Analysis of Pesticides by Ultra-High Performance Liquid Chromatography in Combination with High Resolution Mass Spectrometry (UPLC-Orbitrap MS): Case Study for Application on River Water Samples in Vietnam

Abstract

This study introduced mass spectrometry (MS)-based method for analysis of 118 pesticides after applying solid phase extraction (SPE) for 26 water samples collected in Red River, Northern Vietnam. The excellent limit of quantitation of the developed method ranged from 0.0005 to 0.015 µg L^-1. Intraand inter-day repeatability of the developed method was achieved below 15%. Specific procedure of the liquid chromatography Orbitrap mass spectrometry (LC-Orbitrap MS)-based analytical method for target compounds was identified by retention time, precursor ion, product ions, accurate mass of precursor and product ions and fragmentation pathway in high resolution mass spectrometry. Mass accuracy of all precursor ions and product ions were achieved below 5 ppm in this work. It demonstrated that the concentration of most studied analytes was below the allowable level according to National Technical Regulation in Vietnam (QCVN 08:2023/BTNMT). Alachlor, fenobucarb and acetochlor were often determined at levels in the range between 0.099-8.978, 0.058 18.776, and 0.073 1.460 µg L^-1, respectively. Their concentrations were much higher than regulated pesticides mentioned in QCVN 08:2023/BTNMT (0.02 to 1.0 µg L^-1 for chlorinated pesticides). Therefore, it is necessary to propose a national regulation regarding the levels of the three analytes above to control pesticide pollution in Vietnam.

Keywords:
ultra-high performance liquid chromatography; Q Exactive Orbitrap MS; solid phase extraction; pesticide analysis; river water

Introduction

Pesticides plays an important role in improving the quality of agriculture.¹-³ They are used for the improvement of harvest efficiency and protected plants from insects. Because of their widespread use in agriculture, their occurrence in environment such as ambient air, water and soil over time has been considered.⁴,⁵ Based on their chemical structures, they can be classified into many groups: organochlorine, organophosphate, carbamate, neonicotinoid, strobilurin, and triazole.⁶,⁷ However, they have also been known as toxic persistent organic pollutants (POPs) for environment and human health. Especially, a large number of pesticides are known for their high solubility in water, therefore, their residues in river waters at high levels have the risk of developing diseases in humans.⁷-⁹ Currently, pesticide usage and its residues are subject of strictly surveillance analysis in numerous countries. According to some studies, the maximum residue limit (MRL) of individual pesticide as safety level for environment is set to 0.1 µg L^-1.⁴,⁷,⁹ In Vietnam, according to circular No. 03/2018/TT-BNNPTNT,¹⁰ only 30 compounds in many thousands of multi-class pesticides in the mentioned list were forbidden to use in agricultural activities currently.

The studies on analysis of pesticides in environmental matrices have been performed for many years, with instrumental based methods such as gas/liquid chromatography-mass spectrometry (GC-MS,¹¹ LC MS),¹²,¹³ gas/liquid chromatography tandem mass spectrometry (GC-MS/MS, LC-MS/MS),¹⁴,¹⁵ gas/liquid chromatography in combination with high resolution mass spectrometry, for example GC-TOF-MS, UPLC QTOF MS,¹⁶-¹⁸ LC-Orbitrap MS.⁷ Among these methods, the liquid chromatography in combination with high resolution mass spectrometry (HRMS) for determination of multi-class pesticides has been introduced and many studies were published based on excellent mass accuracy, selectivity and sensitivity.¹⁹,²⁰ Especially, one of the main advantages of HRMS instrument is their ability to screen and quantify a number of compounds in full scan mode or combination between full scan and data dependent MS/MS mode.²¹,²² However, there are not so many reports involving the analysis of multi-residue pesticides in surface water by LC-HRMS, especially in Vietnam to the best of our knowledge. For enhancement of sensitivity and selectivity, several techniques for sample preparation and clean up procedure such as liquid-liquid extraction, solid phase extraction (SPE) have been applied prior to analysis by GC/LC-MS methods in the context of targeted and non-targeted analysis.²³

The aim of this work was to develop an analytical method for screening and quantification of 118 pesticides belonging to different groups present in surface water samples by using an ultra-high performance liquid chromatography in combination with a high-resolution mass spectrometry (UPLC-Orbitrap MS).

Experimental

Multi-class pesticides in the river water samples were enriched by off-line SPE extraction. Target analytes were separated on pentafluoro phenyl propyl (PFP) stationary phase and detected by high resolution mass spectrometry (Orbitrap MS) in both full-scan and data dependent MS² mode. The confirmation of the presence of the targeted compound in the samples was performed by retention time, accurate mass of both precursor ion and product ions and fragmentation pathway employing data dependent MS² measurement (dd-MS²). By conducting spiked experiments in the pool samples, which were preparared by mixture of equal volume of all collected river water samples, the analytical method parameters such as limit of detection (LOD), limit of quantitation (LOQ), linearity and recovery have been calculated and presented for validation of the analytical method. In addition, an isotope-labeled internal standard was used to control the losing of analytes during sample preparation and other effects from LC-Orbitrap MS measurement such as ionization efficiency in the ionization source. Matrix effect (ME), a critical parameter in the mass spectrometry with heated electrospray ionization, has also been investigated by using preand post-extraction spiking experiments. Finally, with the optimized and in-house validated UPLC-HRMS based method, a total of 26 surface river water samples collected in 2023 along to the Red River in the North of Vietnam were applied for analysis of screening and quantification in this study.

Chemicals and reagents

One hundred and eighteen pesticides (high purity, > 98%) in different classes were purchased individually from Sigma-Aldrich (Singapore). The general information of pesticide including chemical formula, chemical abstract service (CAS) number and molecular weight are listed in the Table S1 in Supplementary Information (SI) section. The standard stock solution mixture (5 μg L^-1) was prepared by mixing and diluting individual stock solutions (1000 μg L^-1). The solution of isotope-labeled internal standard (IS) including dimethoate-d₆, dichlorvos-d₆ (Dr Ehrenstorfer, Germany) and malathion-d₁₀ (Toronto Research Chemicals, Canada) was prepared at a concentration of 10 µg L^-1. The eight working standard solutions at concentrations of 2, 5, 10, 25, 50, 100, 250, 500 µg L^-1 containing 250 µg L^-1 of isotope-labeled internal standard were prepared in the mobile phase solution (mixture of methanol and deionized water containing formic acid and 5 mM of ammonium acetate).

