Journal of Chromatography B 1156 (2020) 122296 Contents lists available at ScienceDirect Journal of Chromatography B journal homepage: www.elsevier.com/locate/jchromb Validation of a methodology by LC-MS/MS for the determination of triazine, T triazole and organophosphate pesticide residues in biopurification systems Mario Masís-Moraa, Wilson Beita-Sandía, Javier Rodríguez-Yáñezb, Carlos E. Rodríguez-Rodrígueza,⁎ a Research Center of Environmental Pollution (CICA), University of Costa Rica, 2060 San José, Costa Rica b Laboratorio de Ecología Urbana, Universidad Estatal a Distancia (UNED), 474-2050 San José, Costa Rica A R T I C L E I N F O A B S T R A C T Keywords: Biopurification systems are useful in the management of pesticide residues and provide an option to dispose Biopurification system wastewaters of agricultural origin derived from pesticide application practices. The analysis of pesticide residues Pesticides in the biopurification system biomixture is necessary to determine whether the removal of the target compounds Validation methodology occurs with reliable results. In this study, the pesticide extraction methodology was optimized and validated in a Biomixture biomixture composed of coconut fiber, compost and soil, to determine a total of 43 molecules, distributed among LC-MS/MS triazines (10), triazoles (13) and organophosphates (20) using liquid chromatography coupled to a triple quadrupole mass spectrometer. For the validation, the parameters of linearity, matrix effect, limit of determi- nation (LOD), specificity, selectivity, precision, trueness and robustness in the proposed biomixture were eval- uated. The analyses of those parameters revealed satisfactory results of the method for most of the compounds, with the exception of diclorvos and ciromazine, for which the development of an alternative method is re- commended. Once the extraction methodology was validated, the removal of eight molecules was assayed in a biopurification system used for the simultaneous treatment of a mixture of pesticide commercial formulations. Although most of the compounds were at least partially removed, none of them was eliminated at levels below the LOD. The removal pattern of ametryn, atrazine, chlorpyrifos, malathion and terbutryn was comparable to those obtained in other efficient biomixtures, and the highly recalcitrant triadimenol was eliminated; none- theless, tebuconazole and diazinon were not significantly removed. 1. Introduction point-source pollution by pesticides, hence, reducing its environmental impact. For instance, in situ treatment can reduce pesticide residues The advances in agricultural science (e.g., improved soil and water remaining in knapsack sprayers. However, this requires treatment sys- management practices and the use of agrochemicals, organic fertilizers, tems that are accessible to farmers and easy to operate [6,15–17]. biological control and pesticides) allowed the enhancement of food Bioremediation is regarded as a feasible treatment of wastewaters production. Although the use of pesticides was initially intended to containing high loads of pesticides [15–18]. Particularly, biopurifica- minimize the effect of pests on crops and enhance their productivity, tion systems (BPS) stand out for their low cost and maintenance, easy pesticide application may also cause undesirable effects on human construction and versatility. BPS include biofilters, Phytobac® and health and the environment as they inevitably reach non-target or- biobeds [6,9,11,12,19,20]. These configurations use a biological active ganisms [1–4]. matrix that retains the contaminants and stimulates the rapid de- The presence of pesticides in the environment is worrisome because gradation of the compounds by microbial activity [6,7,9,10,19,21]. The of their toxicity and persistence [3,4]. The effects on humans and other biological matrix is a biomixture that comprises soil, compost or peat organisms are compound-specific and vary depending on toxicity, route and a lignocellulosic material at a 1:1:2 volumetric ratio [6,7,10]. Soil and time of exposure [3,4]. In the environment, aerial fumigations, is the main source of degrading microorganisms; thus, it is desirable to superficial runoff, and infiltrations to groundwater are some common use soils pre-exposed to the target pesticides. Compost or peat is added causes of contamination by pesticides [5–14]. Implementing good for enhancing the adsorption capacity, also helping to control the management practices and adequate treatment systems can mitigate temperature and humidity of the system. The lignocellulosic material is ⁎ Corresponding author. E-mail address: carlos.rodriguezrodriguez@ucr.ac.cr (C.E. Rodríguez-Rodríguez). https://doi.org/10.1016/j.jchromb.2020.122296 Received 27 May 2020; Received in revised form 25 July 2020; Accepted 29 July 2020 Available online 01 August 2020 1570-0232/ © 2020 Elsevier B.V. All rights reserved. M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 a lignin rich source for microorganisms, which favors the growth and in a freezer while the PDS and the calibration standards at < 6 °C. activity of ligninolytic fungi, known for their wide capacity to degrade organic pollutants; in this respect, the use of different components such 2.3. Samples and sample preparation as straw, bagasse, coconut fiber, citrus peel, branches, olive leaves, wood curl, paper, rice pellets, among others [6,7,9,10,17,22], has been The biomixture (pH 6.4; C 4.83%; N 0.32%; C/N 15.2; P 0.22%; Ca applied depending on geographical availability. 0.48%; Mg 0.71%; K 0.19%; S 0.07%; Fe 31 192 mg/kg; Zn 91 mg/kg; BPS are effective degrading molecules with different action modes Mn 521 mg/kg; B 66 mg/kg; EC 0.6 mS/cm) employed consisted of like carbamates [8,23,24], organophosphates [25–28] and triazines coconut fiber, compost and soil (45:12:43, volumetric ratio). This bio- [2,27,29–32]. Pesticides containing such molecules are not applied si- mixture was previously optimized for the removal of carbofuran [8]. multaneously, but rather in cycles in each crop. Therefore, pesticide The biomixture samples were fortified with the target pesticides during active ingredients and their co-formulated materials can be treated in the optimization and validation of the method. Pesticides were ex- the BPS throughout the season as they are used in the crops [9,33]. tracted following a QuEChERS modified procedure described elsewhere Thus, the objectives of this study were (i) to develop and validate a LC- [8]. Carbofuran-d3 and linuron-d6 were used as surrogate and internal MS/MS multiresidue methodology for the determination of more than standard, respectively. Quality controls included blank samples (pesti- 40 pesticides in a conventional biomixture, and (ii) to evaluate the ef- cide-free biomixture), blanks for calibration curve (pesticide-free bio- ficiency of a BPS during the treatment of wastewater containing com- mixture without surrogate or internal standard; extract used for the mercial formulations of diverse pesticides, applying the validated calibration curve in the matrix) and solvent reference (procedure re- methodology. agents without sample). 2. Materials and methods 2.4. Chromatographic conditions 2.1. Chemicals and reagents LC-MS/MS analyses were carried out using an Agilent 1290 Infinity II LC System (Santa Clara, CA, U.S.) Ultra-high performance liquid The analytical standards anilophos (98.9%), azinphos-methyl chromatography (UHPLC) coupled to an Agilent 6460 triple quadrupole (98.8%), cadusafos (97.2%), chlorpyrifos (99.5%), dichlorvos (98.4%), mass spectrometer. Chromatographic separation was done at 40 °C in- dimethoate (99.5%), edifenphos (98.5%), ethoprophos (98.8%), fena- jecting 6 μL of the sample (2 μL loop) in a Poroshell 120 EC-C18 column miphos (99.0%), phoxim (99.4%), heptenophos (98.6%), isazofos (100 mm × 2.1 mm i.d., particle size 2.7 µm) and using a binary mobile (99.2%), isofenphos (99.5%), malathion (99.5%), methamidophos phase consisting of acidified water (formic acid 0.1% v/v, solvent A) (99.5%), monocrotophos (99.5%), pirimiphos-methyl (99.5%), cyro- and acidified methanol (formic acid 0.1% v/v, solvent B) at a flow rate mazine (99.5%), prometon (99.5%), prometryn (99.5%), simetryn of 0.3 mL/min. The conditions were as follows: 30% of solvent B for (99.5%), terbutryn (98.1%), bitertanol (99.5%), cyproconazole 3 min, 15 min linear gradient to 100% solvent B, 4 min at 100% solvent (99.5%), epoxiconazole (99.5%), fenbuconazole (99.5%), flusilazole B, 0.1 min gradient back to 30% of solvent B, and 5 min at initial (98.6%), hexaconazole (99.3%), myclobutanil (98.0%), tebuconazole conditions. The mass spectrometer used a jet stream (electrospray) io- (98.0%), triadimefon (99.5%), triadimenol (98.7%) were purchased nization source operating at a gas temperature of 300 °C; gas flow 7 L/ from ChemService (Penssylvania, U.S.). Standards acephate (99.0%), min, nebulizer 45 psi; sheath gas temperature 250 °C; sheath gas flow coumaphos (99.0%), fenthion (99.0%), triazophos (80.0%), amethryn 11 L/min. The other conditions were capillary voltage 3500 V; nozzle (98.0%), atrazine (99.0%), cyanazine (98.5%), simazine (98.0%), ter- voltage 500 V; heater MS1 and MS2 100 °C. Data acquisition was per- buthylazine (98.5%), difenoconazole (98.7%), paclobutrazole (98.5%), formed using the MassHunter software (Santa Clara, CA, U.S.). propiconazole (99.0%), carbofuran-d3 (98.0%) and linuron-d6 (98.5%) were acquired from Dr. Ehrenstorfer (Augsburg, Germany). 