The present protocol describes the utilization of ammonium formate for phase partitioning in QuEChERS, together with gas chromatography-mass spectrometry, to successfully determine organochlorine pesticide residues in a soil sample.
Currently, the QuEChERS method represents the most widely used sample preparation protocol worldwide for analyzing pesticide residues in a broad variety of matrices both in official and non-official laboratories. The QuEChERS method using ammonium formate has previously proven to be advantageous compared to the original and the two official versions. On the one hand, the simple addition of 0.5 g of ammonium formate per gram of sample is sufficient to induce phase separation and achieve good analytical performance. On the other hand, ammonium formate reduces the need for maintenance in routine analyses. Here, a modified QuEChERS method using ammonium formate was applied for the simultaneous analysis of organochlorine pesticide (OCP) residues in agricultural soil. Specifically, 10 g of the sample was hydrated with 10 mL of water and then extracted with 10 mL of acetonitrile. Next, phase separation was carried out using 5 g of ammonium formate. After centrifugation, the supernatant was subjected to a dispersive solid-phase extraction clean-up step with anhydrous magnesium sulfate, primary-secondary amine, and octadecylsilane. Gas chromatography-mass spectrometry was used as the analytical technique. The QuEChERS method using ammonium formate is demonstrated as a successful alternative for extracting OCP residues from a soil sample.
The need to increase food production has led to the intensive and widespread use of pesticides worldwide over the last few decades. Pesticides are applied to the crops to protect them from pests and increase crop yields, but their residues usually end up in the soil environment, especially in agricultural areas1. Furthermore, some pesticides, such as organochlorine pesticides (OCPs), have a very stable structure, so their residues do not decompose easily and persist in the soil for a long time2. Generally, the soil has a high capacity to accumulate pesticide residues, especially when it has a high content of organic matter3. As a result, the soil is one of the environmental compartments most contaminated by pesticide residues. As an example, one of the complete studies to date found that 83% of 317 agricultural soils from across the European Union were contaminated with one or more pesticide residues4.
Soil pollution by pesticide residues may affect non-target species, soil function, and consumer health through the food chain because of the high toxicity of the residues5,6. Consequently, the evaluation of pesticide residues in soils is essential to assess their potential negative effects on the environment and human health, particularly in developing countries due to a lack of strict regulations on the use of pesticides7. This makes pesticide multi-residue analysis increasingly important. However, the rapid and accurate analysis of pesticide residues in soils is a difficult challenge due to the large number of interfering substances, as well as the low concentration level and the diverse physicochemical properties of these analytes4.
Of all the pesticide residue analysis methods, the QuEChERS method has become the quickest, easiest, cheapest, most effective, robust, and safest option8. The QuEChERS method involves two steps. In the first step, a microscale extraction based on partitioning via salting-out between an aqueous and an acetonitrile layer is performed. In the second step, a cleaning process is carried out employing a dispersive solid phase extraction (dSPE); this technique uses small amounts of several combinations of porous sorbents to remove matrix-interfering components and overcomes the disadvantages of conventional SPE9. Hence, the QuEChERS is an environmentally friendly approach with little solvent/chemical going to waste that provides very accurate results and minimizes potential sources of random and systematic errors. In fact, it has been successfully applied for the high-throughput routine analysis of hundreds of pesticides, with strong applicability in almost all types of environmental, agri-food, and biological samples8,10. This work aims to apply and validate a new modification of the QuEChERS method that was previously developed and coupled to GC-MS to analyze OCPs in agricultural soil.
1. Preparation of the stock solutions
NOTE: It is recommended to wear nitrile gloves, a lab coat, and safety glasses during the entire protocol.
2. Sample collection
3. Sample preparation via the modified QuEChERS method using ammonium formate
NOTE: Figure 1 shows a schematic representation of the modified QuEChERS method.
4. Instrumental analysis by GC-MS
5. Data acquisition
The full validation of the analytical method was performed in terms of linearity, matrix effects, recovery, and repeatability.
Matrix-matched calibration curves with spiked blank samples at six concentration levels (5 µg/kg, 10 µg/kg, 50 µg/kg, 100 µg/kg, 200 µg/kg, and 400 µg/kg) were used for the linearity assessment. The determination coefficients (R2) were higher than or equal to 0.99 for all the OCPs. The lowest calibration level (LCL) was set at 5 µg/kg, which meets the maximum permissible limit established at 10 µg/kg for monitoring purposes in food applications11.
