ACD/LC Simulator

Technical Information

Prediction of Retention Time and pKa from Chemical Structure

ACD/Labs has devised a unique algorithm for the prediction of logD from chemical structure. The logD prediction (the octanol-water partition coefficient for partially dissociated mixtures) is based on the well-characterized logP (the octanol-water partition coefficient for neutral forms) contributions of separate atoms, structural fragments, and intramolecular interactions between different fragments, also taking into account pK_a (the acid-base ionization coefficient). ACD/LC Simulator uses logD values in an algorithm to predict compound retention times using reversed phase HPLC and ion exchange HPLC analyses under particular conditions. In some cases, the values of molar refractivity, molar volume, and molecular weight are also involved in the correlation.

Using chemical structures and experimental retention times for a set of compounds, the program calculates a prediction equation that can be used to extrapolate retention times for new compounds under given experimental conditions. This may also be used to predict retention times for different pH values, allowing you to optimize the mobile phase pH prior to the first injection.

System training in ACD/LogP, ACD/LogD, and ACD/pK_a has been extended to the ACD/LC Simulator algorithm to help you "teach" the software using experimental values for one compound to refine the calculations for the related compounds and classes.

For a long time, scientists have been researching how parameters such as hydrophobic and non-hydrophobic forces affect retention times. ACD/Labs software expresses these qualitative studies as accurate estimations of future experimental observations.

Retention time is a well-defined function of logP, pK_a, and net charge - all properties which ACD/Labs' software predicts accurately.

The emphasis changes somewhat for different columns. For reversed phase liquid chromatography, the algorithm uses logP and pK_a. For ion exchange chromatography, we add total charge. Furthermore, in some cases the values of molar refractivity, molar volume, and molecular weight are also included in the prediction equations. By knowing exact values for these parameters and correlations, we can accurately predict retention times.

See how ACD/LC Simulator predictions compare to experimental chromatograms.