Tackling matrix effects during development of a liquid chromatographic-electrospray ionisation tandem mass spectrometric analysis of nine basic pharmaceuticals in aqueous environmental samples

When developing an LC-MS/MS-method matrix effects are a major issue. The effect of co-eluting compounds arising from the matrix can result in signal enhancement or suppression. During method development much attention should be paid to diminish matrix effects as much as possible. The present work evaluates matrix effects from aqueous environmental samples in the simultaneous analysis of a group of nine specific pharmaceuticals with LC-ESI/MS/MS: flubendazole, propiconazole, pipamperone, cinnarizine, ketoconazole, miconazole, rabeprazole, itraconazole and domperidone. Solutions to diminish signal suppression were examined: optimisation of the sample preparation, decrease of the flow rate, and the use of appropriate internal standards. Several SPE-stationary phases were tested in view of retention of the analytes: Oasis HLB, C8, Phenyl, Strata X-polymer RP sorbent and Strata X-polymeric SCX/RP sorbent. Oasis HLB showed the best retention for all analytes. The Oasis HLB SPE-method was optimised, but analyses showed high matrix suppression. Therefore, a second SPE-method, on a phenyl stationary phase (the second best option), was also optimised. A comparison of the matrix effect was made between the two procedures: the phenyl-method was less subject to matrix effects, however, the average matrix effect (ME%) of 46% indicated that matrix effects where still present. Several optimisation options for the phenyl-method were evaluated: addition of a ferric nitrate solution before extraction, application of an alkaline wash step, and use of a second SPE-cartridge, either a NH2-column or a florisil column. A more efficient sample clean-up was achieved by applying the extract after extraction on the phenyl column and after dilution with chloroform, onto a NH2-column (average ME%: 53%). In addition, applying a post-column split (1:5), further reduced matrix effects (average ME%: 65%). Labelled internal standards are the best way to tackle matrix effects, but no such internal standards were commercially available for the analytes of interest. The thorough search and application of four internal standards (structural analogues) was beneficial and compensates the matrix effect partially (average ME%: 83%). In an attempt to reduce the analysis time Speedisk phenyl columns were applied. Under these conditions matrix effects decreased even more while recoveries were between 91 and 109%. Different kinds of surface water samples were analyzed, and different matrix effects were observed. For this reason, standard addition will be used to perform quantitative analysis. (c) 2006 Elsevier B.V. All rights reserved.