2.5. Optimization of the transitions Commercial formulations of atrazine (Atranex®, 90% w/w), ame- tryn (Agromart®, 50% w/v), chlorpyrifos (Solver™ 48% w/v), diazinon Each molecule was injected individually into the LC-MS/MS system (Zinoncoop 60 EC, 60% w/v), malathion (Bioquim malathion, 5% w/ to optimize the fragmentor voltage and the collision cell energy for all w), tebuconazole/triadimenol (Silvacur® Combi 30 EC, 22.5% and 7.5% the transitions. The optimization was done in five acquisition modes w/v, respectively) and terbutryn (Terbutrex®, 50% w/v) were acquired including (i) MS2 Scan to find the precursor ions (in positive and ne- at local markets. gative electrospray ionization), (ii) Product ion to find the optimized Distilled and deionized (DDI) water (< 18 mΩ) was produced in the fragmentor voltages (ranging from 50 to 210 V) and the main frag- laboratory, formic acid (ACS, ISO, Reag. Ph Eur, 98–100%), glacial ments, (iii) multiple reaction monitoring (MRM) for optimizing the acetic acid (ACS, ISO, Reag. Ph Eur 100%), acetonitrile collision cell energies (range of 1 to 45 V) of each fragment; all those (LichroSolve®, > 99.8%) and methanol (LichroSolve®, 99.8%) were methods were done without column; (iv) MRM with column to find the purchased from Merck (Darmstadt, Germany). Anhydrous magnesium retention time, and (v) dynamic MRM (dMRM) to define a specific sulfate (MgSO4, > 99.5%) and sodium acetate trihydrate acquisition time range; this acquisition method was applied with the (CH3COONa·3H2O, > 99.5%) were obtained from Sigma-Aldrich (St. chromatographic gradient conditions. Louis, MO, U.S.), bondesil-PSA (40 μm particle size) from Agilent (Santa Barbara, CA, U.S.), Sepra-C18 from Phenomenex (Torrance, CA, 2.6. Optimization of the analytical method U.S.) and sodium chloride from JT Baker (PA, U.S.). A 23 full-factorial design was used to study the effects of stirring 2.2. Analytical solutions (manual vs. automated), amount of water added to the sample (5 mL vs. 10 mL) and the amount of magnesium sulfate added for cleaning up Stock solutions of individual analytes ranging from 700 to 3800 mg/ (450 mg vs. 900 mg) on the extraction process. Each experiment was L were prepared, depending on their solubility, in methanol or acet- performed in duplicates. The recovery for each evaluated condition was onitrile. Primary dilution standards (PDS) at 10 mg/L were prepared the measured response. using acidified acetonitrile (0.1% formic acid). Calibration standards from 1 to 500 µg/L were prepared in a mixture (1:1 v/v) of acetoni- 2.7. Method validation trile:water acidified with 0.1% formic acid, and in the matrix extract (matrix-matched standards). The stock solutions were stored at −15 °C The validation of the method was conducted following the 2 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 guidelines of the European Commission-Directorate General for Health Table 2 and Food Safety [34]. Analytical parameters evaluated included line- Distribution of the experimental conditions of the study factors for the Youden- arity, limit of determination (LOD), matrix effects, trueness, precision, Steiner test. robustness, application range, specificity and selectivity. Every sample ID Factor E1 E2 E3 E4 E5 E6 E7 E8 was analyzed applying the methodology described in 2.3. Specificity and selectivity were calculated using the transitions, retention time and F1 A A A A a a a a ion ratio for each compound in each analyzed sample. The linearity of F2 B B b B B B b b F3 C c C C C c C c the calibration curves was evaluated with analytical standard solutions F4 D D d D d d D D at ten concentration levels (5, 10, 20, 50, 90, 135, 170, 200, 250 and F5 E e E E e E e E 500 µg/L). The standards were prepared in blank biomixture extract F6 F f f F F f f F and in acidified (0.1% formic acid) water-acetonitrile (1:1). Each cali- F7 G g g G g G G g bration curve was prepared in triplicate. LOD was estimated on a signal RESULT s t u V w x y z to noise ratio (S/N) ratio > 10. LOD was the lowest spiked level with good criteria of trueness and precision. For LOD, seven blank samples flow, with a more homogeneous distribution on the contact surface; the were spiked at 10 µg/kg. biomixture was thoroughly mixed immediately after pesticide appli- Matrix effects (ME) were calculated as the correlation percentage cation. Composite samples were collected at 0, 9, 14, 21, 28, 38, 47 and between the slopes of the solvent calibration curve and the calibration 53 d after pesticide disposal, by withdrawing small portions of bio- curves prepared with blank biomixture extracts, using the expression mixture with basin and shovel from the upper, middle and lower parts ME (%) = [(Slopematrix/Slopesolvent)-1] × 100. of the system in four different points distributed randomly at each Trueness and precision were evaluated spiking the blank biomixture sampling time; subsamples were pooled to collect around 200 g. At least at four different concentrations (10, 50, 150 and 350 µg/kg), with seven 100 g of the biomixture was kept in custody and stored at −20 °C. The replicates for each spiked level (n = 7) and three analysts, for a total of remaining biomixture was reincorporated into the BPS. When possible, 21 samples. Trueness and precision were determined as recovery and removal data for each compound was modeled according to a first order relative standard deviation percentages (RSD), respectively. model (SigmaPlot 14.0) to estimate removal half-life (DT50) values. The robustness was measured applying the Youden-Steiner test [35]. It was applied at two effect levels of seven factors or conditions. To apply the test, seven changing factors in two conditions were em- 3. Results and discussion ployed (Table 1). This study was developed according to the experiment guide in Table 2. Capital letters indicate the experiment was applied 3.1. Optimization of pesticide molecules with the values of the condition “HIGH”, and the lower-case letters with the value of the condition “low”. Physicochemical properties of each compound were used to decide The robustness was calculated by comparing the difference of the on the optimization experiments to be developed (Table S1, Supple- values of each factor, according to the experiment, in relation to the mentary Material) and to define the ionization mode, precursor ion and value calculated as a critical value, which depends on the total standard the solubility of the pesticide in the organic solvent of the methodology. deviation of the experiment. To evaluate selectivity and specificity, five Then, the fragmentor voltages and the collision cell energies were op- blank samples and five spiked samples at 25 µg/kg were prepared, then, timized for the precursor and the product ions for each molecule. The the signals derived from both kinds of samples were compared to dif- optimization results are shown in Table 3. ferentiate between signals provided by the matrix and the analyte, re- Each compound was first injected individually without the analy- spectively. tical column for identifying the best working conditions. MS2 Scan The application range was calculated with an initial concentration acquisition mode was conducted to identify the precursor ion of each + of 2000 μg/kg in samples that were subsequently diluted to inter- molecule at positive or negative electrospray ionization mode (ESI or − + mediate concentrations of the calibration curve, to evaluate the effi- ESI ). All tested molecules showed better results in ESI and worked ciency of the method to achieve good results at high concentrations. with the protonated molecules [MH] +. The use of adducts of sodium, potassium or ammonium was avoided, as they produce a lower sensi- bility in the next optimization steps [36] (Fig. 1). 2.8. Determination of pesticide removal in a functional BPS Then, the main fragments of the precursor ion were determined in Product Ion acquisition mode. Likewise, the fragmentor voltage that A pilot-scale functional BPS conformed by a 204 L plastic barrel gives a signal with a greater intensity was selected. At least two product containing 104 L (~56.2 kg) of the biomixture was employed to assay ions were selected for each molecule (Fig. 1). the performance of the method at field relevant concentrations. A A third acquisition mode, MRM without column, was used to ensure pesticide solution containing commercial formulations of ametryn, the product ions or main fragments were selected at the maximum atrazine, chlorpyrifos, diazinon, malathion, tebuconazole, terbutryn voltage signal (Fig. 1). Finally, the MRM mode with column was applied and triadimenol was disposed into the biomixture. Disposal was per- with the chromatographic column conditions and a proposed mobile formed using a watering can; uniform application allowed a free drip phase gradient, at optimized values for the individual injection of each molecule. Thus, the retention time of each molecule and the evaluation Table 1 of the transitions, as well as possible interference signals for other Experimental design for the determination of the robustness test using the molecules were obtained. Table 3 summarizes the optimized values for Youden-Steiner test. all molecules. ID Factor Factor