Dear Jan,

This is an interesting question. I think that for the same reason, the only presence of cation vacancy can induce a lattice expansion so as to minimize anion repulsions in the vicinity of the vacancy, but don´t forget the defects are never alone in an ionic structure but form an equilibrium, ie if you have cation vacancies (A or B site), you have (an)other defect(s) occuring to compensate the "negative charge" corresponding to the cation vacancy (see Kroger Vink formalism). I don´t know your precise case but it could be anion vacancies (common in perovskites), cation interstitials (must be small), cation oxidation (difficult in your case as Nb is already 5+), ... and all those defects also affect the lattice (expansion or contraction...)

I hope it answers (at least a bit...)

Cheers

Gilles

I have a same question, which is related to Ba evaporatiion from my BaZrO3 pervoskites.

Whether the lattice expands or contacts by Ba evaporation

We studied this subject in MBE grown epitaxial LaFeO3 films on SrTiO3 (DOI: 10.1039/c7cp01104k)

It was found that B-site vacancies cause lattice expansion but A-site vacancies have little effect, even at >20% concentration!

I am sure this is related to the fact that the A-site (La) cation electronic states are far from the Fermi level compared to the B-sites (Fe). Thus, removal of their electron density has little effect on the electronic structure near the Fermi level (I.e. the valence electrons, which are responsible for all the macroscopic properties, INCLUDING LATTICE PARAMETER!). Since the electron bands nearest the Fermi level are relatively dominated by O and B-site cations (more generally, transition metal cations), It is logical to presume that their removing these atoms would have a larger effect of the electronic and atomic structure.

Of course, forming a La or Fe vacancy requires the corresponding number of O vacancies to maintain charge neutrality in the crystal. The other option is the Fe valence can increase to Fe4+, as is the case for Sr2+ substitution for La3+ (SrFeO3 into LaFeO3, I have several publications on this system as well). If you read my paper, you will see that we attempted to oxidize the cation deficient LaFeO3 samples using the same process we previously used to oxidize SrFeO3. But, we found no evidence that the higher oxidation state of Fe was obtained in any of the samples. This implies that all of the samples had pure Fe3+ and thus the concentration of cation vacancies was equally charge balanced by O vacancies. This also implies that the O vacancies have an insignificant (or otherwise equivalent) effect when charge balancing for La or Fe vacancies.

Size and charge on cation are important to change the lattice parameters while making vacancy.

Yes, there are several articles and I've mentioned one of them here.

Influence of A‐site nonstoichiometry on the electrical properties of BT‐BMT, R Muhammad, A Khesro - Journal of the American Ceramic Society, 2017

If you exceed the limit then symmetry will completely change.