I wanted to solve the penetration of electric field lines into a hetero-structure thin film in COMSOL and see that how much of field can penetrate into substrate (image is attached) . I am using AC voltage.
COMSOL is a very powerful software but is so specific. Also, you system seems to be so simple, I mean in the geometrical aspects at least. I think it would be easier and even more interesting if you could model it using simpler concepts such as impedances in series and parallel. You could increment your model adding distance dependent resistances.
Well, I believe it would be more elegant and maybe easier than COMSOL.
The structure you presented has three layers, the substrate SrTiO3 which is an insulator with ferroelectric characteristocs, NiO which is a wide gap semiconductor which is normally p type with relatively low resistivity, then finally two gold contacts at he top of the NiO.
You applied an AC voltage between the two electrodes and you say that you want to determine the electric filed penetration in the insulating substrate.
At zero frequency, that a DC voltage the current will path completely in the conductive layer of NiO. As the frequency increases, in addition to the conduction current in the NiO layer, there will be a capacitive current in the substrate shunting the conductive path in NiO. As the frequency increases the capacities current increases while the conductive current remains constant. You can model the device by a conductance G in parallel with the substrate capacitance C. The current I= V Y, the admittance Y= G+j wC,
C represents the electric field penetration in the the substrate.
The electric field theoretically speaking will penetrate through the whole substrate since the material is dielectric.
Therefore i could not see what is the objectives of your simulation?
Field effect on different types of materials is well known.
Surely the reason to do a Finite Element Analysis simulation with software like COMSOL is to model the 3D (or 2D) distribution of the electric fields and shielding currents. Simple approaches in terms of impedances are all very well if all you need to know is the integrated currents and fields as a function of time and frequency (which is all you need to know for many applications) but will not give you insights into the current and field spatial distributions - for example if there are going to be any local heating effects near corners and whether the gold concentration distribution where the gold has diffused into the oxide during formation of the ohmic contacts is critical.
I would suggest you model this using simple impedance calculations first, to give some sense of how the impedance varies with frequency and how the field and current density vary through the oxide thickness, ignoring any 2D effects. Then if there are any areas of concern, perform a more detailed 2D or 3D analysis using COMSOL (very quick and straightforward if you are familiar with COMSOL) and use these to optimize the geometry (channel length, area of ohmic contacts, oxide thickness etc.)
The whole point of software like COMSOL is that it allows more complex 2D and 3D analyses to be performed as quickly as simple back-of-the-envelope calcs. The latter are good for giving quick insights and for scoping problems, the former give more realistic and quantitative results.