Condition 1 Condition 2 F1 Water rest time (min) 20 (A) 30 (a) 3.2. Method optimization F2 Acetonitrile stir time (min) 2 (B) 5 (b) F3 Type of magnesium sulfate (brand) Sigma (C) Fluka (c) The matrix of study was a biomixture, a sample with high organic F4 Shaker agitation time (min) 30 (D) 15 (d) and low water content (< 30%). Pesticide extraction from this type of F5 Centrifuge time (min) 7 (E) 3 (e) F6 Centrifuge speed (rpm) 4000 (F) 2500 (f) matrix with low water content was previously evaluated with F7 Water bath temperature (℃) 30 (G) 40 (g) QuEChERS methodologies [2,7,8,23,27,37]. A 23 full-factorial design was used to determine the method with 3 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Table 3 Optimization of transitions for the quantification (Q) and confirmation (q) of the studied analytes by LC-MS/MS. Compound Transition Fragmentor (eV) Collision energy (eV) Retention time (min) Type of transition Precursor ion Product ion Acephate 184 143 60 5 0.91 Q 95 25 q Amethryn 228 186 106 17 8.05 Q 96 25 q Anilophos 368 199 70 10 13.58 Q 171 15 q Atrazine 216 174 106 17 9.42 Q 96 25 q Azinphos-methyl 318 132 60 13 10.65 Q 125 17 q Bitertanol 338 99 82 13 14.05 Q 269 5 q Cadusafos 271 159 70 5 14.32 Q 131 20 q Chlorpyrifos 350 97 90 30 15.74 Q 198 15 q Coumaphos 363 277 116 25 13.67 Q 307 13 q Cyanazine 241 214 100 15 7.05 Q 104 30 q Cyproconazole 292 70 110 15 12.42 Q 125 30 q Cyromazine 167 85 104 17 0.87 Q 60 21 q Dichlorvos 221 109 104 13 7.40 Q 79 29 q Difenoconazole 406 251 126 25 14.42 Q 337 13 q Dimethoate 230 199 70 3 3.45 Q 125 20 q Edifenphos 311 111 90 20 13.46 Q 283 10 q Epoxiconazole 350 121 106 25 12.75 Q 101 40 q Ethoprophos 243 97 84 33 12.52 Q 131 17 q Fenamiphos 304 217 138 21 13.13 Q 234 13 q Fenbuconazole 337 70 116 17 12.95 Q 125 40 q Fenthion 279 247 104 9 13.83 Q 105 25 q Flusilazole 316 165 110 25 13.14 Q 247 15 q Hexaconazole 314 70 116 21 13.85 Q 159 33 q Heptenophos 251 127 110 10 10.23 Q 109 30 q Isazofos 31 120 94 29 12.25 Q 162 13 q Malathion 331 99 82 25 11.83 Q 127 9 q Methamidophos 142 94 90 10 0.98 Q 125 10 q Monocrotophos 224 127 62 13 1.85 Q 193 5 q Myclobutanil 289 70 106 17 11.86 Q 125 37 q Paclobutrazol 294 70 110 15 11.81 Q 125 35 q Phoxim 321 192 94 9 13.88 Q 115 21 q Pirimiphos-methyl 306 164 90 20 13.38 Q 108 30 q Prometon 226 142 116 21 7.15 Q 184 17 q Prometryn 242 200 126 17 9.71 Q 158 21 q Propiconazole 342 159 126 29 13.61 Q 69 21 q Simazine 202 124 106 17 7.42 Q 104 25 q (continued on next page) 4 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Table 3 (continued) Compound Transition Fragmentor (eV) Collision energy (eV) Retention time (min) Type of transition Precursor ion Product ion Simetryn 214 124 106 17 6.08 Q 96 25 q Tebuconazole 308 70 106 21 13.55 Q 125 40 q Terbuthylazine 230 174 104 13 11.33 Q 96 29 q Terbutryn 242 186 96 17 9.83 Q 91 29 q Triadimefon 294 197 94 13 11.74 Q 69 21 q Triadimenol 296 70 72 9 12.35 Q 99 13 q Triazophos 314 162 100 15 12.25 Q 119 35 q Linuron-d6 (IS) 255 160 92 17 11.12 Q 185 13 q Carbofuran-d3 (SS.) 225 165 86 9 7.67 Q 123 21 q IS: internal standard; SS: surrogate standard. best recovery for more molecules in the same experiment. Three critical triazines and triazoles and up to 19% for organophosphates. The ad- factors of the methodology were evaluated: the amount of water added dition of 10 mL of water enhanced the extraction recovery rates of to 5 g of the biomixture (5 mL vs. 10 mL), the type of agitation em- 3–5% out of 37 molecules, particularly organophosphates. This effect ployed during extraction (automated vs. manual), and the amount of was attributed to the time of hydration of the matrix during the ex- magnesium sulfate added for sample drying in the cleaning stage traction, since it allows the opening of the pores of the matrix which (450 mg vs. 900 mg). The experimental conditions are shown in Table leads to a better extraction of the molecules by the solubility of the S2 (Supplementary Material). The examination of the effects of the acetonitrile in water medium [38,39]. main factors and the interactions between factors revealed that in- Recoveries for cyromazine, cyproconazole, acephate and methami- dividual factors produce the greatest inference on the analysis metho- dophos decreased as the volume of water was increased. This is at- dology. tributed to the high water solubility of these molecules, which causes a The first factor evaluated was the amount of water added to the greater affinity to the aqueous phase and subsequent losses by the ad- matrix (Fig. 2); the recoveries exhibited differences of up to 5% for dition of sodium sulfate to dry the sample. However, other molecules 1A) 1B) 1C) 2A) 2B) 2C) 3A) 3B) 3C) Fig. 1. Optimization of voltage fragmentor and precursor ion (A); the quantification product ion (B); and confirmation product ion (C) for ametryn (1), ethoprophos (2), and flusilazol (3). 5 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Fig. 2. Recovery obtained from the design of comparative experiments 23 during the optimization of methodological factors: i. addition of water (5 mL vs. 10 mL) during the extraction; ii. automated or manual agitation during extraction; and iii. the amount of sulphate magnesium (450 mg vs. 900 mg) added in the cleaning step for: (A) Organophosphates; (B) Triazines, (C) Triazoles. 6 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 with higher water solubility like monocrotophos did not exhibit this voltages of individual molecules were optimized (Table 3). Then, a effect. working solution containing the 43 analytes, carbofuran-d3 and li- When comparing the agitation types (Fig. 2) (either with a pro- nuron-d6 was injected. Also, the following 52 additional molecules grammable automated agitation equipment vs. manual agitation), oxamyl, carbendazim, carbendazim-d4, amitraz, methomyl, thia- variability between both methodologies was below 2% for most of the methoxam, thiabendazole, imidacloprid, picloram, imazapyr, pir- pesticides. This finding implies that manual agitation can be used for imicarb, 3-hydroxycarbofuran, acetamiprid, cymoxanil, imazapic, al- this type of samples, as usually employed for the traditional QuEChERS dicarb, 3-ketocarbofuran, monuron, metribuzin, bromacil, propoxur, methodology applied to vegetable matrices [40–42]. The application of hexazinone, carbofuran, thiophanate-methyl, pyrimethanil, bentazone, manual agitation in each extraction stage produced similar results to metsulfuron-methyl, carbaryl, imazalil, metalaxyl, isoproturon, diuron, the use of agitation for 30 min at 2500 rpm. Cyproconazole was the thiophanate, linuron, azoxystrobin, propanil, methiocarb, molinate, only molecule that showed an improvement of 15% when working with dimethomorph, myclobutanil, fenarimol, prochloraz, fipronil, kre- the programmable agitator. Considering the availability of the pro- soxim-methyl, haloxyfop, pyraclostrobin, triflumuron, buprofezin, ha- grammable shaker, that allows the simultaneous processing of several loxyfop-p-methyl, fluazifop-p-butyl, teflubenzuron and pendimethalin samples, the use of programmable agitation was selected for the pur- were added to the previous mixture at a concentration of 200 μg/kg pose of the proposed method. each. None of these additional 52 molecules (excluded from the vali- The third factor of study was the amount of magnesium sulfate used dation) showed interference signals on the optimized triazine, triazole in the cleaning stage (Fig. 2). Magnesium sulfate is added to remove the and organophosphate transitions, thus demonstrating the method is excess of water in the samples, since the proposed methodology re- selective. quires a subsequent step of concentration to dryness. Also, it reduces The specificity of the method was studied with five blank samples the number of co-extracts due to the decrease in polarity in the ex- and five spiked samples; their comparison confirmed the absence of traction acetonitrile phase [40]. The addition of magnesium sulfate did false positives for the studied molecules, thus suggesting it is a specific not show significant differences in the groups of triazines and organo- method. In the case of ametryn, atrazine, cyromazine, cyproconazole, phosphates; however, an increase in the recoveries between 1 and 5 % tebuconazole, triadimenol, chlorpyrifos and fenamiphos, slight signals was observed when using 900 mg of magnesium sulfate and the chro- were identified in their respective transitions, which implies a slight matogram showed less interferences. Yet, most of the triazoles were interference for subsequent detections. The detected signals showed S/ favored when using 900 mg of magnesium sulfate. This is expected as N ratios > 10, which was the criterion to assign a positive signal. The this is the group with the lowest water solubility reported. Based on this S/N ratios and the ion ratio of the spiked samples were higher than the finding, 900 mg of magnesium sulfate were used for improving the LOD, which does not affect their selectivity when applying the meth- recovery of most of the studied molecules. odology (Table S3, Supplementary Material). In summary, the key factor that permitted to achieve better recovery The ion ratio criterion was also used to evaluate both parameters, by was the addition of 10 mL of water to the biomixture, with favorable comparing the ion ratio for each compound with the ion ratio of cali- results on most of the compounds. Although the amount of magnesium bration curves. All the compounds showed good results in the se- sulfate added for cleaning did not result in significantly different values lectivity test (spiked samples), as the ion ratio of the sample extracts for the study molecules, the cleaning with 900 mg of magnesium sulfate were ± 30% of the average for calibration standards. In the case of increased the recovery of triazoles by up to 5% and eliminated inter- specificity, all the compounds exhibited ion ratios out of the selection ferences, which favors parameters for the methodology validation such criterion. as selectivity. 