An alternative supercritical fluid extraction system for aqueous matrices and its application in pesticides residue analysis

Abstract

Supercritical fluid extraction (SFE) is a rapid and convenient method for the isolation of organic compounds from environmental samples. This paper describes a supercritical carbon dioxide (CO₂) extraction system that uses a newly designed extraction cell to recover organic compounds from an aqueous matrix. Analysis of the extracts by gas chromatography‐electron capture detector (GC‐ECD) indicated that the herbicide trifluralin (2,6‐dinitro‐N,N‐dipropyl‐4‐trifluoromethylaniline) could be quantitatively extracted by using the SFE system proposed with small amounts of sample. The percentage of recovery obtained with the SFE system described was twice as high as the result obtained using a conventional solid‐phase extraction technique. Extraction by SFE was completed in a short period of time using a simple and low‐cost home‐made system that did not require the use of organic solvents.     

Evaluation of a multiresidue method for pesticides in cereals using supercritical fluid extraction and gas chromatographic detection

Supercritical fluid extraction (SFE) was evaluated to be applied for residue analysis of 22 gas chromatography/electron capture detector-nitrogen phosphorus detector (GC/ECD-NPD) amenable pesticides in rice, wild rice and wheat. Samples were extracted with supercritical carbon dioxide at 200 atm pressure and 50°C temperature, using methanol as a static modifier. Mean recoveries obtained with the proposed SFE method at two spiking levels with four replicates per level are compared with those obtained with an ethyl acetate-based solvent extraction/gel permeation chromatography (GPC) clean up method. Both methods gave consistent high recoveries for almost all the pesticides from all the commodities with overall mean recoveries higher than 70% with relative standard deviations lower than 20%. Remarkable exceptions were captafol and dimethoate, for which low and/or non-reproducible recoveries were obtained with the SFE method. Residue levels determined with both methods in nine different incurred samples of wheat, containing some of the studied pesticides, were very similar, but, in all cases, slightly higher levels were determined with the SFE method.     

Comparison of supercritical fluid extraction and Soxhlet extraction for the determination of PCBs in seaweed samples

The efficiency of supercritical fluid extraction for the determination of 12 polychlorinated biphenyls from algae samples is compared to Soxhlet extraction. Analytical detection limits for the individual congeners ranged from 0.62 microgl(-1) to 19 microgl(-1). Recovery was tested for both methods using standard addition procedure. At maximum spike level of concentration, the mean recoveries were not significantly different (P>0.05) of all PCBs studied, with the exception of PCBs 28, 52, 77 and 169. Method precision for Soxhlet extraction (< or =3.9%) was slightly better than for SFE (< or =9.2%). Although both methods yield comparable results, SFE offers the advantage of detecting all PCBs studied at lower concentrations, reducing extraction time, and reducing the amount of solvents needed. The optimized methods were applied to the analysis of three real seaweed samples, except for PCB101 the concentrations of all PCBs were low or below the detection limits. The levels of PCB101 found in sample 1 were 6.6+/-0.54 ng g(-1) d.w., in sample 2 the levels were 8.2+/-0.86 ng g(-1) d.w. and in sample 3 they were 7.7+/-0.08 ng g(-1) d.w.     

Solid-phase extraction for multi-residue analysis of pesticides in honey

A fast and simple multi-residue method for the analysis of 15 organophosphorus (OP), 17 organochlorine (OC), 8 pyrethroids (PYR), 12 N-methyl-carbamate (NMC) pesticide residues and bromopropylate in honey is presented. Ready–to–use EXtrelut®NT 20 column, eluted with dichloromethane, was used to extract the pesticide residues from the aqueous-acetone honey sample, obtaining a clean extract directly analyzable. Determination was carried out by gas chromatography (GC) coupled with flame photometric detector (FPD) for OP compounds and by GC coupled with mass spectrometry detector (MSD) for OC and PYR pesticides and bromopropylate. The NMC pesticides were analysed by liquid chromatography-double derivatization coupled with spectrofluorimetric detector (LC/DD/Fl). This method allows the determination of the 53 pesticide residues at low concentrations (0.0005–0.074 mg/kg) and can be used to assess the compliance with the Maximum Residues Levels (MRLs) set by the European Union. The performance of the method was evaluated and specificity, linearity, recovery, repeatability, reproducibility, limit of quantification (LOQ) and limit of detection (LOD) were determined. A good linearity (r²? 0.99) was found in the range 0.0005–0.074 mg/kg for the majority of the compounds studied. Most of the pesticides had recoveries in the range 70–103 % and values of relative standard deviation (RSD) < 20 for repeatability and reproducibility, showing good accuracy and precision of the method. Aldicarb partially degraded in aldicarb sulphoxide during the analytical procedure, giving anomalous values. The LOQ for all pesticides investigated was from 0.0005 to 0.025 mg/kg while the LOD ranged from 0.0002 to 0.008 mg/kg.     