Solid phase extraction cartridge, Bond Elut Plexa (200 mg, 3 mL) was purchased from Agilent Technologies (Singapore). All solvents and chemicals such as methanol (MeOH, LC-MS grade), ammonium formate (HCOONH₄, LC-MS grade) and formic acid (FA, LC-MS grade) were purchased from Fisher Scientific (USA). LC-grade water (type III) was obtained by purifying deionized water in a Milli-Q Integral 3 system (Merck Millipore, France).

Sample preparation

Sample collection

Twenty-six water samples were collected along the Red River in the North of Vietnam through Hung Yen, Ha Nam, Vinh Phuc, Nam Dinh, Thai Binh, Yen Bai, Lao Cai and Phu Tho provinces and Hanoi city, as shown in Figure 1. River water samples were collected following the methodology described by US EPA.²⁴ Generally, suface water samples were collected by using 2-L horizontal water sampling bottle in each sampling location belonging from upstream to downstream of the Red River. The river water samples were collected in 2023 and they were coded and stored in amber glass bottles at -20 °C until the analysis.

Figure 1
Map of the sampling position for river water samples along the Red River (a) LY1-LY8 and (b) TN19-TN36.

Sample preparation for UPLC-HRMS measurement

A Bond Elut Plexa (polymeric-based reversed phase absorbent) solid phase extraction cartridge was used to cleanup and enrichment the investigated pesticides from the river water samples. In brief, solid phase extraction cartridge was first conditioned by using 5 mL MeOH then followed by 5 mL deionized water adjusted to pH 3 with concentrated formic acid. A 100 mL of river water sample at pH 3 (adjusted with concentrated formic acid), spiked with 250 µg L^-1 isotope-labeled internal standards (dimethoate-d₆, dichlorvos-d₆ and malathion-d₁₀) which were described in the sub section “Chemicals and reagents”, were loaded on SPE cartridge at a flow rate of approximately 1 drop per second by gravity force (approximately 3 mL min^-1). Chemical interferences were removed by washing with deionized water, pH 3 (adjusted with concentrated formic acid). The SPE cartridge was dried by a gentle nitrogen gas flow. The elution of target analytes on SPE column was performed by 10 mL MeOH and the eluent was concentrated under a gentle nitrogen gas flow at 40 °C to 1.0 mL. Finally, the samples were filtered through Minisart NY25 Syringe filter 17846, 0.45 µm polyamide (Sartorius, Germany), then the clear solution was subjected to analysis by UPLC-Orbitrap MS.

Instrumental method

A UPLC system (Ultimate 3000, Thermo Scientific, Germany) including quaternary pump, a liquid autosampler and a column thermostat chamber coupled with a Q-Exactive Focus Orbitrap mass spectrometer was used. A heated electrospray ionization source was used for analysis of pesticides in this work. All parameters were set and monitored by Thermo XCalibur software version 4.0 (Thermo Scientific, Germany). A Thermo Hypersil GOLD PFP column (150 mm × 2.1 mm, 3 µm, Thermo Scientific, USA) was used for separation of target analytes. The column temperature was maintained at 40 °C during chromatographic separation. The flow rate was constantly kept at 0.3 mL min^-1. The mobile phases were 0.1% FA + 5 mM HCOONH₄ in H₂O (channel A) and 0.1% FA + 5 mM HCOONH₄ in MeOH (channel B). Their gradient elution started with 2% B in 0.25 min, then raised to 30% B in 0.75 min, after that increased linearly to 100% B in 24 min (held for 5 min) then went back to 2% B in 0.5 min and held for 7 min to re-equilibrate the column for the next injection. The total time of chromatographic separation was 37.5 min. An amount of 5 µL of sample/standard solution were injected in the UPLC Orbitrap MS system using a liquid autosampler. The needle and sample loop in the liquid autosampler were washed with a mixture of methanol:deionized water (1:9 v/v).

A Q-Exactive Focus Orbitrap MS using heated electrospray ionization source (HESI) working at 70000 FWHM resolution (at 200 Da) was operated in both positive and negative ionization modes. The operating conditions were optimized by direct infusion experiments and the optimized operating parameters were as follows: sheath gas pressure at 32 psi, auxiliary gas flow rate at 7 L min ¹, sweep gas flow rate at 5.0 L min^-1, spray voltage +2.8 and -2.5 kV for positive and negative ionization mode, respectively. Capillary and vaporizer temperature were at 320 and 295 °C, respectively; S-lens radio frequency (RF) level was set at 50 V. All target ions of the precursors were detected by full-MS mode (70000 FWHM resolution, at 200 Da) in m/z (mass to charge ratio) 80 1200 in profile data acquisition format. The full scan data dependent mode (full-MS/dd MS²) simultaneously measured the fragmentation spectra for their precursors. Moreover, the full-MS/confirmation mode (with inclusion list of targeted compounds) was also used to confirm the fragmentation pathway of the selected precursors. The dd-MS² and confirmation mode conditions were followed by parameters: 17500 FWHM resolution (at 200 Da), mass isolation window m/z 1.0, maximum and minimum automatic gain control (AGC) target 8 × 10⁴ and 5 × 10³, respectively; normalized collision energy (NCE) 30%, spectrum data collection for confirmation was set in centroid format. The detailed information of the UPLC HRMS analysis for pesticides in river water samples were listed in Table S2 in SI section.

Quality control

For the quality control, the Q-Exactive Orbitrap mass spectrometer was calibrated with Pierce positive/negative ion mass calibration solutions (Thermo Fisher, Germany) before each batch measurement. Besides, the stability of analytical signal for internal standards was evaluated via relative standard deviation (RSD) by analysis of both quality control (QC) and spiked pool samples. In addition, all river water samples were spiked with 250 µg L^-1 internal standards (dimethoate-d₆, dichlorvos-d₆ and malathion-d₁₀) before sample treatment to compensate the loosing of the analytes during sample preparation procedure as well as other effects from measurement such as ionization enhancement/suppression. For each batch, the QC sample was first injected in triplicate to check intensity and retention time shift before measurement, according to previous study.²⁵,²⁶ The order of analysis for measuring each batch was performed as: instrument blank (using mobile phase), QC, working standard solutions, and samples. After every 10 samples, QC and blank samples were injected for checking the stability of analytical signal and carry-over effect, respectively. For individual sequences, the number of blank and QC samples accounted for approximately 20% of the total number of injections.