3.3.2. Limit of determination (LOD) 3.3. Parameters for methodology validation Since the method is intended to be applied in biomixtures used for pesticide treatment, the expected concentrations in this matrix are quite After optimizing the aforementioned factors, the methodology high, in the order of more than 10 mg/kg (particularly at the moment of proposed employed 5 g of sample and the addition of 10 mL of DDI the disposal of pesticide-containing wastewater). The LOD was de- water, 15 mL of acetonitrile acidified (1% v/v acetic acid), 6 g of termined based on the lowest residue concentration that can be quan- magnesium sulfate, 1 g of sodium chloride and 2.6 g of sodium acetate tified for each pesticide, with a S/N > 10 for the quantified and con- trihydrate, followed by shaking for 30 min, 2500 rpm, and cen- firmation ions. trifugation (4000 rpm, 10 °C, 7 min). After centrifugation, an aliquot of The LOD is identified as the lowest level of spiked sample with 3 mL of the extract was placed in a tube with 900 mg of magnesium acceptable recovery and precision; in some cases, it can be equated to sulfate, 150 mg of PSA and cleaned with 75 mg of C18. The sample was the maximum limit of residues (MRL); however, there are no MRL for then stirred and centrifuged again at the same conditions. A sample of this type of matrix. The criteria to accept the LOD was RSD ≤ 20%. The 1.5 mL of the supernatant was gently dried with a nitrogen stream and acceptance criterion was the detection of the two transitions and the ion finally reconstituted to 1.5 mL with acidified (0.1% v/v formic acid) response ratios from the sample, and the average of the calibration water-acetonitrile mixture (1:1), and filtered (0.45 µm PTFE filter) standards lower than ± 30% [34]. The results for these experiments are before being placed into a vial. presented in Table 4. The subsequent validation of the method included selectivity, spe- Cyromazine was the only molecule in the LOD study that presented cificity, precision, intermediate precision, LOD, trueness, linearity, ap- a RSD greater than 20%, which is justified as it is a basic ionic molecule, plication range, matrix effect and robustness. that may be adsorbed to soil due to the known sorption of triazines to the humic groups of soil [43]; however, during the extraction process, 3.3.1. Selectivity and specificity the sample reached an approximate pH value of 4.5, due to the acetate Selectivity is the ability of the method to discriminate between the buffer that was made in situ [43,44], which favors the extraction of analyte of interest and other molecules present in the matrix, while other triazines, but not the cyromazine that has been shown to require a specificity is the ability to obtain a negative result, when the samples do greater amount of acid for better extraction [45]. Some organopho- not have the analyte [34]. By carefully coupling the choosing of the sphates such as cadusafos, ethoprophos, fenthion and malathion had an solvents and reagents of the analytical extraction, to the properties of RSD value close to 15%. For these molecules, the greater variability at the LC-MS/MS technique to identify compounds according the opti- lower concentrations is not due to pH or pKa, but rather to the hy- mization of their mass, the results will be most of the times selective. drogen bonds that are formed between pesticides and ionic compounds The retention times, transitions and fragmentation and collision cell in the soil [43]. 7 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Table 4 Results of average RSD calculated for the precision parameter at three spiked levels (n = 7) and intermediate precision (n = 21); average recovery (%) calculated for trueness (n = 7), and LOD (n = 7) for the method proposed to determine triazines, triazoles and organophosphates in a biomixture composed by soil/compost/ coconut fiber. Compound Precision (as repeatability) (RSD < 20%) Intermediate precision (RSD < 20%) Trueness (Recovery % n = 7) LOD 10 µg/kg (RSD) 50 µg/kg 150 µg/kg 350 µg/kg 50 µg/kg 150 µg/kg 350 µg/kg 50 µg/kg Acephate* 5.06 6.16 8.72 12.25 6.07 6.89 92.1 11.4 Amethryn 3.27 7.77 2.80 4.44 5.04 10.74 106.0 4.2 Anilophos 4.17 4.63 3.47 6.77 5.27 8.39 90.0 6.4 Atrazine 3.83 7.06 3.30 5.72 4.60 7.75 111.5 5.7 Azinphos-methyl 3.25 7.58 2.35 5.90 5.15 6.85 107.6 5.5 Bitertanol 4.70 7.88 3.14 5.07 5.10 13.47 108.0 4.8 Cadusafos 6.92 7.48 5.06 21.77 8.34 12.42 99.9 17.6 Chlorpyrifos 7.81 10.66 3.99 12.76 7.53 10.97 103.2 9.7 Coumaphos 4.50 8.81 5.12 8.43 7.01 12.92 110.2 7.9 Cyanazine 5.19 7.32 3.16 5.38 4.73 10.01 112.1 5.3 Cyproconazole 6.27 8.59 5.25 7.67 6.10 12.24 116.4 7.4 Cyromazine 23.66 17.38 28.57 61.15 40.95 54.86 49.4 38 Dichlorvos 43.14 20.31 19.09 40.19 25.56 25.80 47.1 12.4 Difenoconazole 3.34 7.38 4.53 4.94 4.76 15.64 112.5 4.6 Dimethoate 3.76 7.78 4.65 5.45 5.18 5.57 110.4 5.6 Edifenphos 4.19 4.47 4.40 4.33 6.08 8.30 114.7 4.1 Epoxiconazole 2.75 7.62 2.88 4.19 5.13 13.67 108.2 4.1 Ethoprophos 7.32 10.09 2.78 11.05 9.29 9.93 97.0 14.5 Fenamiphos 4.72 6.47 3.88 8.15 6.39 8.78 105.3 8.1 Fenbuconazole 4.60 7.78 4.40 5.95 4.96 12.78 107.0 5.6 Fenthion 11.60 6.72 2.94 14.47 11.60 14.25 101.3 13.1 Flusilazole 4.01 6.77 3.84 5.33 4.76 13.23 108.1 5.2 Hexaconazole 3.63 8.09 3.67 4.29 5.39 12.80 102.1 4.2 Heptenophos 10.24 5.21 3.43 9.83 6.92 6.73 89.3 7.2 Isazofos 3.03 6.79 2.65 5.39 4.69 8.84 112.1 5.2 Malathion 8.28 4.76 3.41 19.41 9.18 8.54 93.7 13.4 Methamidophos 5.68 7.63 8.19 10.83 8.18 14.30 78.2 9.8 Monocrotophos 4.54 6.96 4.68 6.77 4.84 6.99 107.7 6.8 Myclobutanil 3.35 7.26 3.09 5.62 5.03 11.73 108.1 5.5 Paclobutrazol 3.94 7.89 3.24 4.16 6.08 10.06 106.7 4.0 Phoxim 3.35 5.52 5.14 8.63 5.86 8.73 114.2 8.1 Pirimiphos-methyl 2.74 7.78 3.32 6.74 5.14 13.36 113.0 6.3 Prometon 3.40 7.98 3.15 4.71 5.10 10.04 105.2 4.6 Prometryn 3.32 7.66 2.95 4.50 4.77 10.40 108.5 4.4 Propiconazole 4.16 7.55 3.38 5.43 5.17 14.10 103.7 5.1 Simazine 3.53 7.42 2.50 5.24 4.87 9.52 109.5 5.1 Simetryn 2.72 7.59 1.77 5.13 5.13 11.03 103.7 4.6 Tebuconazole 11.70 6.74 2.78 15.66 6.65 16.54 107.7 13.9 Terbuthylazine 3.28 7.09 3.26 5.07 4.48 10.14 109.3 4.9 Terbutryn 3.10 8.15 2.90 4.36 5.31 10.88 108.1 4.2 Triadimefon 2.52 7.56 2.64 4.35 5.21 12.86 109.9 4.2 Triadimenol 4.84 7.46 4.04 7.90 6.68 10.63 105.5 7.6 Triazophos 4.61 7.40 2.82 6.55 4.75 8.06 113.4 6.4 LOD: Limit of determination. * Acephate had a LOD of 50 µg/kg. 3.3.3. Precision as repeatability and intermediate precision variation at low concentrations, which decreased at higher concentra- After detecting the concentrations for LOD, it was necessary to work tions. All organophosphates but dichlorvos presented results that met at three higher concentrations (50, 150 and 350 μg/kg) to establish the the acceptance criteria. From the remaining 19 molecules, chlorpyrifos precision of the methodology. The critical criterion was an RSD < 20%. showed the highest RSD value. Acephate and methamidophos exhibited The precision as repeatability was determined with the results of one of a different behavior, as their RSD increased with the concentration. The the analysts (n = 7); intermediate precision was determined with three polarity of both molecules and the possibility to have greater hydrogen analysts who performed the methodology at different days (n = 21). bonding between the molecules and the biomixture, may favor their The results are shown in Table 4. greatest data dispersion; nonetheless, the dispersion was not greater After the evaluation of 43 molecules, cyromazine and dichlorvos than that allowed by the validation criterion in this case. For the case of showed the greatest dispersion data along the three concentration levels cadusafos, no precision was considered at low concentration of the intermediate precision parameter (> 40% and > 25%, respec- (RSD = 21.77% at 50 μg/kg). All the triazoles presented acceptable tively). The precision as repeatability and intermediate precision had an values of standard deviation, as the other nine triazines. RSD higher than 20%, which exceeds the acceptance criterion of SANTE The RSD values were in general higher for intermediate precision [34]. The results for these molecules suggest the need to consider the than repeatability. This is expected, as data executed by three different behavior of other parameters to determine whether the methodology is analysts on different days was employed, which consequently gives a appropriate for their analysis. Thus, it is considered that the multiresidue greater variability to the method. method proposed is not precise for these molecules, or in the worst-case scenario, it would be better to test other methodologies for their analysis. The