Did you measure impedance during your polarisation experiment (or before and after)?

One reason could be that the passivation layer on the lithium electrodes breaks down when lithium is dissolved / deposited. Impedance would tell you if this is the case.

Beside the de-passivation of the Li surface, the growth of Li dendrites may significantly contribute to the current increase. The dendrite growth causes a surface incement, which itself can be a sufficient reason for the current increase upon potetiostatic polarization. However, the dendrite growth decreases the anode/cathode distance. Hence, by merely the drop in the ohmic resistance of your gel, the current can increase because one of the 3 impedance terms (anode impedance, cell resisitance, cathode impedance) the middle one decreases. Also, eventually the densrites can also short-circuit the cell after a long time, leading to breakdown (this is why metallic Li is not used in commercial batteries since Li is prone to develop dendrites upon deposition).

If you measure impedance, the drop in the high-frequency resistivity decrease with time is the indication of the decrease of the resistance of the ionic conductor phase. If this is constant but the diameter of the subsequent semicircle decreases, then the current increase is related to the quality of the electrode surface(s).

The good manner would be to establish a sufficiently large frequency from the impedance measurement and using this single signal for measurement only with a moderate amplitude.

Dear Keld,

Thank you so much for your response. You always answer all of my questions. I really appreciated. 

The current during the DC polarization is increasing. I did EIS and I found that the interface resistance is reducing from 300 ohm to 50 ohm. because of that the current is increasing instead of decreasing. You are right the passivation layer on the lithium surface breaks down since the interfacial resistance is reducing. I also tried to stabilize the interfacial resistance by aging for 2 days, but I see the same behavior. I also heat the Li/GPE/Li cell at 70C overnight to stabilize the interfaces, but heating increases the interfacial resistance from 300 to 1200 and I see the same behavior.

My GPE consists of 80 Wt.% (1M LiTFSI in TEGDME) and 20 Wt.% UV curable Acrylate base polymer.           

Do you have any suggestion how I could get the Li Transference number by Bruce and Vincent method? or more precisely, how could I stabilize the interface of GPE/Li?  

Dear Amir Chamaani,

did you notice any change of the mechanical or the optical properties of the GPE upon long time aging ?

Dear Ioannis,

Basically, the GPEs after UV-curing process are freestanding and transparent. I aged them between two Li foils inside the Swagelok cells, so I did not open the cells to check the colors. But I sometimes disassemble the cells and I did not see any changes in color or mechanical properties.   

Do you get a stable current (stable impedance) after long time polarisation (or maybe cycling Li back and forth between the two electrodes . safer)?

Dear Amir Chamaani,

ok.  Two EIS meas. using 2+n (n:=1 or 2,  lithium) electrodes, before and after aging, will show the type(s) of the subsequent changes.

Thanks everyone. Last update is that I found the problem is not related to the UV curable polymer. I also observed the increase in the current during the DC polarization of liquid electrolyte (1M LiTFSI in TEGDME). I think the problem is coming from the fact that the liquid electrolyte does not form stable SEI on the surface of Li metal and during the DC polarization the SEI is etching off and because of that the current is increasing.

Any suggestion?

Is it possible to :

1) show a i-t graph (t in hours) ?

2) reverse polarity (in the same cell) and show a graph, also ?