Adapting Tile-Based Discovery Analysis for Comprehensive Two-Dimensional Gas Chromatography with Vacuum Ultraviolet...

Area of Science:

• Analytical Chemistry
• Chromatography
• Spectroscopy

Background:

• Comparing complex chromatograms is challenging.
• Comprehensive two-dimensional gas chromatography with vacuum ultraviolet spectroscopy detection (GC × GC-VUV) generates rich data.
• Existing methods require adaptation for GC × GC-VUV data analysis.

Purpose of the Study:

• To adapt and evaluate tile-based comparative analysis methods for GC × GC-VUV data.
• To rank analytes and differentiate gas oil sample classes.
• To establish thresholds for distinguishing analyte variations.

Main Methods:

• Application of Fisher ratio, pairwise, and relative variance ranking methods.
• Adaptation of a tile-based scheme for GC × GC-VUV analyte ranking.
• Utilizing multivariate curve resolution-alternating least-squares (MCR-ALS) for VUV spectrum analysis.
• Principal component analysis (PCA) for unsupervised sample class differentiation.

Main Results:

• Fisher ratio analysis identified significant analyte changes (459/557 hits with >2-fold change) in HDT and LCO samples.
• MCR-ALS enabled classification of analytes into compound classes (saturates, aromatics).
• Pairwise analysis established a concentration ratio >1.2 as a significant threshold.
• Relative variance analysis combined with PCA successfully differentiated five gas oil sample classes.

Conclusions:

• Tile-based methods are effective for comparative GC × GC-VUV analysis.
• The study provides a framework for analyte ranking and sample classification.
• These methods facilitate unsupervised differentiation of complex hydrocarbon mixtures like gas oils.