Data evaluation

A TraceFinder software version 3.3 (Thermo Scientific, USA) was used for conversion, data evaluation and final report. The theoretical m/z was calculated using a web-based EnviPAT software²⁷ as accuracy of the targeted pesticides was calculated by equation 1:

(1) $$\text{Mass accuracy}=\frac{m/z_{\text{experimental}}-m/z_{\text{theorical}}}{m/z_{\text{theorical}}}\times {10}^6(\mathrm{ppm})$$

where, m/z_experimental and m/z_theorical were mass-to-charge ratio of targeted pesticide measured by Q-Exactive Focus Orbitrap MS and computed by online EnviPAT software,²⁷ respectively.

The product ions and mechanism of fragmentation pathway of the analytes were conducted with NCE at 30% and compared with commercially available MS/MS spectral libraries like National Institute of Standards and Technology (NIST) and dedicated MassFrontier software version 7.0 (Thermo Scientific, USA) at the same collision energy. The presence of identified target compound was confirmed by retention time on the LC column, accurate mass of precursor and product ions, and fragmentation pathway associated with given measurement uncertainty. Characteristic parameters for the validation of analytical method such as linearity range, limit of detection (LOD), limit of quantitation (LOQ), recovery, and repeatability were investigated and calculated via spiking experiments at low (10 µg L^-1), medium (50 µg L^-1) and high concentration levels (100 µg L^-1) in sample matrix (in the pool sample). Each experiment was conducted in triplicate and mean value association with given standard error was used. LOD and LOQ were calculated by the signal to noise ratio at 3 and 10 times, respectively. In addition, post extraction spiking experiments were conducted to assess the matrix effect (ME) in mass spectrometry in this work. This fact may have caused the increasing or suppression ionization leading to changes in analytical signal robustness in MS measurement, as mentioned in the previous work.²⁸ These values below or higher than 100% mean suppression and enhancement of the analytical signal, respectively.²⁹ The acceptable range of ME is 80-120%, according to some recent publications.³⁰,³¹

Results and Discussion

Chromatographic separation

The chromatographic separation of multi-class pesticides was performed in a Thermo Hypersil GOLD PFP column employing mobile phase containing formic acid and ammonium acetate in deionized water and methanol. The reason for selecting the PFP column for the separation of these compounds was investigated and mentioned in our previous work.²⁸ This LC column showed to be the best choice for the separation of multi-class pesticides based on chromatographic efficiency and peak shape, specially for polar pesticides. In addition, some polar pesticides are retained on this column due to the nature of the stationary phase and physiochemical properties. The total ion chromatogram of all pesticides in positive ionization mode was shown in Figure 2.

Figure 2
The overlay extracted ion chromatogram (EIC of protonated molecular ion) of 118 pesticides on PFP column in positive ionization mode, concentration of all pesticides and internal standard was 200 µg L^-1.

From chromatography point of view, one of the critical parameters in chromatographic separation is the stability of the analyte retention time in the liquid chromatography column, especially polar compounds in reverse phase LC column. For investigation and optimization of this parameter, standard solutions with concentration of all compounds of 50 µg L^-1 in mobile phase and in matrix matched solution, that means that an equal volume of 26 collected water river samples combined into one composited/matrix sample/pool sample, were prepared and injected independently during the day (intra-day, n = 24) and through three measurement days (inter-day, n = 3) on the UPLC-Orbitrap MS at optimized operating conditions. The relative standard deviation of the retention time of each compound was calculated and listed in Table S1 (SI section). As clearly shown in Table S1, the excellent stability of retention time of all selected analytes was achieved. Relative standard deviations of retention times for intraand inter-day measurement of pesticides in the mobile phase solution were below 0.95 and 0.97%, respectively. In addition, the relative standard deviations of retention times for intraand inter-day measurement of pesticides in the matrix-matched solution were below 0.98 and 0.99%, respectively. There was no significant difference in the retention time of the analyte between the mobile phase and matrix-matched solution. It could be concluded that the excellent retention time repeatability of pesticides on the PFP column in both solutions was achieved in this work.

Mass accuracy and fragmentation pathway

Another factor to identify analytes by high resolution mass spectrometry is their mass accuracies of both protonated molecular ion (deprotonated molecular ion in negative ionization mode) and product ions (positive mode). The measurement of accurate mass could be provided by monoisotopic mass based on the elemental composition of analytes and the abundance of their isotopes in nature. In addition, high mass accuracy of the target ion also avoid positive/negative fail in the quantitative analysis due to isobaric mass interferences. In this work, the mass accuracy of all targeted compounds was investigated and assessed by injecting standards at a concentration of 50 μg L^-1 six times independently into the mobile phase and matching matrix in UPLC-Orbitrap MS. The theorical mass-to-charge ratio of the protonated ion of analyte (or deprotonated ion) was calculated online by EnviPAT software.²⁷ The mass accuracy of all target analytes was then calculated by equation 1. The mass accuracy of precursor and product ions of all pesticides in this study using full scan mode on the UPLC Orbitrap MS system are shown in Table S3 (SI section). It should also be noted that the mass accuracy of precursor ion of all pesticides in both standard solution and matrix-matched solution were lower than 3.36 and 3.44 ppm, respectively. The mass accuracies of all product ions were below 4.9 ppm except for product ion of acetochlor (mass accuracy of product ion of this compound was -5.40 ppm). It is clear that these parameters such as the mass accuracy of the precursor ion as well as the production of all pesticides were lower than 5.0 ppm, which is acceptance criteria for confirming the identity of chemical residues using extract mass data according United States Food and Drug Administration (US FDA) office of Food and Veterinary Medicines²⁹ and European Union Offices: Commission Decision (CD) 2002/657/EC.³⁰ For all analytes, the mass accuracy of the precursor ion was always higher than the product ions. It could be explained by resolution settings in the full scan and dd-MS² modes were 70000 and 17500 FWHM resolution, respectively. However, the identification of target analytes was carried out by comparison of retention time, mass accuracies, and fragments with their corresponding standards at certain experimental conditions. Fragments in these measurements were performed in dd-MS² and confirmed in confirmation modes⁷ by comparison with spectra from Thermo MassFrontier software version 7.0. Table S3 (SI section) shows the accurate mass of product ions of all analyzed pesticides.

