A method combining 96-well plate solid-phase extraction with ultra-performance liquid chromatography-tandem mass spectrometry (96-well SPE LC-MS/MS) was developed for the simultaneous determination of eight environmental phenols in urine samples. The samples included seven bisphenol compounds and triclosan. The urine samples were thawed to room temperature, and the target analytes were deconjugated by β -glucuronidase/aryl-sulfatase in ammonium acetate buffer solution at 37℃ overnight. Then, the effects of three kinds of 96-well solid-phase extraction plates and different elution conditions on the purification of the urine samples and the environmental phenol recoveries were compared. The best purification effect was achieved on Oasis HLB 96-well plate (60 mg) solid phase extraction, using 30% (v/v) acetonitrile aqueous solution as the rinse solution. The target analytes were then eluted by methanol solution and evaporated to dryness using a nitrogen blower. After reconstruction with 0.5 mL methanol/water (1:1, v/v) solution, the target compounds were detected by UPLC-MS/MS. To achieve better chromatographic separation, two kinds of analytical columns (C18 and T3) and different types of mobile phases (methanol and acetonitrile as the organic phase) were also compared. The best chromatographic effect was achieved when the treated samples were separated on a C18 column (100 mm×2.1 mm, 1.7 μm) using acetonitrile/water as the mobile phase at a flow rate of 0.3 mL/min. Mass spectra were recorded by negative electrospray ionization under the multiple reaction monitoring (MRM) mode. The sample matrix effect was also evaluated. The absolute matrix effects of bisphenol A, bisphenol F, bisphenol S, bisphenol B, and bisphenol AF were in the range of 3.47% to 15.32%. Since the above mentioned matrix effect was weak, there was no need for compensation measures. On the contrary, tetrachlorobisphenol A, tetrabromobisphenol A, and triclosan showed an absolute matrix effect of 49.58% (moderate), 71.99% (strong), and 86.93% (strong), thus necessitating compensation measures. Therefore, this strategy uses a one-to-one corresponding isotope internal standard method to offset the matrix effect. Six different urine samples were used to evaluate the relative matrix effect. The relative standard deviations (RSDs) of the eight corresponding internal standard peak areas were 3.63%-9.06%, indicating that the relative matrix effect was stable. Under the optimized conditions, linearity ranges were 0.50-50 μg/L for bisphenol A and bisphenol AF; 0.05-50 μg/L for tetrachlorobisphenol A and bisphenol S; 0.01-50 μg/L for bisphenol F and tetrabromobisphenol A; 1.00-50 μg/L for bisphenol B; and 5.00-200 μg/L for triclosan. The correlation coefficients were all greater than 0.9995. At spiked levels of 2.5, 5, and 25 μg/L, the average recovery ratios of the eight target analytes were 81.01%-118.84%, while the intra-day and inter-day precisions were 0.38%-19.41% and 2.54%-17.83%, respectively. The limits of detection (LOD) were 0.002-1.09 μg/L, and the limits of quantitation (LOQ) were 0.007-3.63 μg/L. This method was successfully applied to the determination of the eight environmental phenols in 64 urine samples collected from Beijing area between 2019 and 2020. All the target environmental phenols were detected, except for bisphenol B and bisphenol AF. Bisphenol A and bisphenol S showed the highest detection rates of 100% and 96.9%, respectively. The detection rates of triclosan, tetrabromobisphenol A, tetrachlorobisphenol A, and bisphenol F were 57.8%, 46.9%, 23.4%, and 21.9%, respectively. The medium values of urinary concentration followed the order 1.44 μg/L(triclosan), 0.69 μg/L(bisphenol A), 0.086 μg/L (bisphenol S), 0.0032 μg/L (tetrabromobisphenol A), 0.00050 μg/L (tetrachlorobisphenol A), 0.00 μg/L (bisphenol F, bisphenol B, and bisphenol AF). The aforementioned results imply that the widespread environmental phenolic exposure of Beijing residents is worthy of attention. Compared with traditional solid-phase extraction methods, the method reported in this paper is time-saving, effective, and suitable for the simultaneous analysis of large quantities of samples; moreover, the small sample and organic solvent consumption make this method more environment- and operator-friendly.
Read full abstract