Is the solvation parameter model or its adaptations adequate to account for ionic interactions when characterizing stationary phases for drug impurity profiling with supercritical fluid chromatography?