Global spectral deconvolution based on non-negative matrix factorization in GC × GC-HRTOFMS.

Area of Science:

• Analytical Chemistry
• Chromatography
• Mass Spectrometry

Background:

• Comprehensive two-dimensional gas chromatography high-resolution time-of-flight mass spectrometry (GC×GC-HRTOFMS) is powerful for complex mixture analysis.
• Nontarget screening requires robust methods for accurate compound identification and quantification.
• Spectral deconvolution is crucial for resolving co-eluting peaks in GC×GC-HRTOFMS data.

Purpose of the Study:

• To develop and evaluate a global spectral deconvolution method based on non-negative matrix factorization (NMF).
• To optimize instrumental parameters and NMF settings for high-performance detection in nontarget screening.
• To assess the method's capability in identifying and refining spectra from complex environmental samples.

Main Methods:

• Development of a global spectral deconvolution algorithm using NMF.
• Application of the method to a sediment sample analyzed by GC×GC-HRTOFMS.
• Evaluation of instrumental parameters (scan rate, mass resolution, m/z precision) and NMF settings.
• Utilizing NIST library search for spectral identification and validation.

Main Results:

• A high scan rate (50 Hz) significantly enhanced the number of detected compounds compared to 25 Hz.
• Optimal performance required higher mass resolution (1,000–10,000) and improved m/z precision for accurate mass database generation.
• After processing, 62 unique assignable spectra (match factor ≥900) were obtained, with 54 spectra refined by the deconvolution process.
• The method successfully deconvoluted complex mixtures, yielding well-resolved reference spectra.

Conclusions:

• The developed NMF-based spectral deconvolution method is effective for enhancing compound detection in GC×GC-HRTOFMS.
• The method significantly improves nontarget screening capabilities by providing accurate mass databases and refined spectra.
• This approach facilitates the detection and characterization of compounds in complex matrices, accelerating analytical workflows.