Modelling of analyte profiles and band broadening generated by interface loops used in two-dimensional liquid...

The shape and variance of the analyte bands introduced in the second-dimension column are important when developing 2D-LC methods as it affects parameters such as the first-dimension flow rate, the modulation time and the loop volume. Using numerical simulations and experimental measurements, better understanding and quantification of the dispersion occurring in the open tubular interface loops can be obtained. Previous studies focused on the possible sample loss due to loop overfilling and the dispersion occurring when operating the interface in the First-In-Last-Out mode (FILO). In this work, the First-In-First-Out (FIFO) mode is considered, as its behavior, and relevant operating parameters are very different. Relevant variables are the analyte diffusion coefficient (D_mol), loop filling and emptying flow rates (F_fill & F_empty), loop inner diameter or radius (R_loop), loop volume (V_loop) and the loop filling fraction f. For a straight loop capillary, we find that the dimensionless loop emptying profiles in FIFO-mode only depend on the dimensionless emptying time t_empty^*=V_loopF_empty·D_molR_loop², the loop filling fraction f and, to a lesser extent, the ratio of filling and emptying flow rates F_empty/F_fill. The peak profiles are, however, much broader and tailing than their FILO-mode counterparts. Overfilling the loop (f > 50 %) reduces the peak variance at the cost of ¹D sample loss. Models are developed to predict the peak variance in straight loop capillaries for the most relevant operating conditions, using only dimensionless parameters. The predicted emptying profiles from numerical simulations correspond very well both in shape and relative position with experimental results for straight loops over the entire range of investigated experimental conditions (f = 50 %, F_empty/F_fill = 1-20). When the straight capillary is replaced by a tightly coiled loop, significantly smaller (factor 2x-3x) peak variances are observed compared to straight capillaries. Nevertheless, the straight loop cases remain of interest as they represent an upper boundary (i.e., worst-case scenario) on the predicted variance.