Carbon fibers (12 μm diameter, 5 mm length) were purchased from Cytec Engineered Materials (West Paterson, NJ, USA) and used for carbon fiber microelectrodes elaboration.

Standards of fenitrothion (97%), 3-méthyl-4-nitrophenol (98%), Potassium ferrocyanide (K₄Fe(CN)₆), and K₂HPO₄ (99%), sulfuric acid (98%), Sodium hydroxide (98%), were purchased from Sigma-Aldrich. KH₂PO₄ (98%) was purchased from FLUKA.

Absolute ethanol (99.9%) and acetonitrile (99.9%) were purchased from AnalaR NORMAPUR.

All solutions were prepared with deionized water (pH 7.2, conductivity < 0.1 μS/cm and DOC < 0.1 mg/L).

A helical steel auger with screw-on tools was used for soil sampling.

A RETSCH brand analytical sieve with a 75 µm mesh, certified to ISO 3310/1, is used for sieving soil and sediment samples. An OHAUS brand analytical balance with a precision of 10⁻⁴ g was used for weighing samples; A CU-5000 IEC brand centrifuge was used to separate the liquid and solid phases after extraction; A BüCHI brand R-200; B-490 rotary evaporator is used for concentrating extracts. A PalmSens potentiostat (PalmSens 3 Instrument, Netherlands) was used for electrochemical analysis by using carbon fiber microelectrodes as working electrodes.

An Agilent Technologies chromatograph equipped with an Agilent J&W HP-5 column and Nitrogen Phosphorus Detector was used to confirm the concentration of FNT in soil.

A CS744 Analyzer, LECO brand was used to perform TOC analysis in soils.

The Sahel region of Burkina Faso is located between the 200 and 600 mm isohyets and has two seasons [24]: a rainy season from June-July to September-October and a dry season from October to May. The Sahel zone is characterized by a highly variable tropical climate in space and time with high temperatures between 33˚C and 36˚C throughout the year. Since warmer air can hold more water vapour, evapotranspiration in Sahel is high and affects the water supply and the amount of surface water available for agriculture [25].

All selected sampling sites were treated with FNT formulations during the previous agricultural season.

Sampling sites are shown in Figure 2.

Figure 2. Sampling sites location.

Fenical 450 UL (FNT in UL formulation) was applied using a MicroUlva device. At each treated site, the insecticide was applied in one day. With a treatment of one (1) liter per hectare, the actual dose of 450 g of FNT per hectare was applied in coverage following lines spaced 7 meters apart at a forward speed of 1 m/s in accordance with FAO guidelines for the control of the desert locust [26].

Soil and sediment samples were collected at T0 (immediately before pesticide application), at T1 (1 hour after pesticide application), and then at T7, T15, and T90 seven, fifteen, and ninety days after pesticide application, respectively. The samples were taken from a depth of 0 - 20 cm using an auger and following the method described by Mathieu and Pieltain [27] (Figure 3).

Sampling at each site was carried out in the morning.

Due to practical difficulties encountered related to the wet conditions at the samples, a mixture of 13 subsamples weighing approximately two (2) kg was immediately collected from the sites, placed in a labeled bag and stored in a cooler before being transported to the laboratory. After drying and sieving the mixture of 13 subsamples, six (6) composite samples are taken using a riffle splitter for analysis, including three (3) by SWV and three (3) for confirmation by GC-NPD.

Samples were dried under ventilation and away from sunlight for 48 hours then sieved to 75 microns and pulps stored by freezing to –20˚C [28].

Figure 3. Sampling points repartition.

Samples were immediately taken before the insecticide was applied in order to obtain control samples (T0) that provide information on the initial content of FNT residues or its main metabolite in the soil. These control samples were used for the physico-chemical characterization of the soils.