The extracted ion chromatogram of fenobucarb and MS² confirmation mass spectrum at 30% NCE are shown in Figure 3. As can be seen in Figure 3, protonated molecular ion and two fragment ions were chosen to confirm and quantify the presence of fenobucarb in the sample beside retention time. The isotopic peak of the protonated molecular ion of the fenobucarb was also observed in the full scan mode. The mass-to-charge ratios of the isotopic peaks of protonated molecular ion of fenobucarb in standard solution and real sample matrix were 208.1330 and 208.1331 Da, respectively. The theoretical mass to charge ratio of this compound is 208.1332 Da. The relative intensities of the isotopic peak compared to monoisotopic peak of the protonated molecular ion of fenobucarb in standard solution and real sample matrix were 12.3 and 11.7%. The absolute bias of two values was -0.6%. The theoretical value of relative abundance of this ion to the protonated molecular ion is 13.33%. Both mass to charge ratio and relative intensities of two product ions of the protonated molecular of fenobucarb were perfectly fit with theoretical calculated using online EnviPAT software.²⁷ The mass accuracy of the precursor ion of all analytes was better than product ions as can be seen in Table S3 (SI section).

Figure 3
Extracted ion chromatogram of fenobucarb and MS² confirmation mass spectrum of precursor at 30% NCE at retention time 6.70-6.75 min: (a) in the standard solution sample with m/z 208.1330 Da and two product ions (m/z 95.0497 Da (C₆H₇O⁺) and m/z 57.0705 Da (C₄H₉⁺)) and (b) in the real sample with m/z 208.1331 Da and two product ions (m/z 95.0493 Da (C₆H₇O⁺) and m/z 57.0704 Da (C₄H₉⁺)).

Validation of the analytical method

Linearity range, limit of detection and limit of quantitation

Eight independent standard solutions with concentrations ranging from 2 to 500 µg L^-1 for analytes with 250 µg L^-1 isotope-labeled internal standards were prepared in solvent as the same as mobile phase composition. All standard solutions were injected on the UPLC Orbitrap MS in the optimized operating condition. The peak area of the pesticide and internal standard were integrated, and peak ratio was calculated. The linearity function of the peak ratio and the concentration of pesticides were built and presented in Table 1 for several typical pesticides and Table S4 (SI section) for all investigated pesticides. It could be seen from Tables 1 and S4 that correlation coefficient (R²) between peak ratio and analytical concentration for all targeted analytes were higher than 0.99. These values were completely met to acceptable values for the validation of the analytical method. The limit of detection (LOD) and limit of quantification (LOQ) were investigated by spiking experiments at 10 µg L^-1 of target analytes into matrix-matched solution. After that, the matrix-matched samples were analyzed on the UPLC Orbitrap MS system after sample preparation as mentioned in the sub section “Sample preparation”. Based on US FDA and European Union guidelines,²⁹,³⁰,³²,³³ LOD and LOQ were calculated at S/N = 3 and S/N = 10, respectively, where S/N refers to signal-to-noise ratio in the matrix-matched solution. As clearly shown in Tables 1 and S4, LODs ranged from 0.001 to 0.076 µg L^-1 and LOQs ranged from 0.002 to 0.227 µg L^-1. Therefore, this method achieved an excellent sensitivity for all targeted compounds. Comparing with some recent publications,³⁴,³⁵ the LOD and LOQ of this work were low enough for the analysis of pesticides in the river water samples.

Table 1
Analytical figures of merit of the UPLC-Orbitrap MS for analysis of several typical pesticides in surface water samples

No.	Analyte	Regression equation	R2	LOD / (µg L-1)	LOQ / (µg L-1)
1	acetochlor	y = (0.004 ± 6.04E-5)x - (0.026 ± 2.1314)	0.9987	0.066	0.198
2	alachlor	y = (0.042 ± 0.001)x - (0.267 ± 0.160)	0.9978	0.067	0.200
3	buprofezin	y = (0.028 ± 1.798E-4)x - (0.106 ± 0.036)	0.9997	0.001	0.003
4	butachlor	y = (0.027 ± 2.83E-4)x - (0.130 ± 0.057)	0.9993	0.001	0.003
5	edifenphos	y = (0.025 ± 1.32E-4)x - (0.048 ± 0.027)	0.9998	0.031	0.093
6	fenobucarb	y = (0.037 ± 0.001)x - (0.031 ± 0.059)	0.9959	0.022	0.065
7	hexaconazole	y = (0.004 ± 1.26E-4)x + (0.039 ± 0.025)	0.9931	0.001	0.003
8	mefenacet	y = (0.033 ± 3.46E-4)x - (0.148 ± 0.070)	0.9993	0.052	0.155
9	pirimiphos-methyl	y = (0.025 ± 2.90E-4)x - (0.145 ± 0.059)	0.9991	0.001	0.004
10	propiconazole	y = (0.017 ± 1.35E-4)x - (0.027 ± 0.014)	0.9997	0.001	0.002
11	propisochlor	y = (0.009 ± 5.81E-5)x - (0.006 ± 0.012)	0.9997	0.001	0.003
12	pyrachlostrobin	y = (0.023 ± 7.82E-4)x - (0.267 ± 0.158)	0.9931	0.001	0.004
13	pyrimethanil	y = (0.052 ± 0.001)x - (0.218 ± 0.137)	0.9968	0.001	0.004

Stability of analytical signal

The stability of the analytical signal was evaluated through RSD of the peak areas of the investigated analytes, by analysis of both QC and spiked sample continuously intra-day (n = 24) and inter-day (n = 5). The peak area was integrated using Thermo TraceFinder software at ± 5 ppm mass accuracy window. The RSD of the peak area was calculated and listed in Table S5 (SI section). Intraand inter-day relative standard deviation of the peak areas of all pesticides in the solvent were below 15.4 and 15.9% (exception of ethion, RSD 4.3 and 22.1% for intraand inter-day, respectively). Intraand inter-day relative standard deviation of the peak areas of all pesticides in the matrix-matched solution were below 14.4 and 15.3% (exception of ethion, RSD 5.3 and 24.9%, for intraand inter-day, respectively). The higher RSD of the peak area of ethion could have contributed to the natural chemical properties of this compound. However, it can be concluded that the intraand inter-day stability of analytical signal is good enough for quantification of pesticides in river water samples according to Association of Official Analytical Chemists (AOAC)³⁶ and US FDA-³⁷ guideline to analysis validation.

