Description of chromatographic performance in reversed-phase liquid chromatography based on a generalised model. Part II: Prediction of serial-column coupling.

While reversed-phase liquid chromatography (RPLC) using C18 columns remains a cornerstone technique in analytical separations, its limited selectivity has driven the development of alternative stationary phases and advanced strategies, such as serial-column coupling, to expand chromatographic capabilities. This approach connects columns of different lengths and chemistries in series to create hybrid configurations that enhance separation performance. Beyond improving resolution, serial coupling can also reduce analysis times, particularly combined with modern zero-dead-volume connectors that minimise band broadening, which have renewed interest in this approach. In this work, we explore the serial coupling of three distinct stationary phases (high-carbon-load C18, phenylsilane, and cyanohexylsilane) to separate a set of 15 sulphonamides with diverse polarities, under both isocratic and gradient elution modes. The performance of these column combinations is evaluated using two retention modelling approaches: traditional individual models that rely on solute-specific parameters, and generalised models (GEMs) with shared system descriptors to capture solute-stationary and solute-mobile phase interactions across compound sets. GEMs offer a robust and streamlined framework for retention prediction and rational column selection. Notably, serial coupling substantially impacts chromatographic selectivity, including peak order reversals, which GEMs successfully predict.