Have you checked already the link given below?

(Seyedali Emami, Luísa Andrade, Adélio Mendes: "Recent progress in long-term stability of perovskite solar cells", U.Porto Journal of Engineering, 2015)

http://feupedicoes.fe.up.pt/journals/index.php/upjeng/article/download/132/142

Dear Jan-Martin Wagner,

Thank you! The article is very good. But it does not show in more detail how the charge separation occur. The paper shows an easy to read energy level diagram. By which I could understand that the energy level on perovskite material is higher than in the TiO2 material (ETM).

Sorry my lack of knowledge, but by what I understand, this means that electrons will tend to stick to the TiO2 material since is the lower energy level. So, moving electrons will "statistically" be more present in TiO2 than in the Perovskite as they are generated, caused by the difference on the materials energy levels. The opposite happens to the Perovskite/HTM interface.

The electric field present on PN interfaces in c-Si cells does not take place on Perovskite solar cells.

Am I on the right track or this conclusions are wrong?

Again, thank you for your much appreciated help!

I am new, but very interested, on the field.

Oh, no trouble at all -- I just entered "Perovskite Solar Cells (e.g. TiO2/CH3NH3PbI3/Spiro-oMeTAD)" from your question (without the quotes!) into the standard https://startpage.com/ search engine, and the link I provided was nothing but the first hit...

Well, I may be called an expert for standard bulk silicon solar cells, but I haven't dealt much with perovskite-based cells so far. Interestingly, the band diagram shown in Fig. 2 of the article shows a similarity with that of so-abbreviated HIT cells (cf. https://www.youtube.com/watch?v=FfU9jxGnYzs), namely in that the band offsets of conduction and valence band are completely different: one of them is very large, the other is quite small; ideally, a so-called type-II heterojunction is formed, having a staggered band alignment (cf. https://en.wikipedia.org/wiki/Heterojunction#Energy_band_alignment).

You're completely right, in Fig. 2 there is no electric field that brings about the charge separation, it's all in the alignment of the energy levels that influences the charge carrier distribution -- in exactly the way you described it. This nicely underlines a fact that Peter Wuerfel likes to point out in some of his works and textbooks, that in general there is no need for an electric field in order to accomplish charge separation in a solar cell.

I think it's noteworthy that already for the Perovskite absorber material there is a fundamental difference compared to standard c-Si solar cells (and also HIT cells): Bulk silicon is always doped, since otherwise its conductivity would be way too poor. Therefore, in silicon one always has to distinguish between majority and minority carriers -- which is irrelevant here since, as far as I can see, the Perovskite absorber material is undoped. One of the consequences is that, due to the lack of free charge carriers, no electric field is established at the interfaces with the neighboring layers.

However, for me the lack of free charge carriers raises the questions about the charge transport mechanism inside the Perovskite and -- a strongly related aspect -- about the maximum working thickness of the Perovskite layer. And if the latter is highly restricted due to the former, the next question is about the light absorption strength of a quite thin Perovskite layer and, therefore, about the maximum obtainable photocurrent. But as I said, I'm not an expert in this field...

Thanks for the explanation. I am very glad I understand correctly the concept behind the energy levels and the charge separation.

In the beginning of my readings I though the excitons should be dissociated by an electric field as well, but reading the following article, it is clear that most of the pairs generated by incident photons are actually free charges.

https://dr.ntu.edu.sg/bitstream/handle/10220/25121/The%20photophysics%20of%20perovskite%20solar%20cells.pdf

So, as you intelligently pointed: no need of electric field.

The questions you proposed are very interesting and a good challenge for my studies. I will try to find more articles and some light for those questions. Until now, what I could find is that the maximum thickness is related to the diffusion length. In CH3NH3PbI3 were found diffusion length >1000nm (article above), but as shown on the following article, in a working solar cell, depending on the architecture used, this length can be much smaller (few hundreds nm).

https://www.researchgate.net/publication/259625678_General_Working_Principles_of_CH3NH3PbX3_Perovskite_Solar_Cells

Maybe the light absorption can be related too, as you have pointed. But I did not find anything about it yet.