The matrix effect assessment was carried out by comparing the slopes of the OCP calibration curves in pure solvent and the matrix-matched calibration curves. The matrix effect was calculated using the following equation12:
Matrix effect (%) = (slope of the matrix-matched calibration curve − slope of the pure solvent-based calibration curve)/(slope of the pure solvent-based calibration curve) × 100.
Figure 2 shows the matrix effect distributions for the OCPs studied by applying a modified QuEChERS method using ammonium formate to soil samples. Positive matrix effect percentages correspond to a signal enhancement, while negative percentages mean that there is signal suppression. Specifically, (1) values ranging between −20% and 20% correspond to a soft matrix effect; (2) values ranging between −20% and -50% or between 20% and 50% correspond to a medium matrix effect; (3) and values higher than 50% or lower than −50% mean that there is a strong matrix effect. As observed, more OCPs suffered soft or medium matrix effects, while fewer OCPs suffered strong matrix effects.
The recovery and repeatability were evaluated by spiking blank samples with pesticides at three concentration levels (10 µg/kg, 50 µg/kg, and 200 µg/kg). Figure 3 shows the overall recovery values and relative standard deviation (RSD) values for all the pesticides and spiking levels (n = 9). As can be observed, the great majority of the OCPs studied presented average recovery percentages in the range of 70%-120%, with RSDs lower than 20%, except heptachlor, endrin, and β-endosulfan, which gave slightly higher average recoveries.
Figure 1: Representation of the modified QuEChERS method using ammonium formate to extract pesticide residues from the soil sample. Please click here to view a larger version of this figure.
Figure 2: Distribution of the matrix effects versus retention times (min) for the 17 OCPs. A soft matrix effect corresponds to values between −20% and 20%; a medium matrix effect corresponds to values ranging between −20% and −50% or between 20% and 50%; a strong matrix effect corresponds to values that are greater than 50% or less than −50%. Please click here to view a larger version of this figure.
Figure 3: Average recoveries for the 17 OCPs after spiking 10 µg/kg, 50 µg/kg, and 200 µg/kg (n = 9) in the soil sample. The number of analytes within the acceptable recovery (70%-120 %) and RSD (<20 %) range are provided, along with those labeled outside of that range. Please click here to view a larger version of this figure.
Analyte | Retention time (min) | Quantifier ion | Qualifier ion 1 | Qualifier ion 2 |
α-BHC | 11.35 | 181 | 219 | 111 |
β-BHC | 11.90 | 181 | 219 | 109 |
Lindane | 12.01 | 181 | 183 | 219 |
δ-BHC | 12.39 | 181 | 219 | 111 |
Heptachlor | 13.24 | 272 | 100 | 274 |
Aldrin | 13.94 | 263 | 66 | 265 |
Heptachlor epoxide | 14.86 | 353 | 355 | 81 |
α-Endosulfan | 15.71 | 241 | 239 | 195 |
4,4'-DDE-d8 (IS) | 16.09 | 254 | 184 | 326 |
4,4'-DDE | 16.12 | 246 | 318 | 248 |
Dieldrin | 16.18 | 79 | 263 | 81 |
Endrin | 16.57 | 263 | 317 | 345 |
β-Endosulfan | 16.73 | 195 | 241 | 159 |
4,4'-DDD | 16.89 | 235 | 237 | 165 |
Endosulfan sulfate | 17.61 | 387 | 227 | 272 |
4,4'-DDT | 17.65 | 235 | 237 | 165 |
Endrin ketone | 18.64 | 317 | 67 | 315 |
Methoxychlor | 18.86 | 227 | 228 | 212 |
Table 1: Retention times (min) and quantification parameters for the GC-MS analysis of the OCPs. Alpha-benzenehexachloride (α-BHC); beta-benzenehexachloride (β-BHC); lindane; delta-benzenehexachloride (δ-BHC); heptachlor; aldrin; heptachlor epoxide; α-endosulfan; 4,4'-dichlorodiphenyldichloroethylene-d8 (4,4'-DDE-d8) (IS); 4,4'-dichlorodiphenyldichloroethylene (4,4'-DDE); dieldrin; endrin; β-endosulfan; 4,4'-dichlorodiphenyldichloroethane (4,4'-DDD); endosulfan sulfate; 4,4'-dichlorodiphenyltrichloroethane (4,4'-DDT); endrin ketone; methoxychlor.
