding all of these factors, 1-hexanol, 1-octanol, 1undecanol, and 1-dodecanol were tested as the organic liquid membrane. It was proved that long-chain alcohols are the most desirable organic solvents for extraction of acidic compounds.6,28 Among the examined alcohols, 1-hexanol has the least electrical resistance, and a system operated with 1-hexanol, such as the SLM, underwent instability problems owing to increasing the electric current (Figure 2A). Therefore, 1hexanol offered nonrepeatable results, and the chromatographic signals substantially declined, which was probably caused by the enhancement of electrochemical reactions due to increasing the electric current of the system. On the other hand, 1-dodecanol could not provide a continuous electric field in the entire system, since it possesses a low electrical conductivity; for this reason, the extractability diminished using 1-dodecanol as the SLM. 1-Undecanol was capable of effectively extracting the 2,4D, although it was not a suitable solvent for MCPA. Ultimately, 1-octanol was chosen as the SLM due its ability to give the best results for both of the analytes. Optimization of Extraction Time, Applied Voltage, and pHs of the Sample Solution and Acceptor Phase Using Central Composite Design. The experiments for optimization of the rest of the EM-SPME affected parameters were modeled through face-centered central composite design
the peak areas of both CPA herbicides initially increased via enhancing the concentration of the derivatization reagent up to 200 mM and then reached a plateau with a further rise in TBA concentration. Therefore, 200 mM TBA in methanol was selected as the derivatization solution. Another effective parameter on the efficiency of the analyte desorption as well as the injection port derivatization process is the injection port temperature that was considered in the range of 250−300 °C. The peak areas of the derivatized analytes were enhanced by raising the temperature to 290 °C. Since no changes in the derivatization yields were observed at higher temperatures, 290 °C was chosen to reach the maximum desorption and derivatization efficiencies and also prevent the entrance of the pencil lead from interfering with compounds going into the GC injection port. The carrier gas flow rate can influence the derivatization efficiency through altering the delay time of the analytes in the injection port of the GC apparatus. The flow rate of the carrier gas was scrutinized in the range of 1−4 mL min−1. The maximum peak areas for butyl esters of the herbicides were achieved at a flow rate of 2 mL min−1. Furthermore, desorption time was explored in the range of 0.5−5.0 min while the desorption temperature was 290 °C. The efficiency of thermal desorption improved as the desorption time was increased from 0.5 to 2.0 min. The chromatographic signals slightly increased after 2.0 min, but desorption of some interferences from the pencil lead fiber resulted in crowded chromatograms. Consequently, the desorption time of 2.0 min was selected. EM-SPME Affecting Parameters. Composition of Organic Liquid Membrane. One of the most significant parameters, which could influence the extractability of analytes in EM-SPME technique, is the SLM composition. Extraction selectivity, range of applied voltage, extraction efficiency, and 3137
dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
Table 1. Experimental Factors, Levels, and Analysis of Variance (ANOVA) for the Response Surface Quadratic Model level factor extraction time (min) applied voltage (V) pH of donor phase pH of acceptor phase source sum of squares block model A: t B: V C: pHd D: pHa AB AC AD BC BD CD ̂ A2 ̂ B2 2̂ C ̂ D2 residual lack of fit pure error cor total R2 adj R2
61.46 204.54 122.78 14.64 2.19 12.12 0.9 1.13 1.1 0.051 1.14 × 10−5 0.96 0.28 5.32 5.03 11.34 6.66 6.3 0.35 272.66 0.9685 0.9317
symbol
(−1)
t V pHd pHa
5 20 7 1
d.f.
