It is just a layer of a material. Make a block or cylinder the same shape as the layer, then define a material with the properties of the glucose solution.
You may have to find out what these are from the internet or a book of physical constants, or from a scientific paper. You only need to know the relative dielectric constant and the loss tangent at or near to the frequencies you want to simulate at. The magnetic properties will be the same as for vacuum.
1.In some scientific papers, the researchers used aqueous glucose solution as analyte, which can be diffused fully in the surrounding of the array. On the other hand, in some papers, a layer of fixed thickness(45 or 50 um) is used as the measured analyte. What is the reason behind this?
As for the second method, does the analyte thickness have any effect?
2. Almost in all papers, the researchers mainly show the variation of refractive index for sensing performance. I couldn't find any paper with the relative dielectric constant and loss tangent. Could I use n = √eps at the terahertz (0.1 to 10THz) range?
If it isn't given, then very often the loss tangent is, or is assumed to be, small enough not to make any difference to the results given. You could use 0.
0.001 or lower is not uncommon in materials often used. If you can find the conductivity then you can get the loss tangent as in equations 5 and 6 below.
Article Frequency and gate voltage effects on the dielectric propert...