remaining 41 molecules exhibited higher coefficients of 3.3.4. Trueness The trueness of the method is applied as a validation parameter in 8 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 the absence of an interlaboratory test for the biomixture matrix. The avoids the production of high concentration of other compounds, which trueness is understood as the average recovery of the concentration may increase the surface tension, the viscosity of the droplets in the levels evaluated with a recovery percentage between 70 and 120 % nebulizer, and the proton affinity between the analytes and the co-ex- [34]. The trueness of the method was determined using a low con- tracts [51]. centration, that is, the recovery value of 50 μg/kg. When comparing these results with those obtained in Table 4, it was 3.3.7. Application range observed that dichlorvos (47.1%) and cyromazine (49.4%) presented When an unknown sample is processed, there is a risk that the results below the acceptance criteria (70–120%). For acephate and concentration of the analyte surpasses that of the highest level in the methamidophos, the recoveries were quite good compared to other calibration curve; this implies that the sample must be diluted so that methodologies applied to fruits, vegetables, meats and soils, since they the analyte concentration remains within the validated parameters. In showed values of 92% and 78% respectively [46–49]. The organo- order to corroborate the method properly extracts and detects higher phosphates anilophos (90%) and heptenophos (89%) also exhibited concentrations than those validated with satisfactory recovery, a two- relatively low recoveries, thus, these molecules should be carefully re- part experiment was performed. For the first one, the methodology was vised in a control chart to verify that this average is maintained. The applied to five spiked samples with a concentration of 2000 μg/kg. The recovery values obtained for the triazoles were within the acceptance final extracts were diluted to an intermediate value of the calibration criterion. curve, and then quantified. The criteria for this parameter was the re- More than 80% of the molecules showed recoveries > 100%, a covery (> 70%, < 120%) and RSD (< 20%). finding that could be ascribed to the ionization technique (electrospray Only cyromazine and dichlorvos presented recoveries below 70% system) applied in the LC-MS/MS, which exerts a signal improvement after dilution (Table 5). The remaining 41 molecules were extracted effect. This behavior is classified by some authors as a cause of matrix from the matrix at a concentration of 2000 μg/kg and still had a re- effect [50–52]; nonetheless in this case, the matrix effect was negligible covery between 70% and 120%, with an RSD less than 20%. This im- in most of the molecules and the ion ratio values were acceptable and plies that, although there is no linear relationship between concentra- less than 30% deviated from the reference value in the calibration curve tion and response at high concentrations of the analyte, the method [34]. allows to work at such concentrations using proper dilutions. Moreover, the sample dilution provides the desired effect of an additional decrease 3.3.5. Linearity in the matrix effect. Three independent (not consecutive) calibration curves were pre- pared with 10 concentration levels each in the blank extract sample, 3.3.8. Robustness including concentrations from the LOD value (10 μg/kg) to 1000 μg/kg. The robustness test was carried out to demonstrate that the meth- Three acceptance criteria were considered: (i) correlation coefficient odology is still reliable and accurate, after the variation of several ex- greater than 0.99; (ii) the percentage of residuals for each level must be traction conditions (factors) (Table S4, Supplementary Material). Sev- less than 20%; and (iii) the slope ratio between the three calibration eral experimental designs are used to evaluate the robustness of curves should be higher than 80%. Table 5 shows the results of the analytical methods [57–61]; the Youden-Steiner test was performed in calibration curves of the study molecules. this work. This test consists of a fractional factorial design of resolution Every single molecule yielded acceptable results for each criterion. III, which is represented by the mathematical model 2 7-4III . It works at However, it is important to consider that, for acephate, azinphos-me- two levels of effect, in which seven factors or conditions that can result thyl, coumaphos, cyromazine, dichlorvos, fenamiphos, fenbuconazole, in significantly different results are considered (but not their interac- methamidophos, and pirimiphos-methyl, one of the work levels reached tions), for a total of eight experiments [35,58,62,63]. residual values close to or greater than 10%, which implies that care The robustness was calculated by comparing the difference of the must be taken in estimating the curve, as this could lead to an increase values of each factor, according to Tables 1 and 2, in relation to the of LOD if lower levels need to be eliminated to ensure linearity range. value calculated as critical, which depends on the total standard de- viation of the experiment [35,58,59]. The results are shown in Table S4 3.3.6. Matrix effect of the Supplementary Material. The matrix effect (ME) is the comparison between the response of Forty molecules met the robustness criterion. In particular, for the calibration levels in the organic solvent and the matrix [34,53–56]. amethryn, simazine, dimethoate, bitertanol and hexaconazole the seven The acceptance criterion was ME lower than 20%. The curve that was proposed factors did not statistically affect the behavior of the mole- prepared in the matrix presented the same characteristics as the final cules; on the other hand, four analytes (chlorpyrifos, coumaphos, cy- extract (a phase of water-acetonitrile (1:1), acidified with formic acid romazine and methamidophos) showed high critical values (> 20) that 0.1%). The results are shown in Table 5. represented their high variability; nonetheless they passed the robust- Cyromazine (17%), methamidophos (9.2%) and phoxim (7.6%) ness test. This finding represents a drawback of the statistical method; presented the greatest matrix effect; throughout the validation these as the critical factor increases with RSD > 20%, this method may mask molecules have shown the least favorable performance in the multi- the effect of conditions in the case of molecules with poor precision. residue methodology. The other molecules evaluated in the metho- Other two molecules, dichlorvos and fenamiphos, showed high critical dology presented ME less than 4%. However, none of the molecules values of 14 and 16, but they are still considered as acceptable. failed to meet the acceptance criterion of ME < 20%. Only cyproconazole, dichlorvos, fenamiphos and methamidophos The fact that no considerable matrix effect was observed in the failed the robustness test for one condition. In the case of cyprocona- biomixture was due to the change in solvent, since there was a change zole, the condition was the temperature of the water bath; as the in polarity of the sample and this caused a decrease in the amount of co- temperature increased, a lower recovery of the molecule was obtained, extracts before injection [52]. This finding was also supported by the which implies that temperature control must be used during sample direct observation of less particles remaining in the sample container concentration with nitrogen. In the case of fenamiphos, the critical and the filter used. factor was the agitation time with acetonitrile, as a decrease greater The decrease in co-extracts favors the formation of ions than 20% in the percentage of recovery was obtained with less agita- [51,53,54,56], since it reduces the ionic suppression effect normally tion. For methamidophos, a greater variability was obtained as a cause achieved in the ionization technique by electrospray (which decreases of a shorter agitation time in the centrifuge, which does not allow an the signal), compared to the chemical ionization technique at atmo- adequate phase separation. On the other hand, dichlorvos was affected spheric pressure (which increases the signal) [50–52,56]. This factor by the time it was in contact with water; however, this molecule 9 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Table 5 Results of validation parameters: linearity, application range and matrix effect, for the method proposed to determine triazines, triazoles and organophosphates in a biomixture composed by soil/compost/coconut fiber. Compound Linearity (3 replicates; 10 calibration levels) Application range; spiked at 2 mg/kg (RSD % Matrix effect with n = 21; diluted samples) (Dif < 20%) r2 Highest % of residuals Calibration curve detected < 20% Acephate 0.994 9.99 y = 0.00043749 x – 102.5 (7.0) 0.24 0.00000017 Amethryn 0.999 5.38 y = 0.004057 x – 0.000042 106.6 (5.0) 1.57 Anilophos 0.996 7.88 y = 0.0002312 x – 0.0000019 108.4 (6.5) 0.22 Atrazine 0.992 6.49 y = 0.002688 x – 0.000031 108.2 (3.3) 0.48 Azinphos-methyl 0.996 10.0 y = 0.00006759 x – 106.2 (4.1) 2.20 0.00000063 Bitertanol 0.991 7.50 y = 0.0003322 x – 0.0000027 109.8 (4.9) 0.79 Cadusafos 0.995 7.89 y = 0.0006059 x – 0.0000046 94.1 (3.8) 1.22 Chlorpyrifos 0.994 6.31 y = 0.00009017 x – 103.8 (4.3) 1.19 0.00000088 Coumaphos 0.991 10.38 y = 0.0001379 x – 0.0000019 118.1 (7.8) 4.00 Cyanazine 0.998 8.41 y = 0.0004952 x – 0.0000066 115. 