Electrodiffusion of Polyelectrolytes in Microfabricated Structures.

In free solution, the mobility mu _0 of a polyelectrolyte like DNA is independent of length and electric field, but in gels mu can depend strongly on length L. However, understanding mu(L) in gels is made difficult by the complex, random nature of a gel. In order to simplify the study of mu(L) in hindered environments, we developed a technique to microfabricate effectively two-dimensional chambers with simple, well-understood geometries. The mobility of DNA in a microfabricated array of 1 μm posts with a field ~1 V/cm was measured in two ways: by microscopic tracking of individual molecules, and by macroscopic observation of bands. It was found that mu(L)/mu _0 decreases from 1 sharply with increasing L for short (<1 μm) molecules, and crosses over into a length independent region for L > 15 mum. The behavior of mu(L) can be understood by modeling the observed microscopic dynamics of individual molecules using the Smoluchowski equation. We also show that mu is sensitive to polymer topology. Branched polymer motion is arrested, a surprising result as migration is non-reptative. We used the precise control over geometry afforded by microfabrication to fabricate environments designed to have mu depend strongly on L. A simple refinement to the post array is presented, where a row of posts is followed by an obstacle-free region in a repeating pattern. An appropriate choice for the pattern period can be used to maximize dmu/dL at the length of desired maximum resolution. We also describe a geometry which scaling arguments predict will give linear, increasing separation with a DC field over 2 decades of length, and we discuss considerations for implementing such a geometry in silicon. We believe that microfabricated environments may lead to separation of DNA molecules longer than the current hard-won limit of 10 megabasepairs.