Angelo, 

I am not sure if you have seen this paper before " Impedance techniques at partially blocked electrodes by scale deposition C. Deslouis , C. Gabrielli , M. Keddam , A. Khalil, R. Rosset , B. Tribollet, M. Zidoune". This work provides an excellent background and practical information.

Best regards

Juliano 

Can you given details of your materials and frequency range in which you are interested.

In general, I work without BE because I am just nterested in the conductivity effects.

Best regards,

Hilda

@ Ravindra Dhar: a little bit of my lab setup and materials: I will be testing diverse soil samples, including permafrost. Temperature range: -30C to 50C. Water content: 0 to 50%. Frequency range for multiple instrumentation: 10Hz to 2GHz. Frequency range of interest associated to electrode polarization at 2T-EIS (parallel plate and cylindrical dielectric cells): 10Hz to 200kHz.

@ Juliano Katrib: thank you for the hint. I didn't know this specific papers while I read many other from some of the authors. I am downloading the mentioned paper.

@ H. Mercado-Uribe: I also achieved excellent results of conductivity measurements without the impact of electrode polarization effects in soil samples, but the issues are mainly related to the measurements of the real part of the relative permittivity. Moreover, I developed a technique to quantify the EP impedance and now I am in need  of a sort of "orthogonal" measurement technique in order to understand the effectiveness of what I have been working.

I happened to co-author 2 recent publications, that adopt alternative approaches to the use of blocking electrodes.  Perhaps of some help, if you compare their results and conclusions.

1  "Dielectric relaxation characteristics of chemically reduced graphite oxide," Carbon 95 (2015) 374-9;

2  "Dielectric relaxation and hopping conduction in reduced graphite oxide," J. Appl. Phys. 119, 224102 (2016).

BE is not good idea as it adds new complications. Generally above 10 Hz, data should be free from parasitic effects but in your case high content of water is problem. You will definitely be able to measure conductivity but with increase ionic part which will subsequently enhance permittivity. You may numerically remove it. In your case you should be least interested about permittivity. See attached papers.