Comparison of two principal component analysis methods to evaluate reversed-phase retention data.

Area of Science:

• Analytical Chemistry
• Chromatography
• Chemometrics

Background:

• Herbicidal activity is often linked to specific chemical structures.
• Reversed-phase thin-layer chromatography (RP-TLC) is a valuable technique for separating and analyzing chemical compounds.
• Principal Component Analysis (PCA) is a statistical method used to reduce data dimensionality and identify patterns.

Purpose of the Study:

• To evaluate the retention behavior of twelve 2-nitro-4-cyanophenyl esters in 23 chromatographic systems.
• To assess the impact of using covariance versus correlation matrices in PCA on chromatographic data.
• To determine the optimal method for PCA application in analyzing herbicidal ester retention data.

Main Methods:

• Determining the retention of 2-nitro-4-cyanophenyl esters across 23 reversed-phase thin-layer chromatographic systems.
• Applying Principal Component Analysis (PCA) to the retention data using both covariance (Method A) and correlation (Method B) matrices.
• Comparing the results of PCA, including PC loadings and two-dimensional maps, from both methods.

Main Results:

• The ratio of variances explained by PCA was similar for both covariance and correlation matrices.
• PC loadings and coordinates on nonlinear maps showed poor correlation between the two PCA methods.
• Differences in ester and system distribution were observed on PCA maps, though general trends remained similar.

Conclusions:

• Using the correlation matrix for PCA calculations in this context may introduce distortions.
• The covariance matrix is advocated for PCA in analyzing chromatographic retention data for herbicidal compounds.
• This finding emphasizes the importance of matrix selection in chemometric analysis for accurate chemical profiling.