2013 Volume 38 Issue 3 Pages 144-146
Paraquat and diquat are fast-acting herbicides that have been used extensively in agriculture. The herbicides are known to be toxic, and accidental paraquat poisonings have been reported in Japan. As there are no effective treatments, ingestion above effective toxic levels often results in death. For proper food safety management to avoid poisoning, attention must be given not only to controlling their residues in foods, but also to preventing poisoning by ingestion. Accordingly, it is essential to rapidly analyze for their presence in foods.
Paraquat and diquat are cationic and extremely polar compounds that have N-alkyl pyridinium structures. Due to this property, analysis tends to be burdensome and challenging for analytical scientists.1–4) The Ministry of Health, Labor and Welfare in Japan provides an official method for analyzing paraquat and diquat. However, it requires considerable skill and time, involving the following procedures: sample extraction by refluxing with sulfuric acid and purification through a strong cation exchange mini column; then, fluorescent derivatization with potassium ferricyanide; and finally, quantification by high-performance liquid chromatography-fluorescence detection (HPLC-FLD).5)
In the present study, we examined a highly sensitive method for paraquat and diquat analysis using liquid chromatography-tandem mass spectrometry (LC-MS/MS) instead of HPLC-FLD. In this method, sample preparation procedures are simplified and expedited.
Paraquat standard, diquat dibromide monohydrate standard, HCOOH, and heptafluorobutyric acid (HFBA) [HPLC grade] were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). MeCN [LC/MS grade], MeOH [pesticide residue and polychlorinated biphenyl (PCB) analytical grade], n-hexane (PCB analytical grade), 1 mol/L HCl, and 1 mol/L NaOH aqueous solution were from Kanto Chemical CO., INC (Tokyo, Japan). Polytetrafluoroethylene (PTFE) filters (AUTOVIAL 5, 0.2 µm) were from GE Healthcare Japan (Tokyo, Japan). Oasis WCX (150 mg/6 cc) cartridges were from Waters (Milford, MA, USA).
2. Separation and quantificationConcentrations of herbicides in the samples were determined using a high-performance liquid chromatography ACQUITY UPLC system (Waters) coupled with a triple-stage quadrupole mass spectrometer API 4000 MS/MS system (AB SCIEX, Framingham, MA, USA). Analyst® 1.5 software was used to control the instruments and to process the data in the system. Analytical chromatography was carried out using an ACQUITY UPLC BEH C18 column (2.1 mm×100 mm, 1.7 µm; Waters) operating at 40°C. The mobile phase consisted of (A) 0.025% aqueous HFBA solution and (B) 0.025% HFBA in MeCN. A linear gradient profile was applied at a flow rate of 0.45 mL/min with the following proportions: 0 min (98% A, 2% B), 7 min (98% A, 2% B), 8 min (10% A, 90% B), 9 min (10% A, 90% B), 9.5 min (98% A, 2% B), and 11 min (98% A, 2% B). The injected volume of the samples was 5 µL. An electrospray ionization (ESI) source was operated in positive mode. The operating parameters were optimized under the following conditions: curtain gas, 3 psi; ion source gas 1, 30 psi; ion source gas 2, 30 psi; ion spray voltage, +5500 V; collision gas, 3 (arbitrary units); and ion source temperature, 600°C. Data for quantification and confirmation were acquired in multiple reaction monitoring (MRM) modes. Table 1 lists the precursor (Q1)-to-product (Q3) transitions, the optimum declustering potential (DP), collision energy (CE), and collision cell exit potential (CXP) for each compound. Calibration curves for determining concentrations of samples were obtained by the standard addition method, in which calibration samples were obtained by preparing samples of spiked beer and malt to assure the accuracy of the quantitative value.
| Precursor ion (m/z) | Product ion (m/z) | DP (V) | CE (V) | CXP (V) | |
|---|---|---|---|---|---|
| Paraquat_1 | 93.0a | 171.0 | 51 | 19 | 14 |
| Paraquat_2 | 186.3 | 170.9 | 221 | 25 | 14 |
| Diquat_1 | 183.2 | 156.8 | 66 | 31 | 12 |
| Diquat_2 | 183.2 | 77.9 | 66 | 51 | 8 |
a: m/z of divalent ion.
Three grams of the beer sample was loaded onto an SPE cartridge (Waters Oasis WCX 150 mg/6 cc) previously conditioned with 5 mL of MeOH and 10 mL of H2O. The loaded cartridge was washed with 5 mL of H2O, 5 mL of 50% aqueous MeOH solution, and 5 mL of MeOH to remove impurities. After washing, the herbicides were eluted from the cartridge with 5 mL of 2% HCOOH in 25% aqueous MeOH solution. The eluted fraction was collected into a 15-mL polypropylene centrifuged tube and evaporated by centrifugal solvent evaporator EZ-2 plus at 40°C (Genevac, Ipswich, Suffolk, UK). The residue was dissolved with 500 µL of H2O. Five microliters of HFBA and 5 µL of HCOOH were added to the dissolved sample. The sample was filtrated with a PTFE filter and then analyzed by LC-MS/MS.