mean square
1 14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 12 10 2 27
61.46 14.61 122.78 14.64 2.19 12.12 0.9 1.13 1.1 0.051 1.14 × 10−5 0.96 0.28 5.32 5.03 11.34 0.55 0.63 0.18
(0)
F-value
12:30 50 10 7 p-value
(+1) 20 80 13 13
26.34 221.34 26.4 3.95 21.85 1.62 2.03 1.98 0.093 2.05 × 10−5 1.74 0.5 9.6 9.07 20.44
Electromembrane Surrounded Solid Phase Microextraction Followed by Injection Port Derivatization and Gas Chromatography−Flame Ionization Detector Analysis for Determination of Acidic Herbicides in Plant Tissue Maryam Rezazadeh,† Yadollah Yamini,*,† Shahram Seidi,‡ Elham Tahmasebi,† and Fatemeh Rezaei† †
Department of Chemistry, Tarbiat Modares University, P. O. Box 14115-175, Tehran, Iran Department of Analytical Chemistry, Faculty of Chemistry, K. N. Toosi University of Technology, Tehran 16315-1355, Iran
‡
ABSTRACT: Electromembrane surrounded solid phase microextraction (EM-SPME) of acidic herbicides was studied for the first time. In order to investigate the capability of this new microextraction technique to analyze acidic targets, chlorophenoxy acid (CPA) herbicides were quantified in plant tissue. 1-Octanol, was sustained in the pores of the wall of a hollow fiber and served as supported liquid membrane (SLM). Other EM-SPME related parameters, including extraction time, applied voltage, and pHs of the sample solution and the acceptor phase, were optimized using experimental design. A 20 min time frame was needed to reach the highest extraction efficiency of the analytes from a 24 mL alkaline sample solution across the organic liquid membrane and into the aqueous acceptor phase through a 50 V electrical field, and to their final adsorption on a carbonaceous anode. In addition to high sample cleanup, which made the proposed method appropriate for analysis of acidic compounds in a complicated media (plant tissue), 4.8% of 2-methyl-4-chlorophenoxyacetic acid (MCPA) and 0.6% of 2,4-dichlorophenoxyacetic acid (2,4-D) were adsorbed on the anode, resulting in suitable detection limits (less than 5 ng mL−1), and admissible repeatability and reproducibility (intra- and interassay precision were in the ranges of 5.2−8.5% and 8.8−12.0%, respectively). Linearity of the method was scrutinized within the ranges of 1.0−500.0 and 10.0−500.0 ng mL−1 for MCPA and 2,4-D, respectively, and coefficients of determination greater than 0.9958 were obtained. Optimal conditions of EM-SPME of the herbicides were employed for analysis of CPAs in whole wheat tissue. KEYWORDS: electromembrane surrounded solid phase microextraction, gas chromatography, acidic compounds, chlorophenoxy acid herbicides, plant tissue
■
INTRODUCTION Chlorophenoxy acid (CPA) herbicides are key regulators, preventing arable land from being overgrown by unwanted plants.1 They are widely used in agriculture to control the growth of broadleaf weeds in crops. Considering CPAs’ potential teratogenic and carcinogenic effects and large scale of application, their monitoring in plant tissues is necessary for protecting consumers’ health. Dirty and complicated samples, such as plant tissue, with solid particles and large amounts of interfering species could not be directly introduced into the analysis instrument. Thus, some sample preparation including sample cleanup, preconcentration of analytes of interest, and (in some cases) analytes derivatization is necessary prior to its analysis. Hollow fiber based extraction techniques which are known as green sample preparation methods could offer all of these purposes.2−9 Application of auxiliary energies to sample preparation techniques could be an alternative way to enhance the efficiency, reduce the operation time, and increase the selectivity. The outcome of recent research, carried out to achieve this aim, has been the introduction of an electrical potential stimulated liquid phase microextraction technique termed electromembrane extraction (EME). 2 In 2006, Pedersen-Bjergaard and Rasmussen presented EME based on the migration of ionized species in an electrical field.2 Utilizing electrical field, as the main driving force, makes EME a simple, © 2014 American Chemical Society
efficient, and rapid extraction method for analysis of ionizable compounds in complicated matrices. Some improvements have been done to EME setup to augment its advantages and overcome its drawbacks, comprising a new setup for exhaustive EME extraction,3,4 parallel EME under stagnant conditions using common batteries,5 simultaneous extraction of acidic and basic drugs at neutral sample pH,6,7 electric current controlled EME,10 development of pulsed electromembrane extraction (PEME),11−13 application of carbon nanotube reinforced hollow fiber to EME,14,15 and some installations for lab-on-achip purposes.16−19 Despite the benefits associated with EME, this microextraction technique suffers from some limitations. One of the most important disadvantages of EME is its incompatibility with gas chromatography (GC) instrument, due to its aqueous acceptor phase. There have been some attempts at coupling EME to GC by transportation of analytes into a final organic solvent. Two-phase EME is one of the solutions represented for this objective.20 However, two-phase extraction decreases the cleanup ability of the method, and it could not be performed in dirty matrices. Moreover, increasing the thickness of the organic phase, in this process, noticeably raises the Received: Revised: Accepted: Published: 3134
January 2, 2014 March 18, 2014 March 24, 2014 March 24, 2014 dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
Figure 1. Equipment used for the EM-SPME method, and mechanism of transportation across liquid−liquidliquid−solid boundaries.