The original9 and the two official versions13,14 of the QuEChERS method use magnesium sulfate together with sodium chloride, acetate, or citrate salts to promote acetonitrile/water mixture separation during extraction. However, these salts tend to be deposited as solids on the surfaces in the mass spectrometry (MS) source, which causes the need for increased maintenance of liquid chromatography (LC)-MS-based methods. In terms of overcoming these disadvantages, González-Curbelo et al.15 reported that the more volatile ammonium formate worked well to induce phase separation and the extraction of pesticide residues for both LC- and GC-tandem mass spectrometry (MS/MS). Subsequent studies have also used 0.5 g of ammonium formate per gram of sample to extract pesticide residues from various complex matrices16,17,18,19. In addition, the use of ammonium formate has been shown to provide lower amounts of co-extracted material20, which justifies its use for GC-MS-based methods. The present study, for the first time, reports this version to analyze pesticide residues in soils21.
The GC analysis of pesticide residues in complex matrices such as soils has some limitations because of the action of the co-extracted matrix components on the instrumental response of the pesticides, which causes inaccurate determination and lower sensitivity22,23. Hence, several improvements have been made to minimize the matrix effect, including optimized clean-up steps21. Nevertheless, the matrix effect still takes place and should be corrected as much as possible. In this sense, matrix-matched calibration has been the main approach used because it is very practical in compensating for the enhancement of the chromatographic signal with respect to that of pure solvents24. Thus, in this study, linearity was evaluated by building the calibration curves in pure acetonitrile and using soil extracts, and R2 values higher than or equal to 0.99 for all the OCPs were obtained using both approaches. However, when both calibration curves were compared, appreciable matrix effects were found in the range of –49% to 191% (Figure 2). Although the number of pesticides that suffered a strong matrix effect was only 3 of 17 (endrin, endrin ketone, and methoxychlor), subsequent studies were carried out using the matrix-matched calibration curves to compensate for the matrix effects to a greater extent.
No maximum residue limits (MRLs) have been established for pesticide residues in soils, but an LCL of 5 µg/kg was set for all the OCPs, which is lower than the very demanding standard MRL set at 10 µg/kg by the international legislation for the analysis of pesticide residues in agri-food products (Regulation 396/2005)11. In addition, the LCL of 5 µg/kg provided a signal-to-noise ratio (S/N) of around 10 for all the OCPs. The high sensitivity of this method is similar to or even better than that obtained in other studies that have also analyzed OCPs in soils using the QuEChERS method followed by GC-MS. For example, in one study, 34 OCPs were analyzed using the official version of the QuEChERS method that uses a citrate buffer, and the limits of quantification (LOQs) were equal to or higher than 7 µg/kg25. In particular, the LOQ values of α-BHC, β-BHC, lindane, and δ-BHC were between 206 µg/kg and 384 µg/kg. In another work, lindane and dieldrin were analyzed using the same version of the QuEChERS method, and LOQ values of 42 µg/kg and 292 µg/kg were obtained, respectively26. Likewise, another research work also determined aldrin and heptachlor using QuEChERS and GC-MS, with LOQ values of 13 and 23 µg/kg, respectively27.
The recovery and reproducibility assessment was developed at three concentration levels (low, medium, and high) in triplicate (n = 9). For this purpose, the overall recovery values were determined by comparing the pesticide peak area/IS (4,4'-DDE-d8) peak area ratios obtained from the soil samples spiked at the beginning of the application of the QuEChERS method using ammonium formate with those of matrix-matched calibration. In all cases, each replicate was injected twice in the same sequence. It should be noted that using an IS, an isotopically labeled standard, allows for compensating for the possible losses of the pesticides that take place during the whole procedure, as well as the matrix effect and/or possible variability in the instrument. According to the results, most pesticides met the acceptability criteria of 70%-120% recovery values with RSD ≤20% at each spiking level28, which demonstrated the effectiveness and repeatability of the method. Nevertheless, the overall recovery values (n = 9) were slightly higher than 120% for heptachlor (122%), endrin (121%), and β-endosulfan (130%), though they were consistent (RSDs <13%). In this sense, considering the overall recovery values at three spiking levels, an acceptability criterion of 30%-140% with RSD values ≤20% has been established28.