To consider the temporal evolution of pesticide residues, soils and sediments samples were taken at each site at the following times:

- T1, one hour after pesticides application;

- T7, one week after pesticides application;

- T15, two weeks after pesticides application;

- T90, three months after pesticides application, to verify whether FNT has completely dissipated in the sites.

Samples sites location and nature of matrix sampled are given in Table 1.

Table 1. Nature of the matrix sampled by sampling location.

The samples taken from each site prior to treatment were subjected to particle size analysis. The method used was that described by Devis and Freitas (1984) [29]. One hundred grams of the sample were dried and separated by vibration on a series of superimposed sieves with mesh sizes ranging from 2 µm to 2 mm. The contents of each sieve were then weighed and the fraction of the sample collected was reported as a percentage of the total sample quantity.

Total Organic Carbon (TOC) analysis in soil was performed using digestion and combustion method.

A 20 g of soil sample with a predetermined moisture content was placed in a 250 mL Erlenmeyer flask. After adding 20 mL of acetonitrile, the mixture was stirred for one hour before being centrifuged at 500 rpm for 15 min. At the end of centrifugation, two phases separate: the solid phase consisting of the soil and a liquid phase or aliquot, represented by the extracts. The aliquot was collected, filtered and evaporated to dryness in a rotary evaporator at approximately 50˚C. The dried residues are then solubilized with a minimum amount of absolute ethanol and transferred to a 50 mL flask. The flask is then filled to volume with phosphate buffer at pH 7.2.

The solution thus obtained was then transferred to the electrochemical cell for analysis.

The extraction involved samples from each site, replicated twice to assess the variability in the analytical data.

Stock solutions (1 g·L⁻¹) of FNT and MNP were prepared by dissolving the equivalent mass of FNT and MNP in absolute ethanol. Working solutions 10 mg/L of FNT and MNP are then prepared. 0.1 M of PBS (pH 6.2) was used as supporting electrolyte.

All electrochemical experiments were performed in square wave voltammetry (SWV). Measurements were performed using an electrochemical analyzer PalmSens (PalmSens Instrument, Netherlands) connected to a personal computer using Ivium PC and PSLite software. A three-electrode configuration was employed consisting of a carbon fiber microelectrode as working electrode (CFME) (diameter Φ = 12 μm, an SCE electrode as the reference electrode, and a platinum wire as the auxiliary electrode. The carbon electrode surface was renewed by a homemade electrochemical treatment. Indeed, the carbon fiber microelectrodes were first pretreated electrochemically in a mixture of sulphuric acid H₂SO₄ (0.5 M)/Ethanol (50/50 w/w) followed by a treatment in 0.1 M Phosphate Buffered Saline (PBS) pH 6.2 using following conditions: potential scanning rate: 100 mV·s⁻¹ in the potential range −1.3 to 1.3 V/SCE during 20 cycles. A quality control of the CFME cleanness with visual and electrochemical tests was used. Electrochemical experiments were carried out in a 50 mL glass voltammetric cell at room temperature.

All measurements were carried out under ambient conditions. The appropriate solutions were transferred into the electrochemical cell. The optimization parameters are the following: Scanning was performed from −1.8 to +1 V versus SCE with a step potential of 10 mV, amplitude of 60 mV, and a frequency of 60 Hz.

Before each experiment, the solutions were deaerated by bubbling nitrogen, and the electrochemical cell was kept under a nitrogen atmosphere throughout the experiments.

Analytical curves were carried out in 0.1 M PBS at pH 7.2 and at variable concentrations of FNT. using the standard addition method.

For example, an appropriate volume of the FNT standard solution was added to the PBS. The mixture was then simultaneously stirred and bubbled for ten (10) minutes. The SWV voltammogram corresponding to the electrochemical response of FNT is then recorded ten seconds after having stopped the agitation and the bubbling of the solution.

Voltametric measurements were carried out using a PalmSens Instrument PC233 Potentiostat(Netherlands) controlled using PalmSensVs1.60 software. Data management was controlled by PSLITE 1.7.3 software.

