Unsupervised parameter optimization for automated retention time alignment of severely shifted gas chromatographic...

Area of Science:

• Analytical Chemistry
• Chromatography
• Chemometrics

Background:

• Chromatographic separations often suffer from retention time shifts, complicating data analysis and comparison.
• Traditional alignment methods require training sets or prior knowledge of sample classes.
• Automated and unsupervised alignment is crucial for analyzing completely unknown datasets.

Purpose of the Study:

• To evaluate the performance of piecewise retention time alignment.
• To demonstrate automated, unsupervised parameter optimization for alignment.
• To assess the impact of alignment on peak quantification (height and area).

Main Methods:

• Simulated chromatographic data were used to test piecewise alignment.
• Alignment parameters were optimized using the average correlation coefficient between chromatograms.
• The method was applied to gas chromatography (GC) separations of gasoline and reformate samples.

Main Results:

• Simulated data showed an improvement in average relative shift from 2.0 to 0.3.
• Real GC data (gasoline, reformate) saw average relative shifts improve from 4.7 and 1.5 to 0.5 and 0.4, respectively.
• Peak height and area differences were minimal post-alignment (average relative difference in height: -0.20%; average absolute relative difference in area: 0.15%).

Conclusions:

• Piecewise retention time alignment is effective for correcting severe shifts in chromatographic data.
• Automated, unsupervised optimization allows alignment without training sets, enabling analysis of unknown samples.
• The method preserves peak integrity, ensuring accurate quantitative analysis.