Predicting the "Path of the Carrier" in ITO gratings in Solar Cell

charge
solarcell
chargedistribution
device
chargeprofile

#1

I am currently researching in " Modelling of Nanostructured ITO(Indium Titanium Oxide) Surfaces for increasing the efficiency of organic solar cells". I am currently trying to understand the impact of grating on the electrical properties of material by varying the period, aspect ratio and thickness of the grating structure.

I want to see the path the carrier takes inside the ITO layer and inside the grating and how it varies with varying aspect ratio and period of the grating structure. I have been using the CHARGE result view to try to predict this but have been unable to do so.

Also, since ITO is not available in the material list, it has been defined as a metal with work function of 4.7ev. Since, ITO has various semiconductor properties, the results produced have not been of that much use. In such a case if you could give guidance as to what type material can be used in Device to replicate the properties of an ITO.

The link to the design in DEVICE is
https://drive.google.com/file/d/0B42M3soVszILYjduZGpSamp6YWM/view?usp=sharing


Electrostatic Potential Convergence Problem
Predicting Charge flow in Device
#2

Hi @Arastu, Looking at your file I can see that you are using ITO as the bottom contact in your device. In such a case the solver requires that ITO be defined as a metal. Even if you want to define it as a semiconductor to study the electron transport through it, the solver will treat it as a metal. The workaround could be to place a metallic layer at the bottom of ITO and set that to be the bottom contact. You can then define ITO as a large bandgap semiconductor and then heavily dope it. This will allow you to look at the transport through ITO since the CHARGE solver models charge transport through semiconductors only.

A few things to consider:

  1. Achieving convergence is always difficult when running a simulation with large bandgap semiconductors and chances are that with ITO defined as a large bandgap semiconductor the simulation will be prone to diverging. However I believe that given what you are trying to do, this is probably the best option.

  2. When you dope the ITO specially at the grating, you will have to be careful to ensure that the doping gets applied to ITO only and not into the InN layer. Also keep in mind that the doping objects are additive so the doping from the “bottom diffusion” doping object as it is right now will get added to the doping of the ITO grating. You will have to set the doping values appropriately to ensure that you have the right profile in the end.

  3. On the point of how to define ITO as a semiconductor please take a look at this KB page: https://kb.lumerical.com/en/index.html?materials_creating_new_semiconductor.html. Since we do not have ITO as a default material, you will have to look at published articles to get the material properties in order to create the material model.


#3

I have simplified the design a bit (attached below). I am trying to predict the effect of charge flow and mobility with grating in the ITO structure. I am using InN as an semiconductor to replicate the properties of ITO in for use in organic solar cells for diffraction of light. I just want to ask about the method through which I can get see the charge flow and get the effect on resistance and charge mobility.

My design is attached
https://drive.google.com/file/d/0B42M3soVszILemhWc1BmQ2dCdG8/view?usp=sharing


#4

Hi @Arastu, looking at your file I noticed that the grating are not included in the simulation region. When I looked at the top view of the design I saw that the y-position of the solver region was wrong. When I moved the solver region to include the gratings in the 2D simulation region, the effect of the grating can be simulated.

After I you run the simulation, you can then visualize the current density (Jn and Jp) in semiconductor region. This result is available in the “charge” dataset which is a result available in the solver region

From the current density plot you can see the where the current is flowing through your structure and where the density is high or low. Please let me know if this is what you were looking for.


#5

Hi @aalam, Thanks for the help. I tried the same method as yours, on the same design. Due to some reason I am not getting the charge density output that you are getting. I changed the simulation region so thats not a problem anymore. Can you check why could this be happening.

This is the same design with corrected simulation region
https://drive.google.com/file/d/0B42M3soVszILTll4YlVoUWtsdlU/view?usp=sharing


#6

Hi @Arastu, My mistake; I forgot to mention that the grating was getting overwritten by the ITO layer in your original file. I manually made the mesh order of the grating a smaller number (1) to ensure that they are getting applied.

