LC-MS/MS systematic toxicological analysis: comparison of MS/MS spectra obtained with different instruments and...

To compare the mass spectra obtained using a linear-ion-trap (LIT) tandem mass spectrometer (QTRAP) operated in the "enhanced product ion scan" (EPI) mode with those obtained in the classical triple-quadrupole product ion scan (PIS) mode run on the same as well as on two other instruments (TSQ-Quantum and Quattro-Micro).

After tentative standardization of ion fragmentation and transmission in both polarities using a reference compound (glafenine) on the three instruments, eight test compounds detected in the positive mode and five in the negative mode were systematically infused in different ionization sources and spectral acquisition performed over approximately 5 s. The relative intensity of the ions present in the resulting spectra was quantitatively and statistically compared. Also, the intra-day and inter-day variabilities of these relative intensities, as well as the effect of increasing compound concentration, were studied using QTRAP operated in EPI mode.

The EPI and PIS modes operated on a single LIT MS/MS instrument resulted in significant differences in relative ion intensities in both polarities, and so did the other two instruments despite prior standardization with glafenine. Some fragments could be absent in certain spectra, but no unexpected or unique fragments showed up. Intra-day variability was smaller in the LIT EPI than in the regular PIS mode and in the positive than in the negative polarity. In EPI mode, both intra- and inter-day variabilities increased when the relative intensity decreased. The effect of increasing concentration on the relative intensity of major and minor ions was small but significant in both polarities. Finally, contamination and cleansing of the ionization source also had noticeable effects on MS/MS spectra, though the cause is unclear.

MS/MS spectra do not offer the expected inter-instrument reproducibility despite an attempt at standardizing the fragmentation conditions using a reference compound. However, although the inter-instrument differences in ion relative intensity were significant, the spectra obtained looked almost similar. This suggests that in library searching algorithms, higher weight should be assigned for the m/z ratios than for their relative intensity in the spectra.