Abstract
Proposed is a molecular interpretation of the strong selectivity of field-inversion electrophoresis, recently applied to the separation of long DNA fragments. The sharp transition in the mobility is associated with an antiresonance occurring when chain-length fluctuations compete with the electrophoretic drift. A theory is outlined based on biased reptation (Slater and Noolandi) and length fluctuations (Doi). Systematic preaveraging allows a simple treatment, and important features of the data are correctly accounted for.