Modify your script for grating2 accordingly and the gratings will get applied. You will then be able to see the proper structure and the corresponding current density.


#7

Thanks for the help @aalam.

The solution worked and I can see the charge density now, but I think I am still doing something, because the density is not what I was expecting it to be.

So I am using contacts on the top corners now, but I don’t think effects anything. This is the XZ view of the design

The diagrams below are the zoomed view of the Jn charge density image and the square scale version.

The fact that the ITO parts between the gratings are blue shows that there is less charge density and it makes sense because more charge would flow in the bottom half of un-grated ITO. Now the unexplainable issue with the result that I am seeing is that we have blue lines even in the bottom half, which should theoretically not depend on the grating. I was expecting the whole bottom half would have same density and the part between the gratings in the upper half would be of less density.


#8

Hi Aratsu, I believe the current density is showing what one would expect. I have prepared the figure below to show this.


#10

Hi @aalam,

I was trying to predict the current density from calculating it from the resistivity in the silicon substrate. Since, I am using the silicon material in the default material list, can you see how I would be able to calculate the resistivity since I haven’t introduced Doping in the material.

Also, I tried introducing doping in the design and doing that changes the current density and the recombination graph. I think that my doping is wrong, it would be great if you could check it.
The design is attached.
https://drive.google.com/file/d/0B42M3soVszILaV9xUjhaWC1XcW8/view?usp=sharing

The following are the different results of current density, I am getting for doped and un-doped design.


#11

Thanks for sharing the file. Looking at your design I do thing that the doping objects needed some modifications.

  1. The diffusion doping objects were placed much higher into the metal. Since in a diffusion doping object the doping drops near the bottom (and the edges) I pushed them down so that the silicon under the metal gets doped properly.

  1. The source face in the “bottom diffusion” object was set to “lower z”. However since it is being used to doped the top surface of silicon I changed the source face to “upper z”.

  1. One final thing I noticed is that you have used different types of dopants on the two contacts. So basically you now have a kind on p-i-n device. The current density of such a device cannot be calculated simply by using the resistance of the low doped region. I am guessing that this was unintentional and that you wanted to make both the diffusion objects p-type simply to ensure that you have ohmic contacts with the metal. However since I am not sure I did not change the dopant type.

Here is the modified file: IOT_version1_MOD_v2.ldev (6.6 MB). The doping should work fine now with the changes I have mentioned above.


#12

Thanks for all the help @aalam,

The changes in diffusion region worked and I have got some results that were expected.
In the same design, I have disabled both the doping regions near the contact and only have the constant doping enabled. Now I am trying to calculate the Jsc current due to illumination, and therefore now I have also enabled the Bulk generation in the figure, though I haven’t imported the generation from FTDT, I would still assume it should work.
But because I am getting a very small value of Jsc, I think I have done something wrong in the Bulk generation.
I am using the script to get the Jsc
getdata(“CHARGE”,“emitter.I”)/(getnamed(‘CHARGE’,‘norm length’)*getnamed(“CHARGE”,“x span”));

https://drive.google.com/file/d/0B42M3soVszILNDJnQ1RmaVpHVTQ/view?usp=sharing


#13

Hi @Arastu, I haven’t taken a look at your file yet because I think from your description I have an idea as to why your Jsc value is small. In order to collect photo-generated carriers effectively you need an electric field inside your device which will separate the electrons and holes and drive them towards opposite contacts. This is normally done with the built in electric field of a pn junction. In your device since you only have a constant doping, you do not have this electric field. The photo-generated electrons and holes will diffuse to both contacts randomly and the current due to electrons and holes will cancel each other. As a result you will not get a very small value for the net short-circuit current. To see a reasonable short-circuit current value create a pn-junction inside your device.


#14

A post was split to a new topic: How to apply constant doping to a material