Solid-phase extraction for multi-residue analysis of pesticides in honey

A fast and simple multi-residue method for the analysis of 15 organophosphorus (OP), 17 organochlorine (OC), 8 pyrethroids (PYR), 12 N-methyl-carbamate (NMC) pesticide residues and bromopropylate in honey is presented. Ready-to-use EXtrelutNT 20 column, eluted with dichloromethane, was used to extract the pesticide residues from the aqueous-acetone honey sample, obtaining a clean extract directly analyzable. Determination was carried out by gas chromatography (GC) coupled with flame photometric detector (FPD) for OP compounds and by GC coupled with mass spectrometry detector (MSD) for OC and PYR pesticides and bromopropylate. The NMC pesticides were analysed by liquid chromatography-double derivatization coupled with spectrofluorimetric detector (LC/DD/Fl). This method allows the determination of the 53 pesticide residues at low concentrations (0.0005-0.074 mg/kg) and can be used to assess the compliance with the Maximum Residues Levels (MRLs) set by the European Union. The performance of the method was evaluated and specificity, linearity, recovery, repeatability, reproducibility, limit of quantification (LOQ) and limit of detection (LOD) were determined. A good linearity (r(2)≥ 0.99) was found in the range 0.0005-0.074 mg/kg for the majority of the compounds studied. Most of the pesticides had recoveries in the range 70-103 % and values of relative standard deviation (RSD) < 20 for repeatability and reproducibility, showing good accuracy and precision of the method. Aldicarb partially degraded in aldicarb sulphoxide during the analytical procedure, giving anomalous values. The LOQ for all pesticides investigated was from 0.0005 to 0.025 mg/kg while the LOD ranged from 0.0002 to 0.008 mg/kg.     

Assessing supercritical fluid extraction for the analysis of fipronil, kresoxim-methyl, acrinathrin, and pyridaben residues in melon

An analytical method, using supercritical fluid extraction (SFE), was assessed for residue analysis of the insecticides fipronil, acrinathrin and pyridaben and the fungicide kresoxim-methyl in melons. Extraction conditions were optimized on glass beads, as inert material, evaluating different pressures, temperatures, and CO(2) volumes. Also, the extractability of pesticides from hydromatrix and anhydrous magnesium sulfate was evaluated. Both adsorbents were evaluated for the preparation of SFE-spiked melon samples. Extractions of fortified samples were done using 20 mL of CO(2), 300 atm, and 50 degrees C. Different extraction conditions, including the addition of water and methanol as a static modifier, were used to assess the extractability of these pesticides, and the addition of methanol was demonstrated to improve the extraction of all the studied pesticides.     

Assessing supercritical fluid extraction for the analysis of fipronil,kresoxim-methyl,acrinathrin, and pyridaben residues in melon

An analytical method, using supercritical fluid extraction (SFE), was assessed for residue analysis of the insecticides fipronil, acrinathrin and pyridaben and the fungicide kresoxim-methyl in melons. Extraction conditions were optimized on glass beads, as inert material, evaluating different pressures, temperatures, and CO₂ volumes. Also, the extractability of pesticides from hydromatrix and anhydrous magnesium sulfate was evaluated. Both adsorbents were evaluated for the preparation of SFE-spiked melon samples. Extractions of fortified samples were done using 20 mL of CO₂, 300 atm, and 50°C. Different extraction conditions, including the addition of water and methanol as a static modifier, were used to assess the extractability of these pesticides, and the addition of methanol was demonstrated to improve the extraction of all the studied pesticides.     

Using supercritical fluid extraction to measure the desorption and bioaccessibility of phenanthrene in soils

The aim of this paper was to measure the changing desorbable fraction and bioaccessibility of phenanthrene in two different soils with increasing soil-phenanthrene contact time using supercritical fluid extractions (SFE). Both soils were spiked with 100 mg kg⁻¹ phenanthrene and aged for 28 d. Desorption profiles were measured every 7 d using selective SFE conditions and the results were compared to ¹⁴C-phenanthrene mineralisation assays. Selective SFE showed significant differences in the rates and extents of desorption in the two soils, likely to be due to different organic matter composition. Post-extraction fitting of data yielded consistent SFE extraction times within ageing soils for bioaccessibility prediction.     

