Hi! It depends on your capillary material. If you use fused silica, then you have to consider the EOF, which will carry your solution bulk to the cathode. As the EOF mobility typically exceeds the DNA mobility you will end up with your analytes at your injection end (with your described setup). So either opt for another capillary material or place your detector at the cathodic end...

Thank you, Victor.

That was my thought initially. We were exploring possibilities to improve current sensing in each channel. The material is cyclic-olefin-copolymer (plastic, COC) so the EOF may be very reduced. I have no precise evaluation of surface charges of the COC material. If the EOF from COC is low, I may be able to induce the plug to move in the other direction but very slowly and with mostly peak broadening from diffusion. 

Dear Cedric,

I have not worked with COC capillaries so far, however, I think your original setup sounds feasible.

Victor 

EOF could be lower with COC, thus it is possible to perform DNA separation with injection end negative and waste end grounded. Besides EOF, adsorption is also an important factor to consider if a good peak shape to be expected, as in fused silcia capillary. 

I don't know whether it helps or not, but contact [email protected], see her publications