Simultaneous resolution of overlapping peaks in high-performance liquid chromatography and micellar electrokinetic...

Area of Science:

• Analytical Chemistry
• Chemometrics
• Spectroscopy

Background:

• Hyphenated separation techniques like High-Performance Liquid Chromatography (HPLC) and Microemulsion Electrokinetic Chromatography (MEKC) are powerful analytical tools.
• Data from these methods, when sharing a detection mode (e.g., Diode Array Detection - DAD), can be combined for enhanced analysis.
• Challenges arise from potential wavelength scale shifts between detectors, hindering data augmentation.

Purpose of the Study:

• To introduce a novel method, wavelength shift eigenstructure tracking (WET), for aligning wavelength scales of detectors from different separation techniques.
• To evaluate the effectiveness of data augmentation using corrected data for improved curve resolution.
• To assess the impact of normalization techniques on augmented curve resolution.

Main Methods:

• Development and application of wavelength shift eigenstructure tracking (WET) for detector wavelength scale alignment.
• Augmentation of data matrices from HPLC-DAD and MEKC-DAD after WET correction.
• Application of iterative target transformation factor analysis and augmented curve resolution.
• Investigation of intensity and noise normalization effects.

Main Results:

• WET effectively detects and corrects wavelength shifts between HPLC-DAD and MEKC-DAD detectors.
• Augmented curve resolution significantly improves estimates of chromatographic and electrophoretic profiles and spectra compared to non-augmented methods.
• The benefits of augmented resolution are particularly pronounced for chromatographic peaks with purity less than 0.10.
• Normalization techniques can alter the relative contribution of MEKC and HPLC data but require careful consideration of signal-to-noise ratio and algorithm stability.

Conclusions:

• Wavelength shift eigenstructure tracking (WET) is a robust method for aligning wavelength scales in hyphenated techniques with shared detection.
• Data augmentation combined with WET-corrected data and iterative target transformation factor analysis enhances the resolution of complex mixtures.
• Normalization strategies should be applied judiciously, weighing potential benefits against information loss and algorithmic robustness.