were adsorbed on the solid adsorbent, which was also the anode. After derivatization to enhance the volatility of the CPAs, the carbonaceous sorbent was directly introduced into the GC-FID injection port. The goal of this work is to examine the possibility of employing EM-SPME as an effective and relatively fast extraction technique for analysis of both acidic and basic compounds in complicated matrices.
electrical resistance of the system, which reduces the extraction efficiency as a result of diminishing the strength of the electrical field and lengthening the extraction time.21 EME conjugated with dispersive liquid−liquid microextraction (EME-DLLME) is another route to transferring the analytes into an organic solvent.22,23 However, EME-DLLME is an off-line coupling of two extraction techniques and requires sufficient time for conducting both procedures. Rezazadeh et al. have introduced electromembrane surrounded solid phase microextraction (EM-SPME) as a simple and effective method to benefit from high extraction efficiency, sample cleanup, and fast kinetics related to EME technique as well as solid phase microextraction (SPME) compatibility with GC.24 EM-SPME setup was the same as that of EME, unless one of the electrodes, located in the lumen of the hollow fiber, was substituted with a conductive sorbent. EM-SPME was implemented for extraction of basic drugs utilizing a platinum electrode and a piece of pencil lead as anode and cathode, respectively.24 Thus, the basic analytes migrated in an electrical field from aqueous sample solution through a liquid membrane and into an aqueous acceptor phase, and then they were adsorbed on the solid sorbent, which acted as cathode. It was shown that EM-SPME has a great potential as a microextraction technique and could be established as a simple and inexpensive method for analysis of nonvolatile or ionizable compounds in complex matrices.24 The present study was focused on EM-SPME of acidic herbicides, to explore the capacity of the technique for extraction of this class of analytes by means of a sorbent as anode. Hence, the suggested method followed by GC combined with flame ionization detection (GC-FID) was exploited, for the first time, for extraction of CPA herbicides from plant tissue. In this work, an electrical field was applied to cause the CPAs, which existed in aqueous sample solution as anions, to migrate through the supported liquid membrane (SLM) into an aqueous phase inside the lumen of the hollow fiber. Using organic liquid membrane (SLM), not only the selectivity can be increased but also the sample cleanup highly improves and makes it possible to apply high voltages that increase extraction recoveries. Afterward, the target analytes
■
EXPERIMENTAL SECTION
EM-SPME Equipment. The equipment used for the extraction procedure is illustrated in Figure 1. A 24 mL vial with an internal diameter of 2.5 cm and a height of 5.5 cm was exploited. The platinum electrode employed in this work, with a diameter of 0.25 mm, was acquired from Pars Pelatine (Tehran, Iran). The electrodes were connected to a power supply model 8760T3 with a programmable voltage in the range of 0−600 V and a current output within the range of 0−500 mA from Paya Pajoohesh Pars (Tehran, Iran). During the extraction, the EM-SPME unit was stirred with a speed in the range of 100−1250 rpm via a heater−magnetic stirrer, model 3001 from Heidolph (Kelheim, Germany), using a 1.5 cm × 0.3 cm magnetic bar. A 40 kHz and 0.138 kW ultrasonic water bath with temperature control (Tecno-GazSpA, Parma, Italy) and a Sepand Teb Azma centrifuge (Tehran, Iran) were employed to carry out pretreatment steps for real samples. Chemicals and Materials. 2-Methyl-4-chlorophenoxyacetic acid (MCPA) and 2,4-dichlorophenoxyacetic acid (2,4-D), were purchased from Aldrich (Milwaukee, WI, USA). 1-Octanol, 1-hexanol, 1undecanol, 1-dodecanol, and tetrabutylammonium iodide (TBA), were supplied by Merck (Darmstadt, Germany). All of the chemicals used were of analytical reagent grade. The porous hollow fiber, utilized for supporting the liquid membrane, was a PPQ3/2 polypropylene hollow fiber from Membrana (Wuppertal, Germany), with inner diameter of 0.6 mm, wall thickness of 200 μm, and pore size of 0.2 μm. Ultrapure water was prepared by a Young Lin 370 series aquaMAX purification instrument (Kyounggi-do, Korea). The HB pencil lead fibers with the diameter of 0.3 mm (Owner, Seoul, Korea) were purchased from a local market. The fibers were conditioned via heating under N2 atmosphere from room temperature to 350 °C with a ramp of 3 °C min−1, held for 120 min at 350 °C, and then were allowed to cool to ambient temperature. Standard Solutions and Real Samples. A stock solution, containing 1 mg mL−1 of each analyte, was prepared in methanol and stored at −4 °C, protected from light. Working standard solutions were prepared by dilution of the stock solution with methanol. 3135
dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
The whole wheat plant was analyzed as the real sample. For analysis of the target herbicides in the plant tissue, wheat was harvested after 1 h, 2 days, and 6 days following field spraying. The whole wheat plants were first washed and completely blended. A 20 mL aliquot of a 100 mM NaOH solution was subsequently added to 100 g of the blended sample, and the sample was immersed in the ultrasonic water bath. After 2 min of sonication (at 25 ± 3 °C), the sample was centrifuged at 2500 rpm for 10 min to separate the phases. The final liquid phase was diluted (1:9) with pure water and, following pH adjustment (pH 9.0), was transferred into the EM-SPME vial for analysis. Gas Chromatography Conditions. Separation and detection of MCPA and 2,4-D were executed through an Agilent 7890A gas chromatograph (Palo Alto, CA, USA), equipped with a split−splitless injector, and a flame ionization detector. A 30 m HP-5 Agilent fusedsilica capillary column (0.32 mm, i.d.; 0.25 μm, film thickness) was exploited for separation of the target compounds. Helium (purity, 99.999%) was utilized as carrier gas at the constant flow rate of 2.0 mL min−1. The temperatures of injector and detector were set at 290 and 300 °C, respectively. The injection port was operated at the splitless mode. The oven temperature program was 100 °C for 2 min, increasing to 220 °C with a ramp of 9 °C min−1, then increasing to 280 °C at 100 °C min−1, and eventually holding at 280 °C for 3 min. EM-SPME Procedure. A 24 mL aliquot of the sample solution, containing the analytes in aqueous medium, was transferred into the sample vial. To impregnate the organic liquid membrane in the pores of the hollow fiber wall, a 2.8 cm piece of the hollow fiber was cut out and dipped into 1-octanol for 5 s, and subsequently the excess of the organic solvent was gently wiped away by blowing air with a Hamilton syringe. Pure water, as the acceptor phase, was introduced into the lumen of the hollow fiber by means of a microsyringe, and the lower end of the hollow fiber was mechanically sealed afterward. The pencil lead fiber (the anode) was placed in the lumen of the hollow fiber. After that, the hollow fiber, containing the anode, together with the SLM and the acceptor solution was directed into the sample solution. The platinum cathode was led directly into the sample solution. The electrodes were subsequently connected to the power supply, and the extraction unit was placed on a stirrer with stirring rate of 700 rpm. When the extraction was accomplished, the pencil lead was retracted into the SPME syringe needle. The tractable nature of the pencil lead facilitated its insertion into the SPME syringe. The carbonaceous fiber was well fitted into the syringe by some wearing on top of it, where it was in contact with the needle. Then, the pencil lead was immersed in a concentrated TBA solution, as derivatization reagent, and withdrawn very fast and immediately inserted into the GC injection port for thermal desorption of the analytes at 290 °C for 2 min. The contact time of the fiber and TBA solution should be long enough to form a stagnant thin layer of derivatization reagent around the fiber but it must be too short to prevent dissolving of analytes in TBA solution. Data Analysis and Statistical Methods. Response surface methodology (RSM) is effective for responses that are influenced by many factors as well as by their interactions and was originally described by Box and Wilson.25 Optimization of the parameters affecting the extraction efficiency of the analytes by PEME was performed by a face-centered central composite design. In all cases, design generation and statistical analyses were performed by the software package Design-Expert version 8.0.5 trial for Windows (StateEase, MN, USA).
