RESUMO
Universal quantitative detection without the need for analyte reference standards would offer substantial benefits in many areas of analytical science. The quantitative capability of high-performance liquid chromatography (HPLC) with charged aerosol detection (CAD) was investigated for 50 compounds with a wide range of physical and chemical properties. It is widely believed that CAD is a mass detector. Quantification of the 50 compounds using a generic calibrant and mass calibration achieved an average error of 11.4% relative to 1H NMR. Correction factors are proposed that estimate the relative surface area of particles in the detector, taking into account the effects of the density and charge of analytes. Performing these corrections and quantifying with surface area calibration, rather than mass, shows considerably improved linearity and uniformity of detection, reducing the average error relative to 1H NMR to 7.1%. The accuracy of CAD quantification was most significantly improved for highly dense compounds, with traditional mass calibration showing an average error of 34.7% and the newly proposed surface area calibration showing an average error of 5.8%.
RESUMO
Understanding the quality of a screening collection is the first step to improving it and, as a result, the quality of the screening process. This article outlines how this issue was approached at GlaxoSmithKline and some of the hurdles that needed to be overcome to achieve success. The article focuses specifically on the necessary software and hardware infrastructure needed, and at some of the extra benefits of such a project in terms of data mining and data modelling.