Prediction of internal standards in reversed-phase liquid chromatography -1. Initial study on predicting internal standards for use with neutral samples based on linear solvation energy relationships

This paper describes the results of an initial study on the application of linear solvation energy relationships (LSERs) to the prediction of internal standard compounds in reversed-phase liquid chromatographic (RPLC) method development. Six neutral samples are separated on an Inertsil ODS(3) column by either acetonitrile-water or methanol-water mobile phases under either isocratic or linear gradient conditions. After the separation conditions ar e optimized, the desired positions for internal standard candidates are sel ected based on the "open windows" of the chromatograms. The compounds with the desired retention range are then predicted based on LSERs from a databa se consisting of more than 700 compounds with defined physicochemical prope rties. The prediction requires the use of LSER coefficients under the separ ation conditions for each sample. They are determined a priori by performin g multivariable linear regression on the retention of 20 reference solutes against their physicochemical properties. It can be concluded from the stud y that LSER is an excellent approach to the selection of internal standard compounds for RPLC under either isocratic or gradient elution. The average prediction error is usually within 10%, but no more than 20%. Finally, LSER approach is fast and systematic, and will save a significant amount of tim e and resources during RPLC method development. (C) 2001 Elsevier Science B .V. All rights reserved.