Effect of background correction on peak detection and quantification in online comprehensive two-dimensional liquid chromatography using diode array detection

A singular value decomposition-based background correction (SVD-BC) technique is proposed for the reduction of background contributions in online comprehensive two-dimensional liquid chromatography (LC × LC) data. The SVD-BC technique was compared to simply subtracting a blank chromatogram from a sample chromatogram and to a previously reported background correction technique for one dimensional chromatography, which uses an asymmetric weighted least squares (AWLS) approach. AWLS was the only background correction technique to completely remove the background artifacts from the samples as evaluated by visual inspection. However, the SVD-BC technique greatly reduced or eliminated the background artifacts as well and preserved the peak intensity better than AWLS. The loss in peak intensity by AWLS resulted in lower peak counts at the detection thresholds established using standards samples. However, the SVD-BC technique was found to introduce noise which led to detection of false peaks at the lower detection thresholds. As a result, the AWLS technique gave more precise peak counts than the SVD-BC technique, particularly at the lower detection thresholds. While the AWLS technique resulted in more consistent percent residual standard deviation values, a statistical improvement in peak quantification after background correction was not found regardless of the background correction technique used.

► We compare three background correction techniques for LC × LC-DAD data.
► We compare two peak detection methods for LC × LC-DAD data after background correction.
► The AWLS technique was the only technique to completely remove the background artifacts.
► The derivative based detection approach detected more peaks than the drain algorithm.
► A significant improvement in quantification was not found after background correction.