Capillary electrophoresis of macromolecules in 'syrupy' solutions: facts and misfacts

'Syrupy' solutions of liquid linear polyacrylamide (> or = 10%T, 0%C) appear to be excellent for fractionation of oligonucleotides and, potentially, for DNA sequencing. For such analyses, the silica wall must be coated by covalently bound strings of polyacrylamide; otherwise, the electroosmotic flow will slowly pump out the viscous electrolyte solution. Due to the enormous viscosity (100 Pa s for an 8% T solution) the polymer strings must be prepared in situ, by filling the capillary with the appropriate monomer solution. The reaction, however, cannot be driven to better than 80-85% conversion: in 10%T, the concentration of unreacted monomers will thus be 200-300 mM. This will give a substantial background absorbance (even at 254 nm) and leave a huge amount of potentially harmful reacting species in the background electrolyte. A chemical scavenging method is proposed here: after polymerization, a 100 mM solution of cysteine is driven in from the cathode and allowed to react for up to 10 h. At the end of the reaction period, the excess cysteine and its acrylamido adduct are driven out electrophoretically and the column is reconstituted with its normal background electrolyte. Columns thus preconditioned have been found to perform extremely well and to last as long as the inner coating (and the linear polymer filling) will last. No 'carry over' from run to run was experienced.