Why is P-Type Conductivity Limited in Transparent Conducting Oxides?

[bilalarif]

I was reading about semiconductor transparent conducting oxides and I found this why p-type conductivity is limited in TCOs. But I could not understand its meaning please someone explain me how the holes are localized at valance band edge and how dominated by oxygen- 2p levels?

''Owing to the strong ionic nature of metal-oxygen bonding, holes are typically localized at the valence band
edge, which is dominated by oxygen-2p levels therefore limiting p-type conduction. Two methods have been
suggested to enhance the co-valency between metal oxygen bonding, thereby limiting localization:¹ choosing cations having closed d-shells of energy comparable to that of the oxygen-2p levels (i.e., Cu+, Ag+, and Au+, especially when found in linear coordination with oxygen²), and choosing a structure in which oxygen adopts tetrahedral coordination''

 

[Useful nucleus]

I do not think this statement is very precise. If holes tend to be localized, they would be localized in a gap state close to but not right at the valence band edge. It is also not true that hole localization limits the p-type conductivity. Take for example p-type conduction in perovskite oxides. In many cases it proceeds via hole conduction and these holes are self-trapped on oxygen sites and are derived from 2p states of oxygen.

Could you cite the source paper of this statement?

 

[bilalarif]

here is that paper

 

[Useful nucleus]

I think the sentence highlighted in red encompasses too many information and in fact recent work on hole-self-trapping brought new insights beyond 2004 at the time of writing this paper. In fact even the concept that covalency of metal-oxygen bond reduces self-trapping is questionable. Take for example rare-earth nickelates. The bonding between Ni and O has a clear covalent character, in spite of this hole trapping on oxygen occurs.