Selective supercritical fluid extraction to identify aged sediment-bound PCBs available for uptake by eel

A naturally contaminated sediment was partially extracted with selective supercritical fluid extraction (SFE) to remove a fraction of supposedly bioavailable PCBs from the matrix. Eels (Anguilla anguilla) were cultured in systems with untreated and pre-extracted sediment, respectively, and it was shown that the SFE treatment selectively removed bioavailable PCBs from the sediment, since relative biota-to-sediment accumulation factors (BSAFs) for the eight studied PCB congeners were much lower in the pre-extracted than in the untreated system at the end of the study. Relative BSAF values decreased with about the same relative amount for all eight congeners, independent of the degree of chlorination and the initial concentration in the sediment. The results demonstrate the ability of SFE to selectively remove sediment-bound PCBs that are available for uptake by the eels, thus demonstrating the feasibility of using selective SFE to estimate bioavailability of PCBs in sediments.     

Assessing solvent derivatization and carbon dioxide supercritical fluid simultaneous extraction/derivatization of cyprodinil

Derivatization of cyprodinil with different reagents and solvents has been evaluated to improve the GC/MS characterization of this fungicide. After assessing some preliminary acylation and silylation reactions, derivatization with anhydrous heptafluorobutyric anhydride (HFBA) was selected as the best derivatization option for cyprodinil. The HFBA-cyprodinil derivative was clearly identified and characterized by GC/MS (ion-trap). The spectrum of the HFBA derivative of cyprodinil was characterized by the base peak, 252 m/z ion, and two other ions with relative abundances of 5% (224 m/z ion) and 4% (420 m/z molecular ion). Conversion rates in the range of 83–92% were obtained when 0.1–1 μg cyprodinil were derivatized in vial without solvent at 25ºC temperature for 120 min, with 5 μL HFBA and 5 μL pyridine. Simultaneous extraction-derivatization of cyprodinil in supercritical carbon dioxide was only achieved when no modifier was present, but conversion/recovery rates obtained in the replicate experiments carried out with 15 mL supercritical carbon dioxide at 50°C and 200 atm (n = 5), 300 atm (n = 7), and 400 atm (n = 5) were no reproducible (RSD > 50%) and ranged between 10% and 45% (related to the signal obtained for derivatization in vial).     

Multiresidue analysis and monitoring of pesticides in rice by pressurized liquid extraction

The aim of this study was to assess the performance of a method of analyzing pesticides in rice by using pressurized liquid extraction (PLE) and to perform a preliminary monitoring by using that method. The instrumental quantification limit, instrumental detection limit, method quantification limit, and method detection limit were determined. PLE temperature was also optimized for 6 target pesticides. Mean recoveries of spiked rice with target pesticides (4 ng/g and 40 ng/g) were 83%–109% with the repeatability of the analysis, represented as relative standard deviations, ranged from 1.3% to 11% (n = 5) for PLE at 130°C. These results were satisfactory according to the method of positive list in Japan. In a preliminary analysis of 10 target pesticides in 54 commercial rice samples, ferimzone was detected in only one unpolished rice sample.     

Measuring bioavailability of polychlorinated biphenyls in soil to earthworms using selective supercritical fluid extraction

If the release mechanisms during selective chemical extraction of persistent organic pollutants (POP) mimic release mechanisms in natural systems during biological uptake, then a selective non-exhaustive extraction could give a quantitative measure of the bioavailable POP fraction. Supercritical fluid extraction (SFE) is suggested as a possible technique to estimate the amount of bioavailable polychlorinated biphenyls (PCBs) at contaminated sites and hence serve as a new tool in risk assessment. The uptake of PCBs by earthworm (Eisenia foetida) was investigated. PCB contaminated soil was pre-extracted with selective non-exhaustive SFE (50 degrees C, 350 bar, 1h), which removed on average 70% of the individual PCBs. Earthworms were placed in this pre-extracted soil, as well as in untreated soil. After 10 days, the PCB uptake by earthworms in the two systems was compared. The bioaccumulation factor (BAF) was 83% lower in the pre-extracted system than in the untreated system, demonstrating that SFE extracts primarily bioavailable contaminants. From the data, the bioavailable fraction could also be calculated to be 75%, which is very close to the 70% removed by SFE under the applied conditions. This suggests that the chemical methodology is capable of measuring the bioavailable fraction very accurately in this system.     

