Chromatographic quantification of seven pesticide residues in vegetable: univariate and multiway calibration comparison

Abstract Second order algorithms have been used in many chromatographic data analyses to overcome the interference drawback of an unexpected coeluted constituent, however the majority of the works partially coelute the analytes to reduces the elution time, this require the multiway calibration even if an unexpected constituent is not present at the sample. It is always preferable to perform a univariate quantification than perform a multiway calibration. For this purpose, it is necessary to use a detection system whose selectivity reduces the probability of interferences, for instance a liquid chromatography with mass spectrometry (LC-MS), unfortunately the operating costs can be considerably high. In this work, a comprehensive study of univariate and multiway calibration was performed for the quantification of seven pesticide residues in vegetables using a liquid chromatography with diode array detector (LC-DAD). The univariate and multiway calibration comparison study was performed along the data acquired for a full validation of the univariate methodology, where the analytes were separated in the column previously to detection. At the vegetable samples analysis, it was shown that the results are the same when the interference is not present and only the multiway approach lead to accuracy results when it is present. Performing the multiway calibration only when it is necessary, leading to faster data analyzes, and eliminate the need of another injection, revalidation or resampling when an interferent is present and reduce the analysis cost by the use of a LC-DAD system. In addition, in this work it is possible to evaluate the multiway calibration efficiency comparing the results with the well stablished univariate calibration, which was done for the first time, using the same data, as long as the authors knowledge.