Recovery and matrix effect (ME)

In this work, the SPE column, Bond Elut Plexa (Agilent Technologies, Singapore), was chosen because it is an universal enrichment sorbent for extraction of organic pollutants and suitable for interaction with analytes of a wide polarity range. In general, the pH of the sample, and the eluent components were optimized, and the optimum experimental conditioning were mentioned in the Experimental section. Matrix effect plays an important role in LC-MS analysis, especially in electrospray ionization (ESI) mode.³⁸ For investigation of the matrix effect, preand post-extraction spiking experiments were conducted in the pool samples at three concentrations of analytes (10, 50 and 100 µg L^-1). The pool samples were prepared as mentioned in Experimental section until final volume of 1 mL, after that, 5 µL of sample solution were injected to UPLC Orbitrap MS system using a liquid autosampler. The peak area of the pesticides was considered for the calculation of matrix effect. The ME was calculated by comparing the peak area of a given pesticide in the post-extraction solution and in the solvent (at the same concentration). From Table S6 (SI section), the matrix effect ranged from 41 ± 8.8 (carboxin) to 206.3 ± 12.1% (haloxyfop-R). As clearly shown in the Figure 4, the sample matrix affected 63% of the target analytes in the total number of analytes with ME values higher than 100%, corresponding to ionization enhancement and 6.8% of the number of compounds with ME lower than 80% association with ionization suppression. Overall recovery was calculated by comparing the peak area of the analyte pre-extraction spiking with the theoretical peak area of the analyte in standard at the same concentrations. As described in the sub section “Quality control”, three internal standards of dimethoate-d₆, dichlorvos-d₆ and malathion-d₁₀ were used in this study. Actually, it would be excellent results if this study applied as much as internal standards which were representative for each pesticides. However, it is not viable to have internal standards of all compounds for quantification. Therefore, the results of the sample matrix effect and recoveries for investigated pesticides in this study will be affected by three applied internal standards, and the analytical results will be more accurate if more internal standards were spiked in analytical procedure, especially for analytes detected in the rivers. However, based on results obtained in this study, overall recoveries of target compounds are shown in Figure 5. Overall recoveries of pesticides at three concentrations (10, 50 and 100 µg L^-1) were in the range of 51.3 ± 10.4 (pyridanyl) and 128.9 ± 11.7% (carbofuran). It is worthy note that in this study, the main contribution in overall recoveries of the pesticides was the matrix effect.

Figure 4
Average matrix effect in Orbitrap MS for analysis of pesticides (a) and (b) in surface water, error bar presents standard deviation (n = 3).

Figure 5
Average overall recoveries of pesticides (a) and (b) in surface water, error bar presents standard deviation (n = 3).

Pesticides in samples

This developed method was applied to analyze twenty six river water samples collected downstream of Red River and passing through nine agriculture provinces and cities along Red River delta, one of the main agriculture production areas of Vietnam, where it was known for using a lot of pesticides for rice cultivation and other production systems.³⁹ Samples were collected in Red River in the northern provinces of Vietnam, such as Lao Cai (n = 4), Yen Bai (n = 2), Phu Tho (n = 2), Vinh Phuc (n = 1), Hung Yen (n = 5), Ha Nam (n = 2), Nam Dinh (n = 6), Thai Binh (n = 2) and Hanoi city (n = 2). This study will preliminary determine the presence of polar wide-range pesticide residues in the water river by applying the validated method. The concentrations of pesticides in 26 river water samples were shown in Figure 6, Tables 2 and 3. In general, most of the analytes were hardly detected in real samples or below the maximum value limit for pesticides permitted by the EU regulation level (< 0.1 µg L^-1),³⁹ as well as below the allowable level according to National Technical Regulation on Surface Water Quality in Vietnam (QCVN 08:2023/BTNMT)⁴⁰ (for concentrations of lindan, aldrin/dieldrin and total dichlorodiphenyltrichloroethane (DDT) ranging between 0.02; 0.2 and 1.0 µg L^-1, respectively). As shown in Tables 2 and 3, most of the analytes were not determined qualitatively and quantitatively in samples, even through, some analytes were detected in the samples lower than LOQ. This could be explained because the preand post-spike concentration of analytes was much higher than those in real samples. Therefore, some target compounds in real samples were not detected and/or detected at lower concentrations, so recovery could decrease. Moreover, the method developed in this study used only three internal standards, which have may caused low recovery for the pesticides which were not detect in the river as described in the sub section “Recovery and matrix effect”. However, about 11.0% of the interested pesticides (13/118 investigated pesticides) described in the Tables 2 and 3 were detected at one or more sample collection sites, and especially, all of them have not been banned from use in Vietnam agriculture nowadays. Among that, butachlor, pretilachlor, and fenobucarb were found in most of the 26 collected river water samples. Fenobucarb and pretilachlor were present in collected river samples with high frequency (> 90%), followed by butachlor (65%) then buprofezin (58%). Meanwhile, the observation of acetochlor, mefenacet, propisochlor were in 30% of the total number of samples as shown in Tables 2 and 3. Figure 6 also showed the average amount of concentration for detected pesticides in 26 river water samples collected at eight provinces and Hanoi city along the Red River. The average amount of total pesticides was calculated based on the number of river water samples collected at each province and city where the river water passed through according to the river stream from North to South Vietnam.

