Optimization and validation of automated solid-phase microextraction arrow technique for determination of phosphorus flame retardants in water

In the present work, a very sensitive and fully automated direct immersion PAL SPME Arrow procedure, coupled with GC–MS, has been developed and validated for determination of nine phosphorus flame retardants in different types of water samples (river, drinking and rainwater). PDMS/DVB was selected among three commercially available SPME Arrows (PDMS/DVB, DVB/PDMS/CWR and PDMS/CWR), since it resulted in the best sensitivity. The important experimental parameters were optimized via a central composite design response surface methodology and as result, extraction time of 65 min, extraction temperature of 80 °C and added salt concentration of 19% (w/v), were selected as the optimum values. The optimized method showed linear response over the calibration range (2 – 500 ng L−1), with R2-values higher than 0.9937. The precision (RSD%) measured by replicate analyses (n = 7) was estimated at 2 and 100 ng L−1 and was less than 29% and 21%, respectively. The LOQ of PAL SPME Arrow, calculated as S/N = 10, was between 0.2 and 1.2 ng L−1 (for triphenyl phosphate and tris-(1‑chloro‑2-propyl) phosphate, respectively) with extraction efficiencies between 5.9 and 31% (for tris-(1,3-dichloro-2-propyl) phosphate and tri-n‑butyl phosphate, respectively). To assess the performance of the developed technique for real samples, two river water samples, tap water from two regions and a rainwater sample were analyzed. Most of the target analytes were observed in the river samples with concentrations of 1.0 – 250 ng L−1 and the obtained recoveries at 50 ng L−1 ranged between 60 and 107%. Considering the figures of merit of the optimized method, PAL SPME Arrow-GC–MS showed to be the most sensitive analytical approach for determination of phosphorus flame retardants in water, with satisfying precision and accuracy, compared with conventional SPME-NPD, LLE-GC–MS and SPE-LC-MS/MS.