of SLM composition, compositions of donor and acceptor phases, extraction time, applied voltage, and stirring rate. It is obvious that stirring of the sample solution possesses a positive impact upon the kinetics and efficiency of extraction via enhancing the mass transfer of the analytes and decreasing the thickness of the double layer around the SLM. The influence of the stirring rate on extractability was investigated up to 1250 rpm. A stirring rate of 700 rpm was selected, because of creation of an intense vortex in the sample solution and bubble formation around the hollow fiber at higher rates. The SLM composition was optimized separately, and other EM-SPME variables, comprising extraction time, applied voltage, and pHs of sample solution and acceptor phase, were evaluated by means of experimental design methodology. Derivatization Procedure and GC Analysis. Direct injection port derivatization with ion-pair reagents, known as tetraalkylammonium salt pyrolysis derivatization method, is a rapid and simple alternative to conventional derivatization techniques for aliphatic and aromatic acids.26,27 The product of the reaction of an acidic analyte with tetraalkylammonium salt is a carboxylate ion pair, which is derivatized to its corresponding volatile alkyl ester in a high-temperature GC injection port. RCOOH + N(Bu)4 + X− ↔ RCOO−N(Bu)4 + + HX
(1)
RCOO−N(Bu)4 ↔ RCOO Bu + N(Bu)3
(2)
Two different methods were tested to form the carboxylate ion pair. In the first technique, TBA was added to the acceptor phase of EM-SPME and once the extraction was completed, the pencil lead was directly introduced into the GC injection port. The second method of carboxylate ion pair production was through the immersion of pencil lead in a TBA solution following accomplishment of EM-SPME. Comparison of the attained results demonstrated that good derivatization performance was gained by immersion of the SPME fiber in TBA solution. The main point of the injection port derivatization technique is that both analyte and derivatization reagent should simultaneously exist in the injection port during the process. When TBA was added into the acceptor solution, its adsorption on the surface of the pencil lead was difficult since both of them have positive charges. The charge of TBA is positive in the neutral acceptor phase while the positive charges of pencil lead is attributed to the connection of the fiber to the positive pole of the power supply during the extraction procedure. Therefore, the derivatization efficiency decreases due to reduction of the TBA concentration in the injection port. However, immersion of the fiber into a concentrated TBA solution occurred after completion of the extraction and disconnection of the electrical field. In this way, a thin stagnant layer of TBA solution was formed on the surface of the fiber by means of the porous nature of the pencil lead. When the fiber together with the TBA solution layer is introduced into the GC injection port, the TBA solvent (methanol) is evaporated and thus a considerable amount of TBA is available. Hence, desorption of both TBA and analytes lead to formation of derivatized analytes. The TBA concentration was assessed thereafter, to find the best derivatization conditions. The concentration of the derivatization reagent in methanol was varied in the range of 30−400 mM. After EM-SPME of the analytes, the carbonaceous fiber was immersed in 30 μL of TBA solution and instantly inserted into the GC injection port. It was noticed that
■
RESULTS AND DISCUSSION In order to figure out how the impressive parameters affected the final signals of the analytes, a number of experiments were designed. Two different influential parameters in the final chromatographic signals included EM-SPME related variables and those linked to GC analysis (i.e., derivatization process and desorption conditions). GC instrument associated parameters were optimized in terms of derivatization procedure, TBA concentration, injection port temperature, carrier gas flow rate, and desorption time. EM-SPME affected parameters consisted 3136
dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
Figure 2. (A) Optimization of membrane composition, (B) response surface of the square root of the normalized peak area vs pH of donor phase and pH of acceptor phase, (C) response surface of the square root of the normalized peak area vs pH of the acceptor phase and applied voltage, and (D) response surface of the square root of the normalized peak area vs extraction time and applied voltage.
