i actually want to calculate band structure of SrNbO3.5 with quantum espresso, but i have some trouble with its atomic position. my colleague advice me to use Sr2Nb2O7, but i doubt with the result i get will the same with SrNbO3.5. so, here my question:
does SrNbO3.5 have same properties (e.g band structure) with Sr2Nb2O7?
Yes, you should use Sr2Nb2O7 for SrNbO3.5. The second chemical formula is just a normalized or reduced version of the first. A periodic crystal primitive cell can not have fractional numbers of atoms (even if we write it's formula that way). The reason we sometimes write in this manner is to show how oxygen rich or poor the phase is. For example, in the niobium oxide phases the principle phases are NbO, Nb2O3, NbO2, and Nb2O5. Additional phases like Nb12O29, for example, also exist. Typically these other phases will be written as NbO2.5 and NbO2.417 in order to show their relative oxidation to the other phases. However, I can't actually specify a calculation with only one Nb and two and a half oxygen for a crystalline system, I have to use a multiple of Nb2O5 (more on this below).
Furthermore, having to use a larger unit cell than just the principle formula unit is common. For example, in NbO2, the high temperature rutile phase actually has two formula units (e.g. could be thought of as Nb2O4) and the primitive cell of the low temperature distorted phase has 16 formula unites (e.g. can be though of as Nb16O32) due to their structural symmetries. So it is not uncommon to use a cell that has more atoms than the simply formula unit notation.
Hi Andrew, I am sorry but from my point of view you've overseen something :-). At least one of your statement is incorrect. There is no problem with a description of fractional atoms since they appear not that rarely. Sometimes a position is occupied which has crystallographically a higher number than the chemical formula enables. Just recently I dealt with Ag2S (Argentite), or just use Martensite. Therefore, you have so-called site occupation factors. In your definition they would be always 1 but the aren't. There is certainly a difference between the chemically (ideal) description and non-stochiometry which ends in examples like NbO or NbO.995. But this has primarily nothing to do with the crystallographic description (as you pointed out correctly). Therefore, crystallographers define the unit cell and then give the number of formula units Z, Z=2 in the upper example. It depends on what you want to do. As you wrote correctly, if the lattice parameters are related to Z=2 it is not allowed to use SrNbO3.5 since this does not match. In case of a centered lattice (A, B, C, I) you possibly could create a primitive cell which would use SrNbO3.5. Then it would describe O with some incomplete occupation. Therefore, one should be careful to make such statements. One has to investigate the corresponding phase description and how the lattice is defined (if there is any lattice description necessary). If it is, one need to use exactly the chemical composition which fits to the respective unit cell.
Dear Ms. Ivanova
the space group of Sr2Nb2O7 is Cmc2. It's perovskite-type.
So, should i use Sr2Nb2O7 for SrNbO3.5 ?
p.s.: i also attach the paper which explain crystalstructure of Sr2Nb2O7.
- Badges
- Science topic
- Similar topics
- Physics
- Condensed Matter Physics