Modeling Reverse Bias Breakdown in PN junctions

[dmora97]

Hello!

I want to know if it's possible to use Oghma to model the reverse bias breakdown of a PN junction in a diode consisting of ITO/PTAA/C60/Ag? I.e., Is it possible to model avalanche or Zener breakdown in a simple pn junction so that i can see how these materials breakdown in reverse bias? I eventually plan to add a perovskite interlayer afterwards and allow holes tunneling from ETL into the perovskite and look at the impact of mobile ions within the system.

Sorry for all the chatter. Learning as much as i can in this short time frame...

Thanks in advance!

[ezzrm]

Hi There,
So the quick answer is not at the moment, but I could put it in quite easily (depending when you wanted it for). Currently you can only have Zener effects (Quantum Tunneling) at interfaces, but I think you would need it across the device at each mesh point.

However, before going in this direction my question would be: Are you sure you would have Zener Breakdown which should be reservable and not some runaway effect such as Avalanche Breakdown leading to thermal decay of the device. My feeling is that due to the very low thermal conductivity of these materials the device would heat and degrade quite quickly so you would not be able to go into the Zener region very far or for very long. I think before doing some modelling, I would want to do a JV sweel (+ve to -ve (breakdown)), then redo the measurement and see if the device has changed.

Let me know what you think, happy to add stuff to the model.

Best,
Rod

[dmora97]

Hi Rod,

Thanks for the quick response! We suspect that the breakdown in PSC's is caused by holes tunneling from the ETL into the perovskite's valence band at sufficient -V's (thanks to the mobile ions screening the field at the perovskite interfaces!). I'm interested in modeling this behavior of tunneling at the ETL/PVSK interface that causes current to flow. This type of modeling would be huge to accomplish.

I've lately been using the oghmanano software to look at band diagrams and calculate electric fields at the interface for now since I can't seem to get a PSC to "break down" in reverse bias (for now... ).

There are obviously other electrochemical processes occurring in the perovskite under reverse bias, so i dont expect the software to be able to accurately capture this behavior. There are also likely thermal breakdowns occurring in our real cells and it would be incredible to use this software to look at the Joule heating across a pn diode under reverse bias (is this already possible?).

Basically, modeling the tunneling of holes from the ETL into the perovskite would be amazing if i can see a PSC breakdown and pass current in reverse bias. My second goal is to study the breakdown of normal PN diodes via Zener tunneling at sufficient doping densities or show that the electric field strengths are strong enough to cause Joule heating/dielectric breakdown. I'm struggling to get current to flow in revB with the basic pn diodes for now, and I'd like to know your thoughts on whether modeling this breakdown is possible! I can provide a DOI on a paper for the breakdown of PSC's if that would help guide your thought process!

Thanks

[ezzrm]

Hi there,
In which case it should be able to do that. In the electrical tab, there is a button called interfaces. You can set tunnelling there. There is also a "heterojunction tunnelling example" in the new simulation window. If you have any issues, let me know. At a push I could make a perovskite example for you if you get stuck....
Best,
Rod

[dmora97]

Hi Rod,

I played around with the tunneling example yesterday and set the applied bias from -15 V --> +2 V with the illumination at both 1 & 0 suns. I was able to observe what was expected in forward bias, but in reverse bias I couldn't seem to simulate the breakdown /tunneling of carriers to get reverse current from the heterojunction. however, at 1 sun, I saw what looked like a breakdown (a "knee" around -1V with gradual decaying current), but I couldn't quite get it to do the same in the dark. Is there something I should be doing to see the tunneling breakdown in the dark at reverse biases using this example so i can translate it to a full PSC down the road?

Thanks!

[ezzrm]

Hi,
So I could not quite remember what I had used for tunnelling terms in the model. So I had a look yesterday at the code and the manual that I wrote a couple of years ago, and the terms I put in were for tunnelling between two organic semiconductors, where I assumed charge could percolate from the lower energy side to the higher energy side. So it's more an extra drift term rather than tunnelling per se. As you have an interface with a 'hard' semiconductor I think you will need the classical tunnelling terms.

I will have a think this week if I can include it, it might be quite easy to do but I just need to think how to approach it.

So the question now is, which tunnelling term is appropriate?:

See the table under section 4.16.3.1:
https://www.sciencedirect.com/topics/ch ... -tunneling

Best,
Rod

[ezzrm]

I'm having a go at putting all the physical tunneling mechanisms in the table, into the model. Let's see how far I get... R

[ezzrm]

Hi,
So I've added 'Direct tunnelling' to the model now in version 8.1.009, it is described in section 9.9.1 of the manual. There is also an example under the New simulations->tunneling->Direct tunneling example. I've half added Fowler-Nordheim/Thermionic emission and Hopping conduction. But not finished them. I suspect Direct tunnelling is the one you need though. If you run the example sim with "Direct tunneling" turned off/on you will see the difference.
Best,
Rod

[dmora97]

Thanks so much! I'll take a look at it.