Troubleshooting of the determination of bisphenol A at ultra-trace levels by liquid chromatography and tandem mass spectrometry.

Determination of trace amounts of bisphenol A (BPA) may cause problems mainly related to the presence of BPA in solvents (even in LC-MS grade), laboratory vessels, and plastic equipment used for sample preparation. Variable and sometimes significant amounts of BPA present in the background cause problems in obtaining good repeatability of measurements at the ultra-trace levels. Such observations (i.e., poor repeatability of results) were made during development of the LC-MS/MS method for determination of BPA in human serum samples. The method included gradient separation of the sample's constituents. The BPA peak was present in the chromatograms not only when procedural blanks were injected but also when void injections were made. One of the possible ways to eliminate background contamination is to change the source of solvents, use a different water purification system, and introduce rigorous equipment cleaning procedures. However, despite the use of these recommended guidelines, the peak of BPA was still present in the system blank. It was observed that the intensity of the BPA peak was proportional to the time of column conditioning. It was concluded that BPA, present in the components of the mobile phase, is being enriched in the front of the separation column during its conditioning (i.e., when mobile phase elution strength was low). This paper describes effects of gradient and isocratic elution conditions on LC-MS/MS system blank. The problem of spurious BPA peak, originating from the mobile phase, was solved by replacing gradient with isocratic elution mode. The use of isocratic elution conditions with the mobile phase of relatively high elution strength (50% of acetonitrile) allowed elimination of the peak of BPA coming from the mobile phase and significantly improved the precision of determination of BPA at low concentration levels.