Advanced tuning of the ion management parameters in GC × GC-HRMS using a Fourier transform Orbitrap mass analyzer for pixel-based data handling and multivariate analysis

• Published in: Analytical methods Vol. 14; no. 16; p. 1646
• Main Authors: Moreira de Oliveira, Amilton, Teixeira, Carlos Alberto, Hantao, Leandro Wang
• Format: Journal Article
• Language: English
• Published: England 21-04-2022
• Subjects: Fourier Analysis; Multivariate Analysis; Petroleum
• ISSN: 1759-9679
• DOI: 10.1039/d2ay00314g

GC × GC investigations are well known to generate a substantial amount of information-rich and structurally complex data, requiring advanced data processing strategies like chemometrics. Many workflows are available for data handling and processing, such as the peak-table and pixel-based approaches. The goal of this work is to present a solution based on method development to solve the missing pixel problem that may be encountered in experiments performed with GC and GC × GC coupled to the Fourier transform orbital ion trap (FT-Orbitrap) mass analyzer. Data input is vital for pixel-based chemometric analyses, as some post-processing solutions may lead to significant loss of chemical information in the data set. Hence, a key requisite is that the chemical information is consistently indexed in the data arrays for proper pixel-based data handling and analysis. In this study, we carefully evaluated the ion management parameters to preserve the intrinsic structure and information of the data arrays of the GC × GC-FT-Orbitrap for future pixel-oriented chemometric analysis. The most acceptable conditions yielded acquisition rates up to 42.6 spectra per s, while a routine setting of 24.7 Hz was successfully employed in analyses of different petroleum fractions, producing both consistent tensor sizes and acceptable peak reconstructions. A data acquisition rate of 24.7 spectra per s and a mass resolving power of 15 000 allowed the resolution of a mass split of only 0.004 Da - which is an interesting configuration for challenging applications in petroleomics. Using such advanced settings, the missing pixel problem was reduced from up to 30% to much less than 0.04% of the data array dimension. Thus, the proposed configuration can be employed in studies that require pixel-oriented multivariate data analysis.