The results of characteristic parameters of tudied soils are summarized in Table 2, showing that the studied soils are acidic. Indeed, the ferruginous soils found in Sikré are poor, with a very acidic pH (4.7), and low levels of nutrients and exchangeable bases [30].

As for the sandy-loam soils found at Arbinda, they generally have a slightly acidic to neutral pH, often optimal between 5.5 and 7.0. The sandy-loam soils found at Gaïk N’Goata are known for their significantly lower pH and higher nutrient content [31].

The acidic nature of all these soils could accelerate the acid hydrolysis of the FNT [8]-[10].

Table 2. Characteristics of soils and sediments samples.

The calibration curve was established by Square wave voltammetry using carbon fiber microelectrode as working electrode in PBS 0.1 M pH 7.2 at frequency = 60 Hz, amplitude 60 mV, increments 10 mV and the reference electrode is ECS. The carbon fiber microelectrode is poised at a potential of −1.8 V to +1 V vers ECS and the FNT concentration varies from 10 - 200 µg/L. The obtained result is shown in Figure 4.

Figure 4. SWV voltammograms of calibration curve, FNT concentration range: 10 - 200 μg/L.

It appears clearly that this calibration curve varied linearly upon FNT concentration increase with linear regression model, demonstrated excellent linearity across the 10 - 200 µg/L and was represented by the linear equation: Ip (µA) = 0.0022C (μg·mL⁻¹) + 0.0065 with R² = 0.9997 (Figure 4).

The assessment of calibration function showed that the optimal line of best fit minimizes the sum of squared residuals, indicating that at least one predictor variable has a non-zero coefficient.

That assumes that linear regression is suitable model.

The validation of the reported electrochemical method for the determination of FNT in soil was conducted comprehensively in accordance with EURACHEM/CITAC guidelines and ISO/IEC 17025 requirements, ensuring its suitability for quantitative analysis. The method underwent rigorous evaluation for key performance parameters, including ruggedness (Table 3 and Table 4), specificity/selectivity (Table 5), matrix effects (Figure 5 and Table 4), linearity, accuracy (Table 6), precision (Table 7), limits of detection (LOD) and quantification (LOQ) (Table 8), and measurement uncertainty (Table 9).

Specificity of the method was evaluated by assessing potential interferents. As shown on Table 3, only MNP showed significant interference with the signal of FNT.

The interference of MNP on FNT response could be explained by the fact that the nitro group of MNP is reduced at −1.08 V (vs. ECS), close to the reduction potential of the nitro group of FNT (−1.02 V). This interference of MNP on FNT response has already been reported [32][33].

Table 3. Parameters evaluated in the ruggedness test.

Table 4. Evaluation of results from a ruggedness study of the analytical process.

Table 5. Effect of interferents on detection of FNT.

Table 6. Statistical assessment of method accuracy.

Matrix effects were ruled out through the analysis of multiple blank soil samples (Figure 5), supported by consistent recovery rates (89.12% to 94.76%) (Table 4).

Figure 5. SWV voltammogram of soil blank sample; pH = 7.2; frequency = 10 Hz, amplitude 100 mV, increments 20 mV, Reference electrode: ECS.

The accuracy of the method was established through recovery studies at multiple spiked levels, yielding recoveries between 89.12% and 94.76% with low relative standard deviations (Table 6).

Recovery (%) was calculated using the following formula:

FNT residue levels in spiked soil were assessed using the following formula:

C_soil: concentration of FNT residues in soil;

C_cell (µg/L): concentration of residues in the cell, evaluated using a calibration curve;

C_theor: The theoretical value expected from spiked soil.

m_soil: mass of soil used for the analysis.

V_stand: volume of the FNT standard solution used to spike the soil.

Table 7. Statistical analysis of precision data.

Table 8. Experimental data for LOD and LOQ assessment.

