Considerations in experimental and theoretical collision cross-section measurements of small molecules using travelling wave ion mobility spectrometry-mass spectrometry

Abstract

Travelling wave ion mobility spectrometry-mass spectrometry (TWIMS-MS) has the capability to separate ions based on their mobility through a gas-filled travelling wave (T-wave) device in the presence of a train of transient voltage pulses. By calibration of this device using analytes of previously determined cross-sectional area (from conventional IMS experiments), collision cross-sections of ions can be determined based on their drift time through the T-wave device. Comparison of experimentally determined cross-sections with theoretical calculations from structural models has the potential to provide methodology which can be applied to analytes of previously uncharacterised structure; however, this comparison relies on a high degree of confidence in both experimental and theoretical methods. Focussing on small (≤200 Da) molecules, collision cross-sections have been measured by TWIMS-MS employing a calibration procedure that uses both oligo-glycine peptides and human haemoglobin-derived tryptic peptides in order to extend the calibration range for the measurement of high mobility ions. The effect of TWIMS wave height parameters on the calibration is addressed. Theoretical TWIMS cross-section calculations have been performed using a rapid, Windows-based, in-house developed projection approximation algorithm. These estimates were optimised by comparison with experimental values for a series of small molecules with rigid core structures by systematic variation of the interaction radii of the atoms comprising these species until theoretical measurements were in agreement with experimentally derived TWIMS cross-sections. The effect of varying the interaction radius for the buffer gas was subsequently studied by comparison of theoretical collision cross-sections calculated using helium and nitrogen radii with TWIMS-MS cross-section measurements determined in either helium or nitrogen buffer gases. It was found that the buffer gas used in theoretical calculations should ideally match the buffer gas in which the cross-sections of the calibrants were originally determined by conventional IMS. The resolving potential of the experimental methodology was demonstrated by separation of two isomeric amino acids, leucine and isoleucine, which showed <3% difference in cross-sectional areas. Furthermore, the application of both experimental and theoretical methods to compare the gas phase and solution conformations of seven amino acids was demonstrated. This study describes important considerations for the comparison of TWIMS-MS collision cross-sections with those obtained by theoretical methods, and also demonstrates that TWIMS-MS could be applied specifically to study a range of small molecules including those of pharmaceutical interest, in particular structural isomers and diastereoisomers that may not be distinguishable by mass spectrometry alone.