Automated Feature Mining for Two-Dimensional Liquid Chromatography Applied to Polymers Enabled by Mass Remainder...

Area of Science:

• Polymer Chemistry
• Analytical Chemistry
• Computational Chemistry

Background:

• Automated analysis of complex polymer data is challenging.
• Comprehensive two-dimensional liquid chromatography-mass spectrometry (LC x LC-MS) generates vast datasets.
• Efficient algorithms are needed to extract meaningful information from polymer LC x LC-MS data.

Purpose of the Study:

• To develop a fast, automated algorithm for feature mining of synthetic polymers.
• To analyze homopolymers and alternating copolymers using LC x LC-MS data.
• To enable the proposal of chemical structures for identified polymer compositional series.

Main Methods:

• Data reduction by selecting regions of interest and clustering.
• Isotopic distribution analysis to determine charge states and calculate reduced masses.
• Automated monomer mass selection and mass remainder analysis for end-group composition.
• Mapping compositional series onto chromatograms and separating them in mass-remainder or chromatographic domains.

Main Results:

• A novel algorithm successfully processed LC x LC-MS data in under 3 minutes.
• The algorithm accurately identified compositional series and end-group compositions.
• False positives were assessed, and chemical structure proposals were made for identified series.
• The method was validated using industrial copolyester data.

Conclusions:

• The developed algorithm provides a rapid and automated approach for synthetic polymer analysis.
• This method significantly enhances the ability to characterize complex polymer structures from LC x LC-MS data.
• The automated feature mining and structure proposal capabilities offer valuable insights into polymer composition and end-group analysis.