A library of experimental collision cross sections for food, beverage and plant metabolites generated by reversed-phase liquid chromatography coupled to travelling-wave ion mobility and high-resolution mass spectrometry

Collision cross section (CCS), considered an ion-specific physicochemical property derived from ion mobility spectrometry (IMS), has become a valuable orthogonal descriptor for metabolite characterisation. When coupled with liquid chromatography and high-resolution mass spectrometry, CCS enables a multidimensional analytical approach to address challenging metabolite annotation, particularly for isomeric compounds or when reference standards are unavailable. In food and plant metabolomics, where chemical diversity is vast and reference databases are limited, expanding experimental CCS libraries remains a priority. Here, using reversed-phase liquid chromatography hyphenated with travelling wave ion mobility spectrometry and high-resolution mass spectrometry (RPLC-TWIMS-HRMS), we present a comprehensive library of 597 TWCCSN2 values from 274 standard references, including commercial, non-commercial and in-house isolated compounds relevant to the wine, food and plant metabolome. High intra-and inter-day repeatability was established across 35 metabolites from diverse chemical classes, with 96.5% of the 171 paired measurements showing inter-software agreement within ±2.0% ΔCCS when processed in parallel using two commercial data processing software platforms. Evaluation of five machine learning-based CCS prediction tools against the experimental library revealed up to 68.9% of predicted values within ±3.0%ΔCCS, with chemical class-dependent performance highlighting the importance of continued expansion of experimental CCS resources for confident annotation in metabolomics workflows