Imine-linked porous covalent organic framework used for the solid-phase extraction of estrogens from honey prior to liquid chromatography-tandem mass spectrometry

This study aimed to establish a method for the rapid determination of trace estrogens in honey samples by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) using imine-linked porous covalent organic framework material (IL-COF-1) as the adsorbent for solid-phase extraction (SPE). Estradiol (E1), diethylstilbestrol (DES), estriol (E3), β-estradiol (E2), and ethinylestradiol (EE2) were used as the target analytes. A single factor optimization method was performed to optimize the extraction effect by adding estrogens to honey samples. The optimal conditions were as follows. A total of 30 mg IL-COF-1 was filled in the SPE column. The sample pH was adjusted to 7. The sample was loaded at a flow rate of 3 mL/min and eluted with 5 mL of a 1% (v/v) NH3·H2O-methanol solution. The flow rate of the eluent was 0.4 mL/min. NaCl was not added in the extraction process. HPLC coupled to electrospray ionization and triple quadrupole mass spectrometry was introduced to quantify the estrogens in the extracts. The estrogens were separated on a Thermo Fisher Scientific C18 analytical column (100 mm×2.1 mm, 5 μm). Acetonitrile and 5 mmol/L ammonium acetate solution were used as the mobile phases for gradient elution. The column temperature was set at 40 ℃, and the autosampler temperature was maintained at 10 ℃. The rapid qualitative and quantitative analysis of the five estrogens in the honey samples was operated under multiple reaction monitoring mode in a negative electrospray ion source mode. IL-COF-1 prepared in six batches was used as a filler for the SPE column. The relative standard deviations (RSDs) of the recoveries of the estrogens among different batches were 5.2%-9.1%. The reusability of IL-COF-1 material was assessed. After six SPE cycles on the same solid-phase extraction column, the RSDs of the estrogen recoveries were 2.5%-6.1%, indicating that IL-COF-1 has good reusability. The recoveries of estrogens obtained on an IL-COF-1 solid-phase extraction column within 6 days (tested once a day) were 95.1%-107.4%, and the RSDs were 6.2%-8.9%. These results confirmed that the SPE filler had good stability. The method validation results showed that the linear detection ranges were 1-500 ng/g for E3, E2, and EE2, and 0.1-100 ng/g for E1 and DES withe the correlation coefficients of 0.9934-0.9972. The limits of detection (LODs, S/N=3) were 0.01-0.30 ng/g, and the limits of quantification (LOQs, S/N=10) were 0.05-0.95 ng/g. Five estrogens were added (50 ng/g) for the repeated experiments. The RSDs of the intra-day precision were 3.2%-6.6%. The RSDs of the inter-day precision were 4.2%-7.9%. This method was applied to determine the estrogen levels in four honey samples, and no estrogen was found. The recoveries of the five estrogens in sample spiked at three levels including low, middle, and high levels were investigated, and satisfactory recoveries (80.1%-115.2%) were obtained. The SPE-HPLC-MS/MS method based on IL-COF-1 is rapid, accurate, and sensitive, making it suitable for analyzing and detecting estrogen in honey. Further exploration of the use of IL-COF-1 for the extraction processes is in progress.