Evaluation of pressurized liquid extraction for the analysis of four pesticides in unpolished rice

The aim of to evaluate efficiency of this study was extraction pressurized liquid extraction (PLE) for the analysis of four pesticides, fthalide, etofenprox, fenitrothion, and isoprothiolane, in unpolished rice by comparing with homogenization as a reference technique. The concentrations of four selected pesticides obtained by PLE with acetonitrile at 130 degrees C for 10 min x 2 cycles were comparable to those by homogenization with water-soaking. The repeatability of the analysis, represented as relative standard deviations (RSDs), were 1.4-3.6% (n = 3) for PLE at 130 degrees C and 1.2-3.8% (n = 3) for homogenization with water-soaking. Recovery yields of surrogates were 75-88% and 87-109% for PLE at 130 degrees C and homogenization with water-soaking, respectively, and these were satisfactory according to the method of positive list. This study suggested that PLE can be applied for the analysis of selected four pesticides in unpolished rice as well as homogenization with water-soaking.     

Evaluation of pressurized liquid extraction for the analysis of four pesticides in unpolished rice

The aim of to evaluate efficiency of this study was extraction pressurized liquid extraction (PLE) for the analysis of four pesticides, fthalide, etofenprox, fenitrothion, and isoprothiolane, in unpolished rice by comparing with homogenization as a reference technique. The concentrations of four selected pesticides obtained by PLE with acetonitrile at 130°C for 10 min × 2 cycles were comparable to those by homogenization with water-soaking. The repeatability of the analysis, represented as relative standard deviations (RSDs), were 1.4–3.6% (n = 3) for PLE at 130°C and 1.2–3.8% (n = 3) for homogenization with water-soaking. Recovery yields of surrogates were 75–88% and 87–109% for PLE at 130°C and homogenization with water-soaking, respectively, and these were satisfactory according to the method of positive list. This study suggested that PLE can be applied for the analysis of selected four pesticides in unpolished rice as well as homogenization with water-soaking.     

Simultaneous analysis of 50 pesticides in water samples by solid phase extraction and GC-MS

A multiresidue analysis of 50 pesticides has been developed as a rapid screening method for organic contaminants in water samples. Solid phase extraction was accomplished with two adsorbents in a single column. Simultaneous analysis was performed with gas chromatography-mass spectrometry (GC-MS). The limits of detection for each pesticide are at sub-ppb levels. This method was used to check contamination of water from the Adige river (Northern Italy).     

Multicriteria optimisation of a simultaneous supercritical fluid extraction and clean-up procedure for the determination of persistent organohalogenated pollutants in aquaculture samples

A useful tool based on a single-step extraction and clean-up procedure for the determination of 15 organohalogenated pollutants (including brominated flame retardants) in aquaculture samples, using aluminium oxide basic and acidic silica gel in the supercritical extraction cell followed by gas chromatography with electron capture detection or mass spectrometry has been developed. This effective clean-up step ensures a minimum of chromatographic difficulties related to complex matrix components such as aquaculture feed. The extraction procedure has been screened by a fractional factorial design for the preliminary statistically significant parameters. The factors selected were extraction temperature, pressure, static extraction time, dynamic extraction time and carbon dioxide flow rate. The Doehlert design, followed by a multicriteria decision-making strategy, was then performed in order to determine the optimum conditions for the two most significant factors: pressure (165 bar) and dynamic extraction time (27 min). Under optimal conditions, the procedure developed with GC-MS/MS provides an excellent linearity, detection (0.01-0.2 ng g(-1)) and quantification limits (0.05-0.8 ng g(-1)) for most of the analytes investigated. The feasibility of the proposed supercritical fluid extraction method was validated by analysing two reference materials and fish feed and shellfish samples with satisfactory results.     

Investigating differential binding of polychlorinated dibenzo-p-dioxins/dibenzofurans in soil and soil components using selective supercritical fluid extraction

Supercritical fluid extraction (SFE) with pure carbon dioxide was performed at increasingly strong conditions to investigate differential binding of polychlorinated dibenzo-p-dioxins/dibenzofurans (PCDD/Fs) in two impacted soils, in their sieved size fractions, and in small (a few mg) samples of industry-related waste products separated from impacted soil. The binding strengths of PCDD/Fs were shown to be different in the two soils, and in their different soil particle size fractions. As might be expected based on surface area considerations, one soil showed the strongest binding in the smallest (<5 μm) sieved fraction. However, the other soil showed the strongest binding in the larger sized fractions, possibly indicating that process-related particles could be controlling PCDD/F binding. Selective SFE of various types of particles including black carbon and charcoal (separated from soil), and from a suspected process anode residue did show different PCDD/F binding behavior ranging from quite weak binding (charcoal) to very strong binding (anode particles). Shifts to the stronger SFE fractions in the soils after activated carbon treatment agreed well with the decreases previously found in the uptake of PCDD/Fs by earthworms, as well as decreases in their freely-dissolved aqueous concentrations in soil/water slurries. These results show that, as previously demonstrated for PAHs and PCBs, selective SFE can be a useful tool to investigate differences in PCDD/F binding behaviors in impacted soils and sediments and their component parts, as well as a rapid tool for estimating the effectiveness of activated carbon treatments on decreasing the bioavailability of PCDD/Fs in soils and sediments.     

