[Determination of three diphenyl ether herbicides in rice by magnetic solid phase extraction using FeO@MOF-808 coupled with high performance liquid chromatography].

The complex matrix of rice samples and the small amount of the target analytes in the sample necessitate an effective pretreatment process to enrich the target analytes and minimize matrix interference before instrumental analysis. Magnetic solid phase extraction (MSPE) is a dispersive solid phase extraction technique which allows for the rapid separation of sorbents from the sample solution under an external magnetic field. Compared with other traditional solid phase extraction methods, MSPE has the advantages of convenient operation, minimal interference and absence of column pressure. In this work, a metal organic framework composite (FeO@MOF-808) was synthesized by a facile solvothermal method for using as an effective adsorbent to concentrate nitrofen (NIT), oxyfluorfen (OXY) and bifenox (BIF) in rice samples. Based on the pretreatment, a method was developed by coupling with high performance liquid chromatography-ultraviolet detection (HPLC-UV). The prepared material was characterized by Fourier-transform infrared spectroscopy, X-ray diffractometry, scanning electron microscopy and vibrating sample magnetometry measurements for determining its functional groups, morphology and magnetic strength. The results showed that MOF-808 has a regular octahedral morphology and well-dispersed, and the particle size of the material ranged from 400 to 500 nm with a smooth surface. The spherical FeO particles were uniformly attached to the surface of the octahedral MOF-808 crystals. The maximum saturation magnetization of this composite was 40.35 emu/g which is lower than the saturation magnetization FeO (78.26 emu/g) but still sufficient for the requirements of MSPE. The prepared FeO@MOF-808 was used in the MSPE of three diphenyl ether herbicides (Des) in rice. As is well known, the key factors influencing MSPE are the adsorption and elution processes. In order to establish the optimal extraction conditions, the adsorption parameters (adsorbent amount, extraction time, elution solvent and elution volume) were investigated in detail. A 15 mL mixed standard solution was used in the experiment, and the concentrations of the three Des were 65 ng/mL. All the experiments were performed in parallel three times. The effects of the dosages of FeO@MOF-808 (10, 15, 20, 25 and 30 mg), adsorption time (2, 4, 6, 8 and 10 min), elution solvents (acetone, acetonitrile and methanol) and elution volume (0.5 mL, 0.5 mL×2, 0.5 mL×3, 0.5 mL×4) were investigated. The Des could be adsorbed completely by using 25 mg of FeO@MOF-808 for no more than 6 min. Elution was performed with 0.5 mL×2 of methanol. Various parameters such as limits of detection (LODs), limits of quantification (LOQs), accuracy and precision of the method were evaluated. The method showed good linearity in the range of 2-300 μg/L ( > 0.998). The LODs and LOQs were 0.6, 0.6, 0.4 μg/kg and 2.0, 2.0,1.5 μg/kg for NIT, OXY, BIF respectively. At spiked levels of 5, 10 and 20 μg/kg, the recoveries ranged from 87.3% to 96.7% with relative standard deviations (RSDs) less than 10.8%. The enrichment factors (EFs) of the method for the three Des were between 25 and 29. The method was applied to the pretreatment of the three Des in real samples, and none of the Des could be detected at any of the samples. This method had a lower LOD than that of the national standard method, but its LOD and recovery were similar to those of other reference methods. In summary, the developed method has the advantages of operational simplicity, rapidity and accuracy, and it is suitable for the analysis of herbicide residues in rice samples.