Framework for maximizing throughput in preparative liquid chromatography