Study of the distribution of the polychlorinated biphenyls in the milk fat globule by supercritical fluid extraction

The distribution of polychlorinated biphenyls (PCBs) in the milk fat globule has been studied by sequential extraction of four different lipidic fractions from powdered full-fat milk with supercritical carbon dioxide. Extractions were carried out in the dynamic mode in the pressure range 13.6-23.3 MPa at a temperature of 50 degrees C. The levels of PCBs and short-chain triglycerides (SCT), medium-chain triglycerides (MCT) and long-chain triglycerides (LCT), as well as the cholesterol in these four supercritical fluid extraction (SFE) fractions were determined. Extracts obtained at lower pressures were found to be enriched in PCBs, SCT, MCT and cholesterol, while the concentrations of all these analytes decreased for subsequent extractions. Satisfactory quadratic correlations were found between the PCB and SCT and MCT levels (r2 in the range 0.9-0.999), and between the PCB and cholesterol levels (r2 in the range 0.8-0.999, except for PCB 105) determined in the SFE fractions. These results suggested a similar distribution of the PCBs and the cholesterol in the milk fat globule. This conclusion was also supported by the comparison of the PCBs and the cholesterol chemical structures as well as by the total PCB levels determined by SFE to those obtained by using different extraction methods. The classical requirement for the analysis of this kind of lipophilic pollutants of an exhaustive extraction of the lipids of the matrix to ensure their quantitative recovery has been revised.     

Selective supercritical fluid extraction as a tool for determining the PCB fraction accessible for uptake by chironomid larve in a limnic sediment

Selective supercritical fluid extraction (SFE) at 40 degrees C, 400 bar and 60 min was used to selectively remove bioavailable PCBs from a naturally contaminated limnic sediment. The extraction decreased the sediment concentrations of the nine studied PCBs by 54% (on average). Chironomid larvae were thereafter cultured in pre-extracted sediment as well as in untreated sediment to study the differences in uptake of PCBs in the two cultures. While the prevailing equilibrium partitioning (EqP) theory predicts a 54% decrease in PCB uptake by the chironomids in the pre-extracted sediment, with a maintained biota-to-sediment accumulation factor (BSAF), a decrease in PCB uptake by 91% was observed. For all investigated PCBs the BSAFs were on average 81% lower in the cultures with pre-extracted sediment than in the untreated systems. The data allowed for a calculation of the bioavailable fraction, which was estimated to 60%. This is very close to the 54% removed by selective SFE, demonstrating the possibility of using SFE as a tool to determine the bioavailable PCB fraction in a polluted sediment.     

Multiresidue analysis of organophosphate and pyrethroid pesticides in duplicate-diet solid food by pressurized liquid extraction

An analytical method was developed for determining organophosphate pesticides (OPP) and pyrethroid pesticides (PYR) in duplicate-diet solid food. The method consisted of pressurized liquid extraction (PLE) with dichloromethane followed by cleanup with gel permeation and solid phase extraction columns and gas chromatography/mass spectrometry (GC/MS) analysis. Quantitative recoveries (73-117 %) of the target pesticides were obtained for spiked duplicate-diet food samples. The percent standard deviation (% RSD) of replicate food samples was within ± 20 %. Another method was developed for determining a common OPP metabolite, 3, 5, 6-trichloro-2-pyridinol (TCP) in duplicate-diet food. The method consisted of a PLE with methanol followed by liquid-liquid partitioning, derivatization, and GC/MS analysis. Recoveries of TCP ranged from 83 to 101 % for spiked duplicate-diet food samples. The % RSD of replicate food samples was within ± 15 %. The results confirmed that these methods are reliable and robust, and that they can be used in routine analysis. In addition, a storage stability study for a common OPP, chlorpyrifos (CPF), in solid food samples was performed. The fortified (15)N-(13)C-labeled CPF was stable over 16 mo storage at -20° C in the dark. The developed analytical methods were successfully applied to 278 duplicate-diet food samples from preschool children, demonstrating that these methods are robust and suitable for routine analysis in future exposure monitoring studies.