In conclusion, the QuEChERS method using ammonium formate coupled with GC-MS can successfully determine OCPs in agricultural soil samples. It was shown in this study that the simple addition of 5 g of ammonium formate to induce phase separation between the water and acetonitrile layers ensured suitable extraction with high recoveries of the selected pesticides. However, the matrix effect continued to take place, so other approaches, such as the addition of analyte protectants, should be studied in subsequent works. In any case, this alternative to the official QuEChERS versions may be used to avoid the undesirable solids deposited in the analytical system due to the use of magnesium and sodium salts, especially in routine LC-MS-based analysis. In the latter case, it would be even more interesting since ammonium formate is an aid for ionization in positive electrospray ionization and may enhance the formation of ammonium adducts instead of sodium adducts.
The authors have nothing to disclose.
I would like to thank Javier Hernández-Borges and Cecilia Ortega-Zamora for their invaluable support. I also want to thank the Universidad EAN and the Universidad de La Laguna.
15 mL disposable glass conical centrifuge tubes | PYREX | 99502-15 | |
2 mL centrifuge tubes | Eppendorf | 30120094 | |
50 mL centrifuge tubes with screw caps | VWR | 21008-169 | |
5977B mass-selective detector | Agilent Technologies | 1617R019 | |
7820A gas chromatography system | Agilent Technologies | 16162016 | |
Acetone | Supelco | 1006582500 | |
Acetonitrile | VWR | 83642320 | |
Ammonium formate | VWR | 21254260 | |
Automatic shaker KS 3000 i control | IKA | 3940000 | |
Balance | Sartorius Lab Instruments Gmbh & Co | ENTRIS224I-1S | |
Bondesil-C18, 40 µm | Agilent Technologies | 12213012 | |
Bondesil-PSA, 40 µm | Agilent Technologies | 12213024 | |
Cyclohexane | VWR | 85385320 | |
EPA TCL pesticides mix | Sigma Aldrich | 48913 | |
Ethyl acetate | Supelco | 1036492500 | |
G4567A automatic sampler | Agilent Technologies | 19490057 | |
HP-5ms Ultra Inert (5%-phenyl)-methylpolysiloxane 30 m x 250 µm x 0.25 µm column | Agilent Technologies | 19091S-433UI | |
Magnesium sulfate monohydrate | Sigma Aldrich | 434183-1KG | |
Mega Star 3.R centrifuge | VWR | 521-1752 | |
Milli-Q gradient A10 | Millipore | RR400Q101 | |
p,p'-DDE-d8 | Dr Ehrenstorfer | DRE-XA12041100AC | |
Pipette tips 2 – 200 µL | BRAND | 732008 | |
Pipette tips 5 mL | BRAND | 702595 | |
Pipette tips 50 – 1000 uL | BRAND | 732012 | |
Pippette Transferpette S variabel 10 – 100 µL | BRAND | 704774 | |
Pippette Transferpette S variabel 100 – 1000 µL | BRAND | 704780 | |
Pippette Transferpette S variabel 20 – 200 µL | BRAND | 704778 | |
Pippette Transferpette S variabel 500 – 5000 µL | BRAND | 704782 | |
Vials with fused-in insert | Sigma Aldrich | 29398-U | |
OCPs | CAS registry number | ||
α-BHC | 319-84-6 | ||
β-BHC | 319-85-7 | ||
Lindane | 58-89-9 | ||
δ-BHC | 319-86-8 | ||
Heptachlor | 76-44-8 | ||
Aldrin | 309-00-2 | ||
Heptachlor epoxide | 1024-57-3 | ||
α-Endosulfan | 959-98-8 | ||
4,4'-DDE-d8 (IS) | 93952-19-3 | ||
4,4'-DDE | 72-55-9 | ||
Dieldrin | 60-57-1 | ||
Endrin | 72-20-8 | ||
β-Endosulfan | 33213-65-9 | ||
4,4'-DDD | 72-54-8 | ||
Endosulfan sulfate | 1031-07-8 | ||
4,4'-DDT | 50-29-3 | ||
Endrin ketone | 53494-70-5 | ||
Methoxychlor | 72-43-5 |