Table 2
Pesticide concentration detected in 18 river water samples collected along Red River (Hanoi, Hung Yen, Ha Nam, Vinh Phuc, Nam Dinh and Thai Binh provinces)

Concentration / (µg L-1)
No	Name	Hanoi			Hung Yen						Ha Nam			Vinh Phuc		Nam Dinh							Thai Binh
		TN19	TN20		TN21	TN22	TN23	TN24	TN26		TN25	TN36		TN27		TN28	TN29	TN30	TN31	TN32	TN33		TN34	TN35
1	acetochlor	-	-		-	-	-	-	1.460 ± 0.073		-	-		-		0.134 ± 0.007	0.073 ± 0.004	-	0.153 ± 0.008	-	-		-	-
2	alachlor	-	-		-	-	-	-	3.18 ± 0.159		-	-		-		4.75 ± 0.237	6.316 ± 0.316	0.101 ± 0.005	4.752 ± 0.238	0.218 ± 0.011	0.118 ± 0.006		0.099 ± 0.005	0.165 ± 0.008
3	buprofezin	-	-		-	0.025 ± 0.001	0.019 ± 0.001	-	0.025 ± 0.001		0.018 ± 0.001	0.017 ± 0.001		-		0.023 ± 0.001	0.019 ± 0.001	-	0.019 ± 0.001	-	-		0.017 ± 0.001	-
4	butachlor	-	0.036 ± 0.002		0.036 ± 0.002	0.038 ± 0.002	0.036 ± 0.002	0.037 ± 0.002	-		0.036 ± 0.002	0.036 ± 0.002		0.037 ± 0.002		-	-	0.036 ± 0.002	-	0.036 ± 0.002	0.036 ± 0.002		0.036 ± 0.002	0.037 ± 0.002
5	edifenphos	-	-		-	-	0.026 ± 0.001	-	-		-	-		-		-	-	-	-	-	-		-	-
6	fenobucarb	-	0.087 ± 0.004		0.064 ± 0.003	0.074 ± 0.004	0.058 ± 0.003	0.058 ± 0.003	0.133 ± 0.007		0.062 ± 0.003	7.178 ± 0.359		0.097 ± 0.005		0.864 ± 0.043	2.587 ± 0.129	0.068 ± 0.003	18.776 ± 0.939	0.086 ± 0.004	0.071 ± 0.001		0.063 ± 0.003	0.09 ± 0.004
7	hexaconazole	-	-		-	-	-	-	-		-	0.038 ± 0.002		-		-	-	-	0.041 ± 0.002	0.038 ± 0.002	-		-	0.038 ± 0.002
8	mefenacet	-	-		-	0.06 ± 0.003	-	-	0.065 ± 0.003		-	0.038 ± 0.002		-		0.063 ± 0.003	0.065 ± 0.003	-	0.074 ± 0.004	-	-		-	-
9	pirimiphos-methyl	-	-		-	-	0.032 ± 0.002	-	-		-	-		-		-	-	-	-	-	-		-	-
10	pretilachlor	0.055 ± 0.003	0.056 ± 0.003		0.055 ± 0.003	0.058 ± 0.003	0.057 ± 0.003	0.059 ± 0.003	1.275 ± 0.064		0.053 ± 0.003	0.073 ± 0.004		0.054 ± 0.003		0.151 ± 0.008	0.09 ± 0.005	0.053 ± 0.003	0.153 ± 0.008	0.054 ± 0.003	0.055 ± 0.001		-	0.061 ± 0.003
11	propisochlor	-	-		-	0.039 ± 0.002	-	-	0.047 ± 0.002		-	0.005 ± 0.001		-		0.034 ± 0.002	0.049 ± 0.002	-	0.033 ± 0.002	-	-		-	-
12	pyrachlostrobin	-	-		-	-	0.055 ± 0.003	-	-		-	-		-		-	-	-	-	-	-		-	-
13	pyrimethanil	-	-		-	0.006 ± 0.001	-	-	0.005 ± 0.001		-	-		-		0.003 ± 0.001	-	-	0.012 ± 0.001	-	-		-	-

Table 3
Pesticide concentration detected in 8 river water samples collected along Red River (Lao Cai, Phu Tho, and Yen Bai provinces)

Concentration / (µg L-1)
No	Name	Lao Cai					Phu Tho			Yen Bai
		LY1	LY2	LY3	LY4		LY5	LY6		LY7	LY8
1	acetochlor	-	-	-	-		-	-		-	-
2	alachlor	3.506 ± 0.175	0.194 ± 0.01	0.196 ± 0.01	0.118 ± 0.006		8.978 ± 0.449	0.165 ± 0.008		0.100 ± 0.005	1.166 ± 0.058
3	buprofezin	0.018 ± 0.001	0.017 ± 0.001	0.017 ± 0.001	0.017 ± 0.001		0.018 ± 0.001	0.017 ± 0.001		0.017 ± 0.001	0.018 ± 0.001
4	butachlor	-	-	-	-		-	-		0.035 ± 0.002	0.036 ± 0.002
5	diphenamid	-	-	0.030 ± 0.002	0.030 ± 0.002		0.030 ± 0.002	-		-	-
6	fenobucarb	7.108 ± 0.355	0.120 ± 0.006	0.119 ± 0.006	0.145 ± 0.007		3.531 ± 0.177	0.100 ± 0.005		0.104 ± 0.005	1.901± 0.095
7	hexaconazole	-	-	-	-		-	-		0.038 ± 0.002	-
8	pretilachlor	0.128 ± 0.006	-	0.053 ± 0.003	-		0.096 ± 0.005	0.054 ± 0.003		0.055 ± 0.003	0.072 ± 0.004
9	propisochlor	0.019 ± 0.001	-	-	-		0.038 ± 0.002	-		-	0.011 ± 0.001

Figure 6
Average amount of total concentration for detected pesticides in 26 river water samples collected along Red River.