8 (4.7) 0.48 Cyproconazole 0.991 6.16 y = 0.0008256 x – 0.0000026 110.1 (4.4) 1.00 Cyromazine 0.997 14.33 y = 0.00043203 x + 29.4 (49) 17.29 0.00000018 Dichlorvos 0.995 9.85 y = 0.0002714 x – 0.0000019 61.7 (12.4) 1.64 Difenoconazole 0.990 6.90 y = 0.0009981 x – 0.0000095 99.4 (6.7) 2.06 Dimethoate 0.998 5.60 y = 0.00040682 x – 100.7 (3.9) 0.92 0.00000034 Edifenphos 0.997 5.56 y = 0.0002329 x – 0.0000021 113.3 (4.0) 1.05 Epoxiconazole 0.993 8.44 y = 0.000928 x – 0.000011 109.1 (3.8) 1.87 Ethoprophos 0.995 6.09 y = 0.0002974 x – 0.0000025 103.0 (3.7) 1.10 Fenamiphos 0.991 12.03 y = 0.0002977 x – 0.0000041 108.1 (5.9) 2.71 Fenbuconazole 0.992 9.70 y = 0.0005668 x – 0.0000077 111.4 (4.9) 0.10 Fenthion 0.995 8.65 y = 0.00005111 x – 102.7 (6.8) 1.32 0.000000095 Flusilazole 0.993 8.78 y = 0.001284 x – 0.000016 110.2 (4.4) 0.50 Hexaconazole 0.993 4.91 y = 0.001271 x – 0.000015 108.8 (4.3) 0.02 Heptenophos 0.996 7.48 y = 0.0002239 x – 0.0000016 100.2 (3.5) 2.85 Isazofos 0.995 7.89 y = 0.001438 x – 0.000011 109.4 (4.0) 0.66 Malathion 0.995 5.84 y = 0.000245 x – 0.000013 95.2 (4.3) 1.87 Methamidophos 0.997 10.44 y = 0.00017877 x – 102.8 (17.2) 9.15 0.00000098 Monocrotophos 0.996 5.62 y = 0.0003956 x – 0.0000032 109.1 (4.7) 2.57 Myclobutanil 0.994 4.74 y = 0.0009990 x – 0.0000084 111.0 (4.3) 0.62 Paclobutrazol 0.995 6.58 y = 0.002872 x – 0.000031 112.1 (5.7) 2.99 Phoxim 0.997 8.43 y = 0.00008963 x – 98.5 (4.5) 7.59 0.00000027 Pirimiphos-methyl 0.994 11.07 y = 0.003116 x – 0.000039 117.8 (4.2) 0.89 Prometon 0.993 5.11 y = 0.004279 x + 0.000054 105.1 (3.2) 0.65 Prometryn 0.994 5.97 y = 0.005715 x – 0.000071 107.5 (4.6) 0.13 Propiconazole 0.992 5.73 y = 0.0007152 x – 0.0000092 111.2 (4.1) 1.55 Simazine 0.993 5.07 y = 0.000963 x – 0.000012 108.3 (4.5) 0.56 Simetryn 0.994 4.89 y = 0.001779 x – 0.000021 101.7 (4.6) 0.83 Tebuconazole 0.994 6.09 y = 0.0015468 x – 107.5 (4.0) 1.66 0.00000088 Terbuthylazine 0.994 7.66 y = 0.006191 x – 0.000071 111.3 (4.0) 0.64 Terbutryn 0.993 7.24 y = 0.006781 x – 0.000088 106.3 (4.5) 0.20 Triadimefon 0.995 6.57 y = 0.0007691 x – 0.0000069 109.2 (4.1) 0.30 Triadimenol 0.992 5.28 y = 0.001368 x – 0.000011 103.2 (3.9) 1.24 Triazophos 0.996 5.68 y = 0.001367 x – 0.000011 105.4 (4.2) 0.02 consistently exhibited lower recoveries and unsatisfactory RSD. determination of 41 of the evaluated molecules in the biomixture. Summarizing, every analyte showed satisfactory results for the de- termination of LOD, linearity, specificity and selectivity. Besides these 3.4. Removal of pesticide-containing wastewater in a BPS: application of parameters, the proposed methodology exhibited low matrix effects; the method only cyromazine had a value out of the acceptance criterion. For the parameters of trueness and precision, cyromazine and dichlorvos were The methodology was applied to monitor the removal capacity of a the only two molecules that showed unsatisfactory results, with values 104 L BPS. The synthetic wastewater applied on the biomixture con- of recovery < 70% and RSD > 20%. The lack of precision and accuracy tained eight pesticides (ametryn, atrazine, chlorpyrifos, diazinon, ma- suggests the use of another methodology to work with these molecules. lathion, tebuconazole, terbutryn and triadimenol), at similar con- The parameter of robustness demonstrated that, for most of the mole- centrations to those expected after the disposal of wastewater residues cules, slight changes do not affect the methodology performance. Only from field application, according to the recommendation in the com- four molecules exhibited unsatisfactory results for one different meth- mercial formulations. Initial concentrations in the BPS ranged from odology condition. Overall, the proposed methodology, except for di- 3.9 mg/kg to 51.1 mg/kg (malathion and tebuconazole, respectively). chlorvos and cyromazine, met the acceptance criteria for the Triadimefon was not added to the wastewater, and was initially 10 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Fig. 3. Removal of eight pesticides from commercial formulations during their treatment in a BPS for 53 days. Plotted values are means ± SD. detected at 0.09 mg/kg. Its origin is likely due to contamination in the 4. Conclusions formulation containing tebuconazole/triadimenol, as triadimenol is a known transformation product of triadimefon [64]. The modified QuEChERS methodology was validated for the analysis Although most of the compounds were at least partially removed of pesticides in a solid matrix (biomixture) made up of soil, compost and (Fig. 3), none of them was eliminated at levels below the LOD. The coconut fiber, aimed to remove pesticides from wastewater of agricultural triazole tebuconazole and the organophosphate diazinon were not sig- origin. The method was proved under several validation parameters, nificantly removed. Previous works on BPS report unsuccessful elim- where the results were satisfactory for most of the triazines, triazoles and ination of triazoles, including tebuconazole and triadimenol [2,28,65]; organophosphates evaluated, except for dichlorvos and cyromazine, nonetheless, in this case triadimenol was partly removed at the end of which did not meet the acceptance criteria for some parameters. the treatment after 53 d (up to 51.8%). Many investigations indicate The developed LC-MS/MS methodology allows working with low that triadimenol is a metabolite of triadimefon [64,66]; however, in this and high pesticide concentrations, with good recovery percentages experiment triadimefon was not added and an increase in its levels was between 70% and 120%, coefficients of variation of less than 20%, with observed (up to 5.14 mg/kg). This finding could be due to an oxidation linearity conditions that exceed the coefficient of determination value of the triadimenol in the biomixture, favored by the conditions of of 0.99, and with matrix effects lower than 20%, for 41 out of the 43 temperature and humidity, a scarcely studied reaction described by evaluated molecules. Deas & Clifford [67] in transformations with fungi. Other biomixtures The validated methodology was successfully applied to determine have shown the ability to remove diazinon in peat-based biomixtures, the efficiency of a pilot scale biopurification system, employed for the with DT50 in the range of 4.9 to 10.8 d, with an accelerated effect after removal of wastewater residues containing eight pesticides from com- successive applications [68]. Contrary to diazinon, other organopho- mercial formulations. sphates were removed from the biomixture at different rates; chlor- pyrifos at an estimated DT50 of 10.5 d, while malathion concentration CRediT authorship contribution statement decreased to only 1.9% after nine days of treatment (DT50 = 1.6 d). The removal of chlorpyrifos was significantly faster than data from soil Mario Masís-Mora: Conceptualization, Methodology, Formal ana- (DT50 27–386 d) [69] and slightly faster than reported in other bio- lysis, Investigation, Funding acquisition, Writing - original draft, mixtures, for which DT50 values are within the range 15–59 d Writing - review & editing. Wilson Beita-Sandí: Formal analysis, [65,70–72]. The fast elimination of malathion in biomixtures was also Writing - review & editing. Javier Rodríguez-Yáñez: described in a peat-based matrix (DT50 3.8 d) [25] and a compost-based Conceptualization, Writing - review & editing. Carlos E. Rodríguez- mixture (DT50 7.1 d). From the three triazines tested, atrazine exhibited Rodríguez: Conceptualization, Project administration, Writing - review the faster removal (estimated DT50 11.2 d), followed by ametryn (DT50 & editing, Funding acquisition. 13.4 d) and terbutryn (DT50 19.4 d). The removal of atrazine has been widely described in biomixtures, with DT50 values ranging from as low Declaration of Competing Interest as < 10 d (after single or repeated applications) [29,32,65] to more than 20 d [2,73]. Comparable removal patterns to those observed in The authors declare that they have no known competing financial this work have been achieved for ametryn [2] and terbutryn [30] in interests or personal relationships that could have appeared to influ- compost-based biomixtures. ence the work reported in this paper. 11 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 Acknowledgements containing wastewaters, FEMS Microbiol. Lett. 345 (2013) 1–12, https://doi.org/ 10.1111/1574-6968.12161. This work was supported by the Vicerrectoría de Investigación, [18] J.M. Castillo Diaz, L. Delgado-Moreno, R. Núñez, R. Nogales, E. Romero, Enhancing pesticide degradation using indigenous microorganisms isolated under high pesti- University of Costa Rica (Project 802-B4-503) and the Ministry of cide load in bioremediation systems with vermicomposts, Bioresour. Technol. 214 Science, Technology and Telecommunications of Costa Rica (Project FI- (2016) 234–241, https://doi.org/10.1016/j.biortech.2016.04.105. 093-13). The authors would like to thank the cooperation of the master [19] G.F. Antonious, On-farm bioremediation of dimethazone and trifluralin residues in runoff water from an agricultural field, J. Environ. Sci. Heal. - Part B Pestic. Food program “Management of Natural Resources with an emphasis in Contam. Agric. Wastes. 