I am assuming conductivity is a linear function of the current. In order to answer this, I would need to know how long you wait after applying voltage before measuring the DC current what is the frequency of the AC voltage you use. The onset of electronic current usually occurs in less than a picosecond after applying the voltage but ionic current might take more time to attain its full magnitude.

Imre @ Thanks for the informative advice. I use solartron 1260A imedance analyzer and use frequency range from 1-10E6 Hz to measure ac conductivity from room temperature to 800C. I use Nyquist plots to find the grain and grain boundary conductivity. For dc measurements, I consider initial value that occurs in a second to measure electronic current in dc measurement. I wish to distinguish the ionic and electronic contributions to the total conduction. Please advice.

Hi Kapil. In case of AC, since the ions might respond more slowly to the applied voltage, the ionic current should not contribute much to the total current at high frequencies (say, 1 MHz). Therefore, at 1 MHz all of your conductivity should arise from electronic conduction only. You should also look at the Bode plot of Z'' vs f. There could be a relaxation peak there from the ions. At frequencies lower than the peak, ions should contribute to the conduction but at frequencies well above it, only electronic conduction should occur.