It is clearly shown in Figure 6 that the average concentration in river water samples collected at Nam Dinh province was the highest (6.77 ± 0.335 µg L^-1), followed by Phu Tho (6.59 ± 0.326 µg L^-1), Ha Nam (3.86 ± 0.189 µg L^-1) and Lao Cai (2.96 ± 0.148 µg L^-1). In contrast, the quality of surface water in Hanoi city, Vinh Phuc and Thai Binh provinces showed to be smaller than others: average concentrations of pesticides in the collected samples at three provinces and Hanoi city were 0.20 ± 0.006, 0.27 ± 0.009, and 0.39 ± 0.015 µg L^-1, respectively. Considering the contribution of individual pesticides, the analytical results of this study showed that the reported concentration for some pesticides was exceeded from ten to hundred times higher than the allowable limit level established by European Commission³⁰ (not detection or < below 0.1 µg L^-1), as well as the National Technical Regulation on Surface Water Quality in Vietnam (QCVN 08:2023/BTNMT).⁴⁰ Among that, fenobucarb concentration was up to approximately 18.776 µg L^-1 in one sample collected in Nam Dinh province. Besides, alachlor, fenobucarb and acetochlor were often determined in river water samples at levels ranging from 0.099 ± 0.005 8.978 ± 0.006 µg L^-1, 0.058 ± 0.003 18.776 ± 0.939 µg L^-1, and 0.073 ± 0.004 1.460 ± 0.073 µg L^-1, respectively. Alachlor, fenobucarb and acetochlor were found at approximately high concentrations in comparison with others in this study. Their concentrations were much higher than the regulated pesticides mentioned in QCVN 08:2023/BTNMT.⁴⁰ According to regulation TT10/2020/TT-BNNPTNT of Ministry of Agriculture and Rural Development (MARD) in Vietnam,¹⁰ almost pesticides found in Red River water samples have been on list of pesticides permitted to use in agriculture field in Vietnam. Despite being detected at low level in this study, many pesticides are in the list of pesticides that pose a risk to humans based on its metabolite to sulfone, sulfide and disulfonyl forms.⁵ Therefore, further investigation with large number of river water samples is necessary to know clearly about the presence of hazardous pesticides in environment as well as the potential effect on human health in the investigated areas, because this Red River is the main source of raw water supply before treatment for people living along this river, as well as in delta of this river, in Vietnam.

Conclusions

This study developed an analytical method for screening and quantitative determination for multi-residue pesticides in river water samples collected in the Red River system, located in northern of Vietnam by using SPE technique for preparation and ultra high pressure liquid chromatography-high resolution mass spectrometry (UPLC-Orbitrap MS) for analysis. This analytical procedure, to the best of our knowledge, for the first time in Vietnam, was used for simultaneous quantification and determination of 118 pesticides in full scan mode and confirmation of fragmentation MS² mode. This validated analytical method was completely suitable for application to determine pesticides in 26 river water samples. It clearly demonstrated that in the collected river water samples were found acetochlor, alachlor, and fenobuchlor at approximately high concentration in comparison with others related in QCVN 08:2023/BTNM.⁴⁰ Therefore, it is necessary to propose a national regulation regarding these concentrations for the three analytes above to control pesticide pollution in Vietnam.

Supplementary Information

The data used in this work are available free of charge at http://jbcs.sbq.org.br as PDF file.

Acknowledgements

This research was funded by National Foundation for Science and Technology Development (Nafosted) under the grant number 104.04-2018.331.