47 (2012) 608–621, https://doi.org/10.1080/03601234. Environmental Management” of the Universidad Estatal a Distancia 2012.668454. (UNED, Costa Rica). [20] L. Coppola, F. Comitini, C. Casucci, V. Milanovic, E. Monaci, M. Marinozzi, M. Taccari, M. Ciani, C. Vischetti, Fungicides degradation in an organic biomixture: impact on microbial diversity, N. Biotechnol. 29 (2011) 99–106, https://doi.org/ Appendix A. Supplementary material 10.1016/j.nbt.2011.03.005. [21] C. Vischetti, E. Capri, M. Trevisan, C. Casucci, P. Perucci, Biomassbed: A biological system to reduce pesticide point contamination at farm level, Chemosphere 55 Supplementary data to this article can be found online at https:// (2004) 823–828, https://doi.org/10.1016/j.chemosphere.2003.11.042. doi.org/10.1016/j.jchromb.2020.122296. [22] M. Asgher, H.N. Bhatti, M. Ashraf, R.L. Legge, Recent developments in biode- gradation of industrial pollutants by white rot fungi and their enzyme system, Biodegradation 19 (2008) 771–783, https://doi.org/10.1007/s10532-008-9185-3. References [23] C.E. Rodríguez-Rodríguez, K. Madrigal-León, M. Masís-Mora, M. Pérez-Villanueva, J.S. Chin-Pampillo, Removal of carbamates and detoxification potential in a bio- [1] K. Adou, W.R. Bontoyan, P.J. Sweeney, Multiresidue Method for the Analysis of mixture: Fungal bioaugmentation versus traditional use, Ecotoxicol. Environ. Saf. Pesticide Residues in Fruits and Vegetables by Accelerated Solvent Extraction and 135 (2017) 252–258, https://doi.org/10.1016/j.ecoenv.2016.10.011. Capillary Gas Chromatography, J. Agric. Food Chem. 49 (2001) 4153–4160, [24] K. Madrigal-Zúñiga, K. Ruiz-Hidalgo, J.S. Chin-Pampillo, M. Masís-Mora, V. Castro- https://doi.org/10.1021/jf001528q. Gutiérrez, C.E. Rodríguez-Rodríguez, Fungal bioaugmentation of two rice husk- [2] A. Huete-Soto, M. Masís-Mora, V. Lizano-Fallas, J.S. Chin-Pampillo, E. Carazo- based biomixtures for the removal of carbofuran in on-farm biopurification systems, Rojas, C.E. Rodríguez-Rodríguez, Simultaneous removal of structurally different Biol. Fertil. Soils 52 (2016) 243–250, https://doi.org/10.1007/s00374-015-1071-7. pesticides in a biomixture: Detoxification and effect of oxytetracycline, [25] A.M. Bozdogan, N. Yarpuz-Bozdogan, H. Aka-Sagliker, M. Eren Oztekin, Chemosphere 169 (2017) 558–567, https://doi.org/10.1016/j.chemosphere.2016. N. Daglioglu, Determination of absorption and degradation of some pesticides in 11.106. biobed, J. Food Agric. Environ. 12 (2014) 347–351. [3] K.-H. Kim, E. Kabir, S.A. Jahan, Exposure to pesticides and the associated human [26] P. Fogg, A.B.A. Boxall, A. Walker, A.A. Jukes, Pesticide degradation in a “biobed” health effects, Sci. Total Environ. 575 (2017) 525–535, https://doi.org/10.1016/j. composting substrate, Pest Manage. Sci. 59 (2003) 527–537, https://doi.org/10. scitotenv.2016.09.009. 1002/ps.685. [4] H.M.G. van der Werf, Assessing the impact of pesticides on the environment, Agric. [27] V. Lizano-Fallas, M. Masís-Mora, D. Espinoza-Villalobos, M. Lizano-Brenes, Ecosyst. Environ. 60 (1996) 81–96, https://doi.org/10.1016/S0167-8809(96) C.E. Rodríguez-Rodríguez, Removal of pesticides and ecotoxicological changes 01096-1. during the simultaneous treatment of triazines and chlorpyrifos in biomixtures, [5] K. Bunzel, M. Liess, M. Kattwinkel, Landscape parameters driving aquatic pesticide Chemosphere 182 (2017) 106–113, https://doi.org/10.1016/j.chemosphere.2017. exposure and effects, Environ. Pollut. 186 (2014) 90–97, https://doi.org/10.1016/ 04.147. j.envpol.2013.11.021. [28] S. Murillo-Zamora, V. Castro-Gutiérrez, M. Masís-Mora, V. Lizano-Fallas, [6] M.del P. Castillo, L. Torstensson, J. Stenström, Biobeds for Environmental C.E. Rodríguez-Rodríguez, Elimination of fungicides in biopurification systems: Protection from Pesticide Use – A Review, J. Agric. Food Chem. 56 (2008) Effect of fungal bioaugmentation on removal performance and microbial commu- 6206–6219, https://doi.org/10.1021/jf800844x. nity structure, Chemosphere 186 (2017) 625–634, https://doi.org/10.1016/j. [7] J.S. Chin-Pampillo, K. Ruiz-Hidalgo, M. Masís-Mora, E. Carazo-Rojas, chemosphere.2017.07.162. C.E. Rodríguez-Rodríguez, Adaptation of biomixtures for carbofuran degradation in [29] V. Castro-Gutiérrez, M. Masís-Mora, E. Carazo-Rojas, M. Mora-López, on-farm biopurification systems in tropical regions, Environ. Sci. Pollut. Res. 22 C.E. Rodríguez-Rodríguez, Impact of oxytetracycline and bacterial bioaugmentation (2015) 9839–9848, https://doi.org/10.1007/s11356-015-4130-6. on the efficiency and microbial community structure of a pesticide-degrading bio- [8] J.S. Chin-Pampillo, K. Ruiz-Hidalgo, M. Masís-Mora, E. Carazo-Rojas, mixture, Environ. Sci. Pollut. Res. 25 (2018) 11787–11799, https://doi.org/10. C.E. Rodríguez-Rodríguez, Design of an optimized biomixture for the degradation of 1007/s11356-018-1436-1. carbofuran based on pesticide removal and toxicity reduction of the matrix, [30] J.C. Cambronero-Heinrichs, M. Masís-Mora, J.P. Quirós-Fournier, V. Lizano-Fallas, Environ. Sci. Pollut. Res. 22 (2015) 19184–19193, https://doi.org/10.1007/ I. Mata-Araya, C.E. Rodríguez-Rodríguez, Removal of herbicides in a biopurification s11356-015-5093-3. system is not negatively affected by oxytetracycline or fungally pretreated oxyte- [9] E. Karanasios, N.G. Tsiropoulos, D.G. Karpouzas, On-farm biopurification systems tracycline, Chemosphere 198 (2018) 198–203, https://doi.org/10.1016/j. for the depuration of pesticide wastewaters: recent biotechnological advances and chemosphere.2018.01.122. future perspectives, Biodegradation 23 (2012) 787–802, https://doi.org/10.1007/ [31] G.D. Gikas, M. Pérez-Villanueva, M. Tsioras, C. Alexoudis, G. Pérez-Rojas, M. Masís- s10532-012-9571-8. Mora, V. Lizano-Fallas, C.E. Rodríguez-Rodríguez, Z. Vryzas, V.A. Tsihrintzis, Low- [10] T. De Wilde, P. Spanoghe, C. Debaer, J. Ryckeboer, D. Springael, P. Jaeken, cost approaches for the removal of terbuthylazine from agricultural wastewater: Overview of on-farm bioremediation systems to reduce the occurrence of point Constructed wetlands and biopurification system, Chem. Eng. J. 335 (2018) source contamination, Pest Manag. Sci. 63 (2007) 111–128, https://doi.org/10. 647–656, https://doi.org/10.1016/j.cej.2017.11.031. 1002/ps.1323. [32] G.R. Tortella, R.A. Mella-Herrera, D.Z. Sousa, O. Rubilar, J.J. Acuña, G. Briceño, [11] T. De Wilde, P. Spanoghe, K. Sniegowksi, J. Ryckeboer, P. Jaeken, D. Springael, M.C. Diez, Atrazine dissipation and its impact on the microbial communities and Transport and degradation of metalaxyl and isoproturon in biopurification columns community level physiological profiles in a microcosm simulating the biomixture of inoculated with pesticide-primed material, Chemosphere 78 (2010) 56–60, https:// on-farm biopurification system, J. Hazard. Mater. 260 (2013) 459–467, https://doi. doi.org/10.1016/j.chemosphere.2009.10.011. org/10.1016/j.jhazmat.2013.05.059. [12] P. Fogg, A.B.A. Boxall, A. Walker, A. Jukes, Effect of Different Soil Textures on [33] P.N. Holmsgaard, S. Dealtry, V. Dunon, H. Heuer, L.H. Hansen, D. Springael, Leaching Potential and Degradation of Pesticides in Biobeds, J. Agric. Food Chem. K. Smalla, L. Riber, S.J. Sørensen, Response of the bacterial community in an on- 52 (2007) 5643–5652, https://doi.org/10.1021/jf040023n. farm biopurification system, to which diverse pesticides are introduced over an [13] A. Matilainen, E.T. Gjessing, T. Lahtinen, L. Hed, A. Bhatnagar, M. Sillanpää, An agricultural season, Environ. Pollut. 229 (2017) 854–862, https://doi.org/10.1016/ overview of the methods used in the characterisation of natural organic matter j.envpol.2017.07.026. (NOM) in relation to drinking water treatment, Chemosphere 83 (2011) [34] European Comision (EC), Analytical quality control and method validation proce- 1431–1442, https://doi.org/10.1016/j.chemosphere.2011.01.018. dures for pesticide residues analysis in food and feed. SANTE/12682/2019, 2019. [14] S. Reichenberger, M. Bach, A. Skitschak, H.-G. Frede, Mitigation strategies to reduce [35] E. Karageorgou, V. Samanidou, Youden test application in robustness assays during pesticide inputs into ground- and surface water and their effectiveness; A review, method validation, J. Chromatogr. A 1353 (2014) 131–139, https://doi.org/10. Sci. Total Environ. 384 (2007) 1–35, https://doi.org/10.1016/j.scitotenv.2007.04. 1016/j.chroma.2014.01.050. 046. [36] R. Raina, Chemical Analysis of Pesticides Using GC/MS, GC/MS/MS, and LC/MS/ [15] P.A. Karas, C. Perruchon, K. Exarhou, C. Ehaliotis, D.G. Karpouzas, Potential for MS, in: Pestic. - Strateg. Pestic. Anal., InTech, Rijeka, Croatia, 2011, pp. 105–130. bioremediation of agro-industrial effluents with high loads of pesticides by selected https://doi.org/10.5772/13242. fungi, Biodegradation 22 (2011) 215–228, https://doi.org/10.1007/s10532-010- [37] M.C. Bruzzoniti, L. Checchini, R.M. De Carlo, S. Orlandini, L. Rivoira, M. Del Bubba, 9389-1. QuEChERS sample preparation for the determination of pesticides and other or- [16] K. Sniegowski, K. Bers, K. Van Goetem, J. Ryckeboer, P. Jaeken, P. Spanoghe, ganic residues in environmental matrices: a critical review, Anal. Bioanal. Chem. D. Springael, Improvement of pesticide mineralization in on-farm biopurification 406 (2014) 4089–4116, https://doi.org/10.1007/s00216-014-7798-4. systems by bioaugmentation with pesticide-primed soil, FEMS Microbiol. Ecol. 76 [38] L. Correia-Sá, V.C. Fernandes, M. Carvalho, C. Calhau, V.M.F. Domingues, (2011) 64–73, https://doi.org/10.1111/j.1574-6941.2010.01031.x. C. Delerue-Matos, Optimization of QuEChERS method for the analysis of organo- [17] C.E. Rodríguez-Rodríguez, V. Castro-Gutiérrez, J.S. Chin-Pampillo, K. Ruiz-Hidalgo, chlorine pesticides in soils with diverse organic matter, J. Sep. Sci. 35 (2012) On-farm biopurification systems: role of white rot fungi in depuration of pesticide- 1521–1530, https://doi.org/10.1002/jssc.201200087. [39] J. Vera, L. Correia-Sá, P. Paíga, I. Bragança, V.C. Fernandes, V.F. Domingues, 12 M. Masís-Mora, et al. Journal of Chromatography B 1156 (2020) 122296 C. Delerue-Matos, QuEChERS and soil analysis, An Overview., Sample Prep. 1 by liquid chromatography-high-resolution mass spectrometry, J. Agric. Food Chem. (2014), https://doi.org/10.2478/sampre-2013-0006. 