If you are interested on perovskite as well, we can cooperate and maybe exchanged articles and ideas.

All the best!

Article General Working Principles of CH3NH3PbX3 Perovskite Solar Cells

Dear Marcelo Alatzatianou Rodrigues  and Jan-Martin Wagner 

The working mechanism of perovskite solar cells it's an issue for me as well. It's seems not be a pacified subject. See this paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4598624/

There in 2009, the perovskite solar cells research was restricted to the groups working with sensitized solar cells or organic solar cells. So, the first works had assume the perovskite as a sensitizer and employed models similar to that applied in sensitized solar cells. 

It's well established that the sensitized solar cells, or Grätzel cells, exhibits a mean null built-in electric field - the separation and transport of charge it's explained by the favorable electronic level alignment in the interfaces of materials (see attached file - Gratzel ). The Fermi's golden rule models approach are used to calculate the interfacial electron transfer rates. 

The perovskite material contrasts with the previous known "sensitizers". The charge carrier relatively longer diffusion lengths are one of the main differences. This crucial aspect lead to acceptance of the p-i-n model as the most suitable, at a first glance.

Adding to this, we have the ionic electronic nature of carrier transport in perovskite solar cells ( that should be taken in account to explain the hysteresis behavior seen in the I-V solar cell curve) - see article attached.

Hope we can discuss it!

Best Regards,

Nuno

Dear Respected colleagues,

Charge separation in perovskite solar cells occurs by the built in electric field in the intrinsic perovskite layer. This built in electric field is due to the contact difference of potential between the electron and the hole transport layers. This is assuming that the metallic electrodes form ohmic contacts with the transport layers.

Therefore, the transport mechanism of the short circuit current in the perovskite solar cells is similar to the transport in the pin solar cells. 

This is because the perovskites, on absorbing photons build excitons which are loosely bound like the metallic semiconductors for example silicon. Therefore, once generated they will be dissociated because of the thermal energy. 

The diffusion length of the minority carries will be of the order of one micrometer depending on the quality of the perovskite material, which is sufficient to collect the photogenerated carriers in an equal thickness of the material.

The carrier transport layers conduct the the carriers by drift and so they require some voltage drop to conduct their current. 

Therefore, their conductivity must be high enough to reduce their resistance.

Respected Abdelhalim abdelnaby Zekry sir

How does charge separation occur in HTM (hole transport material) free perovskite solar cells? also can you please explain how internal electric field is generated in perovskite material.

In a very simple way

In perovskite solar cell, HTM-Perovskite-ETM forms a p-i-n structure. Due to the work-fuction difference there exists a potential difference across the perovskite layer, just like the depletion region of a p-n junction. Think it as large depletion region in a p-n junction. So the charge gets separated at the same way as in conventional solar cell.

In photodetectos such p-i-n (n-i-p) structures are widely used.

Md. Sazzadur Rahman Sir thanks for your explanation.

Dear Srest,

Hope every thing is well with you and your country,

As it is said by the the colleague Md., like exactly what occurs in the pn junction or the pin diode there is contact potential difference between any two different materials. There is a built in electric field associated with this contact potential difference. Here the contact difference of potential will be between the ETL and the HTL.

As this a conceptual questions i would like that you refer to the following references:

Conference Paper Generic Analytical Models for Organic and Perovskite Solar Cells

and Book Electronic Devices

You will understand well the contact difference of potential and the associated built in field in various types of junctions.

If you find them useful it will be good if you recommend them for your students and your colleagues.

Best wishes

Abdelhalim abdelnaby Zekry Sir

Thanks for your concern sir. we are trying to deal with this COVID situation.

Hope you and your family members are safe and sound.

Also thank you for your reply to my query. Your article is great!

Hi

This article may help you:

https://www.nature.com/articles/ncomms9397