References

• ¹ Haron, S. H.; Ismail, B.; Latif, M. T.; Int. J. Basic Appl. Sci. 2012, 12, 85. [Link] accessed in January 2025
  » Link
• ² Thuy, P. T.; Geluwe, S. V.; Nguyen, V. A.; Bruggen, B. V. D.; J. Mater. Cycles Waste Manage. 2012, 14, 379. [Crossref]
  » Crossref
• ³ Yadav, I.; Devi, N.; Environ. Sci. Eng. 2017, 6, 140. [Link] accessed in January 2025
  » Link
• ⁴ Dujaković, N.; Grujić, S.; Radišić, M.; Vasiljević, T.; Laušević, M.; Anal. Chim. Acta 2010, 678, 63. [Crossref]
  » Crossref
• ⁵ Gunasekara, A. S.; Truong, T.; Goh, K. S.; Spurlock, F.; Tjeerdema, R. S.; J. Pestic. Sci. 2007, 32, 189. [Crossref]
  » Crossref
• ⁶ Alder, L.; Greulich, K.; Kempe, G.; Vieth, B.; Mass Spectrom. Rev. 2006, 25, 838. [Crossref]
  » Crossref
• ⁷ Casado, J.; Santillo, D.; Jonson, P.; Anal. Chim. Acta 2018, 1024, 1. [Crossref]
  » Crossref
• ⁸ Paíga, P.; Sousa, S.; Vera, J.; Bitencourt, L.; Vieira, J.; Jorge, S.; Silva, J. G.; Correia, M., Domingues, V. F.; Delerue-Matos, C.; Environ. Sci. Pollut. Res. 2021, 28, 66787. [Crossref]
  » Crossref
• ⁹ Climent, M. J.; Herrero-Hernandez, E.; Sanchez-Martin, M. J.; Rodriguez-Cruz, M. S.; Pedreros, P.; Urrutia, R.; Environ. Pollut. 2019, 251, 90. [Crossref]
  » Crossref
• ¹⁰ Ministry of Agriculture and Rural Development (MARD); Circular No. 03/2018/TT-BNNPTNT: List of Pesticides Permitted and Banned in Vietnam, 2018. [Link] accessed in January 2025
  » Link
• ¹¹ Schwanz, T. G.; Carpilovsky, C. K.; Weis, G. C. C.; Costabeber, I. H.; J. Chromatogr. A 2019, 1585, 10. [Crossref]
  » Crossref
• ¹² Lehotay, S. J.; J. AOAC Int. 2007, 90, 485. [Crossref]
  » Crossref
• ¹³ Wong, J.; Hao, C.; Zhang, K.; Yang, P.; Banerjee, K.; Hayward, D.; Iftakhar, I.; Schreiber, A.; Tech, K.; Sack, C.; Smoker, M.; Chen, X.; Utture, S. C.; Oulka, D. P.; J. Agric. Food Chem. 2010, 58, 5897. [Crossref]
  » Crossref
• ¹⁴ Stachniuk, A.; Szmagara, A.; Czeczko, R.; Fornal, E.; J. Environ. Sci. Health, Part B 2017, 52, 446. [Crossref]
  » Crossref
• ¹⁵ Lee, J.; Shin, Y.; Lee, J.; Lee, J.; Kim, B. J.; Kim, J. H.; Chemosphere 2018, 207, 519. [Crossref]
  » Crossref
• ¹⁶ García-Reyes, J. F.; Hernando, M. D.; Molina-Díaz, A.; Fernández-Alba, A. R.; TrAC, Trends Anal. Chem. 2007, 26, 828. [Crossref]
  » Crossref
• ¹⁷ Malato, O.; Lozano, A.; Mezcua, M.; Agüera, A.; Fernandez Alba, A. R.; J. Chromatogr. A 2011, 1218, 7615. [Crossref]
  » Crossref
• ¹⁸ Rajski, M.; Gómez-Ramos, M.; Fernández-Alba, A. R.; Compr. Anal. Chem. 2016, 71, 119. [Crossref]
  » Crossref
• ¹⁹ Gómez-Ramos, M. M.; Ferrer, C.; Malato, O.; Agüera, A.; Fernández-Alba, A. R.; J. Chromatogr. A 2013, 1287, 24. [Crossref]
  » Crossref
• ²⁰ Havlíková, L.; Matysová, L.; Hájková, R.; Šatínský, D.; Solich, P.; Talanta 2008, 76, 597. [Crossref]
  » Crossref
• ²¹ Huang, H.; Li, Z.; He, Y.; Huang, L.; Xu, X.; Pan, C.; Guo, F.; Yang, H.; Tang, S.; J. Chromatogr. B:Anal. Technol. Biomed. Life Sci. 2021, 1179, 122847. [Crossref]
  » Crossref
• ²² Rajski, Ł.; Petromelidou, S.; Díaz-Galiano, F. J.; Ferrer, C.; Fernández-Alba, A. R.; Talanta 2021, 228, 122241. [Crossref]
  » Crossref
• ²³ Hajeb, P.; Zhu, L.; Bossi, R.; Vorkamp, K.; Chemosphere 2021, 287, 132306. [Crossref]
  » Crossref
• ²⁴ United States Environmental Protection Agency (US EPA); LSASDPROC-201-R6: Surface Water Sampling; US EPA Science Division: Georgia, 2023. [Link] accessed in January 2025
  » Link
• ²⁵ Want, E. J.; Masson, P.; Michopoulos, F.; Wilson, I. D.; Theodoridis, G.; Plumb, R. S.; Shockcor, J.; Loftus, N.; Holmes, E.; Nicholson, J. K.; Nat. Protoc. 2013, 8, 17. [Crossref]
  » Crossref
• ²⁶ Want, E. J.; Wilson, I. D.; Gika, H.; Theodoridis, G.; Plumb, R. S.; Shockcor, J.; Holmes, E.; Nicholson, J. K.; Nat. Protoc. 2010, 5, 1005. [Crossref]
  » Crossref
• ²⁷ EnviPat: Isotope Pattern Calculator, https://www.envipat.eawag.ch/index.php, accessed in January 2025.
  » https://www.envipat.eawag.ch/index.php
• ²⁸ Nam, V. D.; Trung, N. Q.; Thuy, L. M.; Xuyen, N. T.; Tri, M. T.; Hai-Anh, V.; Lan-Anh, N.; Tien, D. D.; Hoi, B. V.; Cam-Tu, V.; Dung, L. V.; Lan-Anh, P. T.; Hong-An, V. T.; Chu, D. B.; J. Anal. Methods Chem. 2019, 2019, 489634. [Crossref]
  » Crossref
• ²⁹ US Food and Drug Administration (FDA); Guidance for Industry Mass Spectrometry for Confirmation of the Identity of Animal Drug Residues; FDA: Rockville, Maryland, 2003. [Link] accessed in Januray 2025
  » Link
• ³⁰ Commission of the European Communities; 2002/657/EC: Commission Decision of 12 August 2002 Implementing Council Directive 96/23/EC Concerning the Performance of Analytical Methods and the Interpretation of Results; Official Journal of the European Union, 2002, L 221, p. 8. [Link] accessed in January 2025
  » Link
• ³¹ Trufelli, H.; Palma, P.; Famiglini, G.; Cappiello, A.; Mass Spectrom. Rev. 2011, 30, 491. [Crossref]
  » Crossref
• ³² Sante/11813/2017: Analytical Quality Control and Method Validation Procedures for Pesticide Residues Analysis in Food and Feed, European Union Reference Laboratories for Residues of Pesticides: Stuttgart, German, 2018. [Link] accessed in January 2025
  » Link
• ³³ CAC/GL 40-1993: Guidelines on Good Laboratory Practice in Pesticide Residue Analysis, Codex Alimentarius: Rome, 2010. [Link] accessed in January 2025
  » Link
• ³⁴ Nakhjavan, B.; Bland, J.; Khosravifard, M.; Molecules 2021, 26, 6627. [Crossref]
  » Crossref
• ³⁵ Campanale, C.; Massarelli, C.; Losacco, D.; Bisaccia, D.; Triozzi, M.; Uricchio, V. F.; TrAC, Trends Anal. Chem. 2021, 144, 116423. [Crossref]
  » Crossref
• ³⁶ Association of Official Analytical Chemists (AOAC); Appendix F: Guidelines for Standard Method Performance Requirements; AOAC: Rockville, Maryland, 2016. [Link] accessed in January 2025
  » Link
• ³⁷ US Food and Drug Administration (FDA); Bioanalytical Method Validation Guidance for Industry; FDA: Silver Spring, USA, 2018. [Link] accessed in January 2025
  » Link
• ³⁸ Kruve, A.; Künnapas, A.; Herodes, K.; Leito, I.; J. Chromatogr. A 2008, 1187, 58. [Crossref]
  » Crossref
• ³⁹ Nguyen, T. H.; An Overview of Agricultural Pollution in Vietnam: The Crops Sector; World Bank: Washington, DC, 2017. [Link] accessed in January 2025
  » Link
• ⁴⁰ QCVN 08:2023/BTNMT: National Technical Regulation on Surface Water Quality, Ministry of Natural Resources and Environment: Hanoi, Vietnam, 2023. [Link] accessed in January 2025
  » Link

Edited by

• Editor handled this article:
  Maria Cristina Canela (Associate)

Publication Dates

• Publication in this collection
  21 Feb 2025
• Date of issue
  2025

History

• Received
  29 July 2024
• Accepted
  29 Jan 2025