Department of Mechanical Engineering
Northwestern University
Abstract-
Electroosmotic flow (EOF), first observed by Reuss in 1809, has become the
method of choice for moving fluids around in microfluidic devices.
In addition to the obvious convenience of not requiring moving parts,
EOF has the added advantage of a flat velocity profile, and therefore
low dispersion, in electrophoretic separations. This is however true
only if the wall zeta potential is constant. In many applications,
the wall zeta potential is variable, either through accident
or design. For example, in electrophoresis of proteins, the protein
molecules often stick to the wall due to electrostatic attraction,
changing the zeta potential behind the migrating zone. An approximate
analytical solution, valid for "slowly varying" straight microfluidic
channels will be presented. Except for the requirement that relative
variations in the axial direction over a channel diameter be small
("slow" variations), the cross-sectional shape and zeta-potential
can vary spatially in a completely arbitrary manner. The solution
shows that the volume flux through the channel is a linear function of
the applied pressure head and electric potential drop with coefficients
("fluidic impedances") determined solely by channel geometry and charge
distribution. Evaluation of the flow profile only requires a knowledge of
the 2D Green function for the channel cross-section. The theoretical
results are used to explain experimental data on the increase
in elution times in capillary zone electrophoresis due to wall adsorption
of analytes, and, the loss of resolution (number of theoretical plates)
of a zone of neutral marker in a partially coated capillary.
GALCIT Home Page
|
|