Modeling nonbilinear total synchronous fluorescence data matrices with a novel adapted partial least squares method.

Area of Science:

• Analytical Chemistry
• Chemometrics
• Spectroscopy

Background:

• Nonbilinear data structures pose significant challenges for accurate analyte quantitation, especially in the presence of uncalibrated interferents.
• Second-order data, such as that generated by Total Synchronous Fluorescence Spectroscopy (TSFS), often exhibits nonbilinear characteristics.
• Existing methods like PARAFAC, U-PLS/RBL, and MCR-ALS struggle with the complexity of nonbilinear interferents, requiring extensive modeling parameters.

Purpose of the Study:

• To introduce a new residual modeling algorithm, U-PLS/IMNB/MCR-ALS, designed to address nonbilinear data challenges.
• To improve the quantitation of analytes in second-order spectroscopic data containing uncalibrated interferents.
• To demonstrate the algorithm's efficacy compared to established methods using simulated and real experimental data.

Main Methods:

• Development of the unfolded partial least squares with interference modeling of nonbilinear data by multivariate curve resolution by alternating least squares (U-PLS/IMNB/MCR-ALS) algorithm.
• Application of the algorithm to simulated nonbilinear datasets exhibiting interferents.
• Validation using Total Synchronous Fluorescence Spectroscopy (TSFS) data, including a real-world example for ciprofloxacin determination in the presence of norfloxacin.

Main Results:

• The U-PLS/IMNB/MCR-ALS model effectively handles nonbilinear data structures and interferents.
• Simulated data analysis showed superior performance compared to PARAFAC, U-PLS/RBL, and MCR-ALS.
• The algorithm successfully determined ciprofloxacin in real TSFS matrices with norfloxacin as an interferent.

Conclusions:

• U-PLS/IMNB/MCR-ALS offers a robust and efficient solution for analyte quantitation from nonbilinear second-order data.
• The method provides better performance and handles complex interferents more effectively than previous approaches.
• This algorithm represents a significant advancement for analyzing challenging spectroscopic datasets in analytical chemistry.