Premium
Isotachophoresis with counterflow in an open capillary: Computer simulation and experimental validation
Author(s) -
Liu Bingwen,
Ivory Cornelius F.
Publication year - 2013
Publication title -
journal of separation science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.72
H-Index - 102
eISSN - 1615-9314
pISSN - 1615-9306
DOI - 10.1002/jssc.201300066
Subject(s) - isotachophoresis , capillary action , dispersion (optics) , chemistry , mechanics , taylor dispersion , countercurrent exchange , thermal diffusivity , chromatography , analyte , capillary electrophoresis , analytical chemistry (journal) , diffusion , materials science , thermodynamics , optics , composite material , electrolyte , physics , electrode
The purpose of applying a countercurrent flow to isotachophoretic migration is to increase the effective separation channel length during ITP. However, severe dispersion induced by applying a counterflow can be detrimental to ITP. This paper uses numerical simulations in a 2D axisymmetric domain to investigate the dispersion caused by a parabolic counterflow in open‐capillary ITP. Counterflow in these simulations was generated by applying a back pressure to stop the isotachophoretic stack, i.e., forming stationary ITP zones. It is found that dispersion is strongly related to analyte molecular diffusivity: R ‐phycoerythrin, due to its small diffusivity, showed ∼20‐fold increase in zone width in stationary counterflow ITP, compared to ITP in the absence of counterflow, while fluorescein only had ∼10% increase in zone width under similar operating conditions. Applying the Taylor–Aris dispersion formula in counterflow ITP simulations provided only a rough estimate of the dispersion, e.g., overestimation of analyte zone widths. Experiments on counterflow ITP were conducted in a silica capillary that was covalently and dynamically coated to exclude electroosmosis effect. The counterflow was generated by adjusting the relative height of the fluids in the two reservoirs at the capillary ends. Good qualitative agreement between simulations and experiments was found.