Three grams of the malt sample was homogenized with 10 mL of 0.5% HCl, and 2 mL n-hexane was added. The homogenized sample was shaken at 300 stroke/min for 10 min and centrifuged at 3000 rpm for 10 min. Five milliliters of the supernatant was separated and poured into a 50-mL polypropylene centrifuged tube and was neutralized with 0.2 mol/L aqueous NaOH solution. The neutralized solution was centrifuged at 3000 rpm for 2 min. All of the supernatant was loaded onto an SPE cartridge (Waters Oasis WCX 150 mg/6 cc) previously conditioned with 5 mL of MeOH and 10 mL of H2O. The loaded cartridge was washed with 5 mL of H2O, 5 mL of 50% aqueous MeOH solution, and 5 mL of MeOH to remove impurities. Then, herbicides were eluted from the cartridge with 5 mL of 2% HCOOH in 25% aqueous MeOH solution. The eluted fraction was collected into the 15-mL polypropylene centrifuged tube and evaporated using the centrifugal solvent evaporator at 40°C. The residue was dissolved with 500 µL of H2O. Five microliters of HFBA and 5 µL of HCOOH were added to the dissolved sample. The sample was filtrated with a PTFE filter and then analyzed by LC-MS/MS.
4. Performance evaluationRepeatability [relative standard deviation (RSD) %] and accuracy (%) were assessed using samples spiked with paraquat and diquat at a level of 10 µg/kg; measurements were repeated six times on the same day. The coefficient of linearity was determined by the standard addition method using the samples spiked with paraquat and diquat at levels of 5, 10, 50, 100, 500, and 1000 µg/kg. The limit of quantification (LOQ) was 10 µg/kg, which was the spiked point for repeatability and accuracy.
Due to their extremely polar properties, paraquat and diquat cannot generally be retained on a C18 column in reversed phase chromatography. Meanwhile, they are strongly cationic compounds that have N-alkyl pyridinium structures, and therefore HFBA, a strong acidic ion pair reagent, can be applied to retain them effectively on a C18 column. The concentration of HFBA in the mobile phase was optimized from resolution to 0.025% (Fig. 1). When it was over 0.025%, the sensitivity of LC-MS/MS was decreased remarkably.
Due to their strongly cationic properties, paraquat and diquat can be held and purified efficiently by a weak cation exchange SPE cartridge. An Oasis WCX cartridge, a mixed-mode weak cation exchanger that has its π–π interaction and electrostatic interaction with the herbicides, is expected to show good selectivity for the analysis. The washing conditions of the sample loaded on the Oasis WCX were examined with three solvents: water, 50% aqueous MeOH, and MeOH. The herbicides were not detected in any of the wash fluids, and thus the above washing condition was adopted. After removal of matrix components, the herbicides were eluted with a weak acidic solution. Five microliters of 2% HCOOH in 25% aqueous MeOH solution showed acidity sufficient to dissociate the interaction between the herbicides and the carboxylate anions of the Oasis WCX.
In order to efficiently extract the herbicides from the malt sample, strong acidity was required for the extraction solvents. The extraction efficiencies (%) of the herbicides were as follows: (Extraction solvent, pH, rate of paraquat, rate of diquat)=(H2O, 5.0, 19%, 27%), (10% CH3COOH, 2.5, 52%, 69%), (0.5% HCl, 0.5, 62%, 74%). The extraction rate was not improved for HCl with the acidity stronger than 0.5%, and thus 5 mL of 0.5% HCl (pH=0.5) was selected as an extraction solvent.
3. Result of the performance evaluationGood results were obtained from the performance evaluation in both beer and malt samples (Table 2).
| Sample | Compound | Repeatability (%) | Accuracy (%) | Linearity (r) |
|---|---|---|---|---|
| Beer | Paraquat | 2.5 | 97.6 | 1.000 |
| Diquat | 3.5 | 100.4 | 0.999 | |
| Malt | Paraquat | 7.2 | 97.9 | 0.998 |
| Diquat | 12.0 | 93.9 | 1.000 |
A method for the simultaneous determination of paraquat and diquat by LC-MS/MS has been established. The procedures developed for sample preparation involve extraction with dilute HCl and purification with a weak cation exchange SPE cartridge. With this method, a rapid analysis of paraquat and diquat was achieved while providing high sensitivity (LOQ=10 µg/Kg). By eliminating burdensome steps, such as refluxing with sulfuric acid or fluorescent derivatization, the time required per sample preparation was reduced considerably to half a day. Moreover, our sample preparation method has been so simplified that any analyst can conduct a precise analysis. While accidental and intentional poisonings by paraquat and diquat have been a concern, this method with fast and easy handling of samples will be applied extensively in the field of analytical chemistry for food safety management.
