Critical evaluation of the role of external calibration strategies for IM-MS

Abstract The major benefits of integrating ion mobility (IM) into LC–MS methods for small molecules are the additional separation dimension and especially the use of IM-derived collision cross sections (CCS) as an additional ion-specific identification parameter. Several large CCS databases are now available, but outliers in experimental interplatform IM-MS comparisons are identified as a critical issue for routine use of CCS databases for identity confirmation. We postulate that different routine external calibration strategies applied for traveling wave (TWIM-MS) in comparison to drift tube (DTIM-MS) and trapped ion mobility (TIM-MS) instruments is a critical factor affecting interplatform comparability. In this study, different external calibration approaches for IM-MS were experimentally evaluated for 87 steroids, for which TWCCSN2, DTCCSN2 and TIMCCSN2 are available. New reference CCSN2 values for commercially available and class-specific calibrant sets were established using DTIM-MS and the benefit of using consolidated reference values on comparability of CCSN2 values assessed. Furthermore, use of a new internal correction strategy based on stable isotope labelled (SIL) internal standards was shown to have potential for reducing systematic error in routine methods. After reducing bias for CCSN2 between different platforms using new reference values (95% of TWCCSN2 values fell within 1.29% of DTCCSN2 and 1.12% of TIMCCSN2 values, respectively), remaining outliers could be confidently classified and further studied using DFT calculations and CCSN2 predictions. Despite large uncertainties for in silico CCSN2 predictions, discrepancies in observed CCSN2 values across different IM-MS platforms as well as non-uniform arrival time distributions could be partly rationalized.