extraction time are directly affected by the composition of organic liquid membrane. Analytes should have a proper affinity for SLM so that adequate selectivity and favorable extraction efficiency could be acquired. In fact, SLM provides the major electrical resistance for the system and the range of applied voltage could be extended via raising the electrical resistance of SLM. Accordingly, system stability and extractability may improve utilizing organic solvents with high electrical resistance such as SLM. Nevertheless, the organic phase should possess satisfactory conductivity to enable ion transportation across the phases. Regarding all of these factors, 1-hexanol, 1-octanol, 1undecanol, and 1-dodecanol were tested as the organic liquid membrane. It was proved that long-chain alcohols are the most desirable organic solvents for extraction of acidic compounds.6,28 Among the examined alcohols, 1-hexanol has the least electrical resistance, and a system operated with 1-hexanol, such as the SLM, underwent instability problems owing to increasing the electric current (Figure 2A). Therefore, 1hexanol offered nonrepeatable results, and the chromatographic signals substantially declined, which was probably caused by the enhancement of electrochemical reactions due to increasing the electric current of the system. On the other hand, 1-dodecanol could not provide a continuous electric field in the entire system, since it possesses a low electrical conductivity; for this reason, the extractability diminished using 1-dodecanol as the SLM. 1-Undecanol was capable of effectively extracting the 2,4D, although it was not a suitable solvent for MCPA. Ultimately, 1-octanol was chosen as the SLM due its ability to give the best results for both of the analytes. Optimization of Extraction Time, Applied Voltage, and pHs of the Sample Solution and Acceptor Phase Using Central Composite Design. The experiments for optimization of the rest of the EM-SPME affected parameters were modeled through face-centered central composite design
the peak areas of both CPA herbicides initially increased via enhancing the concentration of the derivatization reagent up to 200 mM and then reached a plateau with a further rise in TBA concentration. Therefore, 200 mM TBA in methanol was selected as the derivatization solution. Another effective parameter on the efficiency of the analyte desorption as well as the injection port derivatization process is the injection port temperature that was considered in the range of 250−300 °C. The peak areas of the derivatized analytes were enhanced by raising the temperature to 290 °C. Since no changes in the derivatization yields were observed at higher temperatures, 290 °C was chosen to reach the maximum desorption and derivatization efficiencies and also prevent the entrance of the pencil lead from interfering with compounds going into the GC injection port. The carrier gas flow rate can influence the derivatization efficiency through altering the delay time of the analytes in the injection port of the GC apparatus. The flow rate of the carrier gas was scrutinized in the range of 1−4 mL min−1. The maximum peak areas for butyl esters of the herbicides were achieved at a flow rate of 2 mL min−1. Furthermore, desorption time was explored in the range of 0.5−5.0 min while the desorption temperature was 290 °C. The efficiency of thermal desorption improved as the desorption time was increased from 0.5 to 2.0 min. The chromatographic signals slightly increased after 2.0 min, but desorption of some interferences from the pencil lead fiber resulted in crowded chromatograms. Consequently, the desorption time of 2.0 min was selected. EM-SPME Affecting Parameters. Composition of Organic Liquid Membrane. One of the most significant parameters, which could influence the extractability of analytes in EM-SPME technique, is the SLM composition. Extraction selectivity, range of applied voltage, extraction efficiency, and 3137
dx.doi.org/10.1021/jf500017r | J. Agric. Food Chem. 2014, 62, 3134−3142
Journal of Agricultural and Food Chemistry
Article
Table 1. Experimental Factors, Levels, and Analysis of Variance (ANOVA) for the Response Surface Quadratic Model level factor extraction time (min) applied voltage (V) pH of donor phase pH of acceptor phase source sum of squares block model A: t B: V C: pHd D: pHa AB AC AD BC BD CD ̂ A2 ̂ B2 2̂ C ̂ D2 residual lack of fit pure error cor total R2 adj R2
61.46 204.54 122.78 14.64 2.19 12.12 0.9 1.13 1.1 0.051 1.14 × 10−5 0.96 0.28 5.32 5.03 11.34 6.66 6.3 0.35 272.66 0.9685 0.9317
symbol
(−1)
t V pHd pHa
5 20 7 1
d.f.
mean square
1 14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 12 10 2 27
61.46 14.61 122.78 14.64 2.19 12.12 0.9 1.13 1.1 0.051 1.14 × 10−5 0.96 0.28 5.32 5.03 11.34 0.55 0.63 0.18
(0)
F-value
12:30 50 10 7 p-value
(+1) 20 80 13 13
26.34 221.34 26.4 3.95 21.85 1.62 2.03 1.98 0.093 2.05 × 10−5 1.74 0.5 9.6 9.07 20.44