Table 9. Sources of uncertainties and standard uncertainty estimate.

Samples were analyzed by means of calibration curve presented in Figure 4.

The results of the electrochemical analysis are presented in Figure 5 and Figure 6.

Figure 6. Temporal variation of FNT residues content in soils at Sikré and Arbinda.

Figure 7. Temporal variation of FNT residues content in sediments at Sikré and Gaik N’goata.

The FNT residues content found in soil and sediment samples, are respectively illustred in Figure 6 and Figure 7 as histograms representing, temporal evolution of FT residues content.

These histograms show that all soil and sediment samples collected before and after treatment (T0, T1, T7, T15, and T90) contain electroactive residues that are reduced on the surface of the carbon fiber microelectrode. The highest concentrations of these residues were recorded in both soil and sediment samples collected from all sites at T1 (one hour after treatment).

Furthermore, the spatiotemporal evolution of residues in the soil is similar to that in the sediments. Specifically, an increase in residues was observed when moving from T0 (taken before spraying) to T1, and a decrease in content is observed from T1 to T90.

Moreover, the concentration levels of residues observed in samples taken at T90 suggest that the residues found in samples taken at T0 are not FNT residues. In other words, the residues found in the samples taken at T0 and T90 are electroactive residues that interfere with the FNT residues on the surface of the carbon fiber microelectrode.

Indeed, the FNT molecule is hydrolyzable, photodegradable, and biodegradable [4][5]. The half-life of FNT reported in the literature is estimated to be between 4.4 and 53.7 days [8]. Under Sahelian conditions, the half-life of FNT has been estimated to be 24 hours [10], 36 hours [34], and 6 days [35]. The sites were treated in October, at the end of the rainy season in the Sahel. At the time of sampling, the humidity and light conditions were ideal for accelerating the volatilization as well as biotic and abiotic degradation of FNT in the Sahel region.

Table 10. Results of the comparative analysis of FNT by GC-NPD.

Considering the slight difference in redox potential of the nitro group of MNP compared to that of FNT, the interference of MNP to FNT response (Table 5), could be responsible of cathodic peak observed in the voltammograms of samples taken at T0 and T90.

To confirm this assumption, soil samples taken at T0, T1 and T90 were analyzed by GC-NPD (Table 10).

The results of the chromatographic analysis (Table 10) confirmed that the samples taken at T0 and T90 do not contain any FNT residues. However, in the sample taken at T1, the FNT residue content was estimated to 312 µg/kg. These results differ from those obtained by the electrochemical method.

Given the slight variation in residue levels observed at times T0 and T90 in all matrices, on the one hand, and the results of the GC-NPD confirmation, on the other, a correction was applied to the results of the samples collected at times T0, T1, and T90, according to the formula below.

where:

C_corrected (t): Corrected concentration at t;

C_SWV (t): Concentration at t found by SWV;

C_SWV (t₀): Concentration at t₀ found by SWV.

This correction which applies to all matrices, accounts for the highest background interference, considers the small variation in the concentration of interferents over time.

The corrected results show that the residues content in the samples taken at T0 and T90 is less than <6.56 µg/kg, which is the detection limit of our method. Furthermore, in the sample taken at T1, the residue content is approximately 346.5 µg/kg. This content is within the same range as that obtained by the chromatographic method (312 µg/kg).

These results confirm that the residues found in samples taken at T0 and T90 are not FNT residues. Although FNT is weakly volatile, studies reported in the literature [36][37] show that volatilization is one of the main pathways for the dissipation of FNT residues in the Sahel. Low in arid environments, it is thought to be particularly significant during the hot and humid period. In contrast, MNP, formed under aerobic conditions, is known for its chemical stability [38], its low mobility [39] in soil, and its low volatility [38] under natural conditions. It degrades poorly in arid environments [38][40]. This physico-chemical property of MNP could explain its persistence in our sampling area, which is indeed an arid area.