There is conceptual answer for your question. It is so that the as deposited perovskite acts as an intrinsic semicondcutor. So, in order to wok as an efficient solar cell one has to sandwich it between the HTL with high hole concentration and ETL with high electron concentration. Thus it is so a pin diode structure. It is then that one can add the metal electrodes to the HTL and the electron transport layer to electrically access the solar cell. The formed metal semicondcutor contacts will be tunneling and behaves low ohmic.
In case you remove the ETL and the HTL you form then the socalled Schottky diode solar cell where one side must be rectifying and the other side must be ohmic. Such Schottky solar cells are normally less efficient than the pin diode structure.
An other solution is to dope the perovskite to form homo p-n junction and then cotact the pn junction by metal from two sides acting as ohmic contacts.
please tell me if there is any report of etl and htl free psc?
if fto is used as one electrode then such a psc will be like n+/i diode at FTO/perovskite interface(normally the thickness of FTO electrode would be few hundred nm so it can be modeled as highly doped n type semiconductor).
i have simulated such a psc and found it to be solar cell with moderately high PCE. effect of ETL was found to be minimal and effect of HTL was much more profound.
What you found is right. If you made your perovskite p-type, the HTL will be also ineffective. It is so that the FTO performs as an ETL and metal electrode at the same time.
So, what you found is plausible.
I think when you search the literature you will find the effect of the thickness of the ETL and the HTL.
But one thing is very important. One does not seek the functionality but one seeks the highest possible efficiency. In addition, These layers have also protective functions against the environment.