to add to previous comments:

This is a good question: p-type silicon (substrate) is much more used for those applications (probably it is cheaper and easier to produce and making metallic contacts to the n-type side, ...), but indeed the best would be to use larger n-type substrate (and thin p-type for the junction) to avoid any recombination of the photo-generated carriers (electrons) in the p-type region.

Dear Poshan,

The p-type silicon substrate is the solar radiation absorbing material in such cells. In order to absorb the longer wavelength the thickness of the material must be greater than 1/alpha where alpha is the absorption coefficient. As silicon is indirect gap material its alpha is relatively small and one needs about 250 um thickness for the radiation near the cut off of Si. The conductive oxide layer on p-Si acts as an electron transport layer.

So it must be of wideband gap and thin enough to pass most of solar radiation to the underlying p-Si. It must also be thin to decrease its resistance to the passage of the electrons to cathode.

Best wishes

If this oxide is used as a protective layer for the semiconductor then it makes sense to have it thin so that you won't lose photogenerated carriers that have to go through it. It's hard to tell without seeing the papers you discuss.

In some cases the n-type conductivity of STO is a bug and not a feature (despite being presented as such), which is the common result of insufficient oxygen stoichiometry.

to add to previous comments:

This is a good question: p-type silicon (substrate) is much more used for those applications (probably it is cheaper and easier to produce and making metallic contacts to the n-type side, ...), but indeed the best would be to use larger n-type substrate (and thin p-type for the junction) to avoid any recombination of the photo-generated carriers (electrons) in the p-type region.