Quantifying injection solvent effects in reversed-phase liquid chromatography

Abstract

Peak distortion due to the injection was measured as a function of injection solvent strength, volume, mass, retention factor, and column selectivity. The concept of a method's sensitivity (s) to injection solvent strength was mathematically defined as a vector of theoretical plate counts compared to an ideal vector that does not change with injection solvent strength. Near ideal sensitivity (s>0.90) was measured on all columns with all analytes in low volume injections of 1.25μL. Increasing the injection volume reduces the measured sensitivity from ideality to a greater extent than increasing the injection mass, with differing values for each column. Using column parameters measured from the hydrophobic-subtraction model and fitting parameters from the acetonitrile excess adsorption isotherm, differences among the columns studied are explained. Decreased ligand density and increased silanol activity provide a consistent peak shape with changes in injection volume or solvent strength. For method development, a quick test is suggested with the ratio of hydrophobic-subtraction column parameters, H/A, to predict the injection solvent sensitivity of a column. As H/A decreases, the sensitivity to injection solvent worsens. Sensitivity to organic modifiers other than acetonitrile are predicted with cited sorbed layer thickness, such that MeOH>EtOH>IPA≈THF≈MeCN, i.e., a strong MeOH diluent is more ideal (better) than a strong MeCN diluent.