Sodium borate
Sodium borate as an electrophoresis buffer

Brody and Kern (Sodium boric acid: a Tris-free, cooler conductive medium for DNA electrophoresis, Biotechniques 36, 214-216, Feb, 2004) recently described a sodium borate buffer for nucleic acid electrophoresis.
"... to our knowledge no one has substantially investigated the simplification and substitution of components of these buffers to achieve a more efficient and inexpensive conductive medium for DNA electrophoresis."
Considering time and effort put into DNA electrophoresis it's indeed disturbing how little effort has gone into optimizing buffers. The authors' inference that sodium is superior to Tris is questionable. There are two good reasons for using Tris with boric acid. Firstly it's a buffer and contributes to buffering in the electrode wells. More importantly Tris has a lower ionic mobility than sodium (see Jovin table), and at equivalent cation concentration and voltage will give less heating than sodium.

An obvious reason for sodium borate superiority in the Biotechniques article is low salt concentration (standard TBE is 89mM Tris). In theory Tris buffers are superior to sodium buffers at equivalent cation concentrations (where the two buffers are most usefully compared). Boric acid is well known to complex polyols. If the latter affects electrophoretic resolution, then boric acid concentration could influence separation independent of pH and salt. If boric acid complexation to Tris is problematic, then other low mobility cations could be substituted.

Other factors shouldn't be forgotten. As Brody and Kern mention, maintaining electrode well pH is important (electrophoresis system geometry affects this). Measuring well pHs after electrophoresis is good practice. Dispersing heat (thin gels surrounded by heat conducting media) is also helpful. Salt concentration in the sample affects resolution. If the sample has more salt than the gel, bands will spread. A partial solution is letting salts diffuse from the sample before applying voltage. For reproducibility I recommend weighing buffer components (and avoiding the pH meter).

Tris buffers have been tried at low concentration (personal communication, Scott Kern) and appear to be inferior to comparable levels of alkali metal cations. This suggests specific cation effects on DNA migration in agarose gels which have interesting theoretical and practical implications. Might specific ion effects in electrophoresis be useful in general?