Can Theoretical Models Predict Direct Bandgap Materials for Solar Cells?

[Mayan Fung]

I am working on some research on solar cells which prefer a direct bandgap then an indirect one for the light-harvesting layer. In many works of literature, people present some measurements to tell whether the material has a direct/indirect bandgap. However, is there any criteria for a preliminary estimation on the bandgap of a material? Or in other words, is there any theory-based guess that we can make other than a simulation, eg DFT?

 

[ZapperZ]

I am working on some research on solar cells which prefer a direct bandgap then an indirect one for the light-harvesting layer. In many works of literature, people present some measurements to tell whether the material has a direct/indirect bandgap. However, is there any criteria for a preliminary estimation on the bandgap of a material? Or in other words, is there any theory-based guess that we can make other than a simulation, eg DFT?

What do you mean by "... preliminary estimation on the bandgap of a material..."? Are you asking for theoretical calculations on the magnitude of the bandgap? Or that it is a direct or indirect?

Both of those came out of band structure calculations for the material. From the band structure, you can look at the bands closest to the Fermi level, i.e. where is the highest maximum of the valence band, and where is the lowest minimum of the conduction band. If they occur at the same k-value, then it is a direct band gap material. If they do not occur at the same k-value, then it is an indirect band gap material.

But this all stems from knowing the actual band structure. To get that, one does band structure calculations using whatever technique is appropriate, including DFT.

Zz.

 

[crashcat]

You want to avoid DFT, but if your materials are crystalline you can do a low-computational-cost DFT simulation on your personal computer, just a couple hours run time, to give a first look at whether the gap might be direct or indirect. Not so easy for amorphous materials, you'd need a beefier computer or a really long run time.

I don't know of any simple rule based on crystal structure or anything to predict. But I am naive in this so that doesn't mean there isn't one.

 

[Mayan Fung]

I asked this because I am working on perovskite solar cells. Perovskite has the chemical formula of ABX3 where A is a +1 cation, B is a +2 cation and X is a -1 anion.

In this field, people like to mix different ions together. For example, they may use 1/3 Br and 2/3 I, or some 5/6 A1 cations and 1/6 A2 cations. While the compound ABX3 has a direct bandgap and compound PQR3 has a direct bandgap. I am curious why they are pretty sure that mixing A with P or mixing X with R will also result in a compound with direct band gap.

 

[ZapperZ]

I am curious why they are pretty sure that mixing A with P or mixing X with R will also result in a compound with direct band gap.

There are many version and many ways to tell this story.

It may be that they have a good guess in the beginning, test it out, work out the recipe, and then stumble upon just the right combination and processes. And then they try other variations to the recipe and discover a pattern, etc. Never discount the trial-and-error step that occurs in the beginning that later on provided guidance on what to come in the future.

This is no different than the early days of cuprate superconductors. In the early days, people were almost throwing darts at the periodic table as they discover families of compounds within the cuprate superconductors. And we still do not have a widely-accepted theoretical description of this high-Tc superconductors, but yet, we do know how to make them.

Zz.