Matrix effects on mass spectrometric determinations of four pharmaceuticals and personal care products in water, sediments and biota
Simple analytical methods were developed for the extraction and determination of four pharmaceuticals and personal care products (PPCPs) from water, sediments, and biota. PPCPs were determined using tandem LC–MS in electrospray ionization mode, and interactions with matrix co-eluents were investigated. Extractions of water samples were performed using solid-phase extraction (SPE), sediments were extracted by pressurized liquid extraction (PLE), and biota was extracted by liquid extraction. The selected analytical methods yielded recoveries ‡ 61% in all matrixes. Matrix interactions were investigated throughout the linear range of quantification of each compound, revealing that dissolved salts had relatively minor effects on ionization (between 14% suppression to 12% enhancement), but that sediment and biota extracts caused significant matrix effects (ranging from 56% suppression to 25% enhancement). The direction and magnitude of matrix interactions reflected the physico-chemical properties of each analyte, particularly their pKa. Among the compounds analyzed in electrospray positive mode, carbamazepine was insensitive to matrix interactions, because it is a strong proton acceptor (pKa = 14.0). In contrast, atorvastatin (pKa = 4.5), a weaker proton acceptor, was particularly sensitive to matrix effects. For those compounds analyzed in negative-ion mode, sample alkalinity was found to be important. With a pKa of 10.4, 17a ethinylestradiol generally exhibited matrix enhancement with increased sample alkalinity. However, the presence of acidic co-eluents contributed to matrix suppression. Lastly, TCS was particularly sensitive to matrix suppression, as its circumneutral pKa (7.9) caused even slight changes in sample pH to considerably impact ionization. We conclude that while different matrixes have clear impacts on ionization of these PPCPs, matrix effects can be quantified and overcome.