63 (2015) 5169–5177, https://doi.org/10.1021/jf505168v. [40] M.Á. González-Curbelo, B. Socas-Rodríguez, A.V. Herrera-Herrera, J. González- [57] B. Dejaegher, Y. Vander Heyden, Ruggedness and robustness testing, J. Chromatogr. Sálamo, J. Hernández-Borges, M.Á. Rodríguez-Delgado, Evolution and applications A 1158 (2007) 138–157, https://doi.org/10.1016/j.chroma.2007.02.086. of the QuEChERS method, TrAC – Trends Anal. Chem. 71 (2015) 169–185, https:// [58] S.L.C. Ferreira, A.O. Caires, T.da S. Borges, A.M.D.S. Lima, L.O.B. Silva, W.N.L. dos doi.org/10.1016/j.trac.2015.04.012. Santos, Robustness evaluation in analytical methods optimized using experimental [41] AOAC, AOAC Official Method 2007.01 Pesticide Residues in Foods by Acetonitrile designs, Microchem. J. 131 (2017) 163–169, https://doi.org/10.1016/j.microc. Extraction and Partitioning with Magnesium Sulfate, 2007. 2016.12.004. [42] S.J. Lehotay, K.A. Son, H. Kwon, U. Koesukwiwat, W. Fu, K. Mastovska, E. Hoh, [59] P.K. Sahu, N.R. Ramisetti, T. Cecchi, S. Swain, C.S. Patro, J. Panda, An overview of N. Leepipatpiboon, Comparison of QuEChERS sample preparation methods for the experimental designs in HPLC method development and validation, J. Pharm. analysis of pesticide residues in fruits and vegetables, J. Chromatogr. A 1217 (2010) Biomed. Anal. 147 (2018) 590–611, https://doi.org/10.1016/j.jpba.2017.05.006. 2548–2560, https://doi.org/10.1016/j.chroma.2010.01.044. [60] P. Stefanelli, T. Generali, S. Girolimetti, D.A. Barbini, Internal quality control as a [43] B. Gevao, K.T. Semple, K.C. Jones, Bound pesticide residues in soils: a review, tool for planning a robustness study regarding a multiresidue method for pesticides Environ. Pollut. 108 (2000) 3–14, https://doi.org/10.1016/S0269-7491(99) found in olive oil, Accred. Qual. Assur. 18 (2013) 313–322, https://doi.org/10. 00197-9. 1007/s00769-013-0986-7. [44] S.J. Lehotay, A. De Kok, M. Hiemstra, P. Van Bodegraven, Validation of a fast and [61] Y. Vander Heyden, A. Nijhuis, J. Smeyers-Verbeke, B.G. Vandeginste, D. Massart, easy method for the determination of residues from 229 pesticides in fruits and Guidance for robustness/ruggedness tests in method validation, J. Pharm. Biomed. vegetables using gas and liquid chromatography and mass spectrometric detection, Anal. 24 (2001) 723–753, https://doi.org/10.1016/S0731-7085(00)00529-X. J. AOAC Int. 88 (2005) 595–614, https://doi.org/10.1093/jaoac/88.2.595. [62] M.L. Oca, L. Rubio, M.C. Ortiz, L.A. Sarabia, I. García, Robustness testing in the [45] K. Xia, J. Atkins, C. Foster, K. Armbrust, Analysis of Cyromazine in Poultry Feed determination of seven drugs in animal muscle by liquid chromatography–tandem Using the QuEChERS Method Coupled with LC-MS/MS †, J. Agric. Food Chem. 58 mass spectrometry, Chemom. Intell. Lab. Syst. 151 (2016) 172–180, https://doi. (2010) 5945–5949, https://doi.org/10.1021/jf9034282. org/10.1016/j.chemolab.2015.12.019. [46] R.P. Carneiro, F.A.S. Oliveira, F.D. Madureira, G. Silva, W.R. de Souza, R.P. Lopes, [63] F. Leonardi, M. Veschetti, S. Tonnarini, F. Cardellini, R. Trevisi, A step towards Development and method validation for determination of 128 pesticides in bananas accreditation: A robustness test of etching process, Appl. Radiat. Isot. 102 (2015) by modified QuEChERS and UHPLC–MS/MS analysis, Food Control 33 (2013) 93–97, https://doi.org/10.1016/j.apradiso.2015.05.002. 413–423, https://doi.org/10.1016/j.foodcont.2013.02.027. [64] A.W. Garrison, J.K. Avants, W.J. Jones, Microbial Transformation of Triadimefon to [47] L. Han, Y. Sapozhnikova, S.J. Lehotay, Method validation for 243 pesticides and Triadimenol in Soils: Selective Production Rates of Triadimenol Stereoisomers environmental contaminants in meats and poultry by tandem mass spectrometry Affect Exposure and Risk, Environ. Sci. Technol. 45 (2011) 2186–2193, https://doi. coupled to low-pressure gas chromatography and ultrahigh-performance liquid org/10.1021/es103430s. chromatography, Food Control 66 (2016) 270–282, https://doi.org/10.1016/j. [65] M. Masís-Mora, V. Lizano-Fallas, G. Tortella, W. Beita-Sandí, C.E. Rodríguez- foodcont.2016.02.019. Rodríguez, Removal of triazines, triazoles and organophophates in biomixtures and [48] V. Kumar, N. Upadhyay, V. Kumar, S. Sharma, A review on sample preparation and application of a biopurification system for the treatment of laboratory wastewaters, chromatographic determination of acephate and methamidophos in different sam- Chemosphere 233 (2019) 733–743, https://doi.org/10.1016/j.chemosphere.2019. ples, Arab. J. Chem. 8 (2015) 624–631, https://doi.org/10.1016/j.arabjc.2014.12. 06.001. 007. [66] N. Singh, Factors Affecting Triadimefon Degradation in Soils, J. Agric. Food Chem. [49] S. Saito-Shida, S. Nemoto, R. Matsuda, Multiresidue analysis of pesticides in ve- 53 (2005) 70–75, https://doi.org/10.1021/jf048884j. getables and fruits by supercritical fluid extraction and liquid chromatography- [67] A.H.B. Deas, D.R. Clifford, Reductive and oxidative metabolism of triazo- tandem mass spectrometry, J. Food Hyg. Soc. Japan. 55 (2014) 142–151, https:// lylmethanes by two basidiomycete fungi, Pestic. Biochem. Physiol. 22 (1984) doi.org/10.3358/shokueishi.55.142. 276–284, https://doi.org/10.1016/0048-3575(84)90020-8. [50] C. Ghosh, C.P. Shinde, B.S. Chakraborty, Influence of ionization source design on [68] G.R. Tortella, E. Salgado, S.A. Cuozzo, R.A. Mella-Herrera, L. Parra, M.C. Diez, matrix effects during LC-ESI-MS/MS analysis, J. Chromatogr. B Anal. Technol. O. Rubilar, Combined microbiological test to assess changes in an organic matrix Biomed, Life Sci. 893–894 (2012) 193–200, https://doi.org/10.1016/j.jchromb. used to avoid agricultural soil contamination, exposed to an insecticide, J. Soil Sci. 2012.03.012. Plant Nutr. 14 (2014) 869–880, https://doi.org/10.4067/s0718- [51] F. Gosetti, E. Mazzucco, D. Zampieri, M.C. Gennaro, Signal suppression/enhance- 95162014005000069. ment in high-performance liquid chromatography tandem mass spectrometry, J. [69] K.A. Lewis, J. Tzilivakis, D.J. Warner, A. Green, An international database for Chromatogr. A 1217 (2010) 3929–3937, https://doi.org/10.1016/j.chroma.2009. pesticide risk assessments and management, Hum. Ecol. Risk Assess. An Int. J. 22 11.060. (2016) 1050–1064, https://doi.org/10.1080/10807039.2015.1133242. [52] P.J. Taylor, Matrix effects: the Achilles heel of quantitative high-performance liquid [70] K. Kravvariti, N.G. Tsiropoulos, D.G. Karpouzas, Degradation and adsorption of chromatography–electrospray–tandem mass spectrometry, Clin. Biochem. 38 terbuthylazine and chlorpyrifos in biobed biomixtures from composted cotton crop (2005) 328–334, https://doi.org/10.1016/j.clinbiochem.2004.11.007. residues, Pest Manag. Sci. 66 (2010) 1122–1128, https://doi.org/10.1002/ps.1990. [53] S. Chawla, H.K. Patel, H.N. Gor, K.M. Vaghela, P.P. Solanki, P.G. Shah, Evaluation [71] G.R. Tortella, O. Rubilar, M.d.P. Castillo, M. Cea, R. Mella-Herrera, M.C. Diez, of Matrix Effects in Multiresidue Analysis of Pesticide Residues in Vegetables and Chlorpyrifos degradation in a biomixture of biobed at different maturity stages, Spices by LC-MS/MS, J. AOAC Int. 100 (2017) 616–623, https://doi.org/10.5740/ Chemosphere. 88 (2012) 224–228, https://doi.org/10.1016/j.chemosphere.2012. jaoacint.17-0048. 02.072. [54] J Hajšlová, J. Zrostlıḱová, Matrix effects in (ultra)trace analysis of pesticide residues [72] C. Vischetti, E. Monaci, A. Cardinali, C. Casucci, P. Perucci, The effect of initial in food and biotic matrices, J. Chromatogr. A 1000 (2003) 181–197, https://doi. concentration, co-application and repeated applications on pesticide degradation in org/10.1016/S0021-9673(03)00539-9. a biobed mixture, Chemosphere 72 (2008) 1739–1743, https://doi.org/10.1016/j. [55] A. Kruve, A. Künnapas, K. Herodes, I. Leito, Matrix effects in pesticide multi-residue chemosphere.2008.04.065. analysis by liquid chromatography–mass spectrometry, J. Chromatogr. A 1187 [73] A. Huete-Soto, H. Castillo-González, M. Masís-Mora, J.S. Chin-Pampillo, (2008) 58–66, https://doi.org/10.1016/j.chroma.2008.01.077. C.E. Rodríguez-Rodríguez, Effects of oxytetracycline on the performance and ac- [56] P. Yang, J.S. Chang, J.W. Wong, K. Zhang, A.J. Krynitsky, M. Bromirski, J. Wang, tivity of biomixtures: Removal of herbicides and mineralization of chlorpyrifos, J. Effect of sample dilution on matrix effects in pesticide analysis of several matrices Hazard. Mater. 321 (2017) 1–8, https://doi.org/10.1016/j.jhazmat.2016.08.078. 13