When i am doping Co to a ground state cubic Ni2MnGa, after full structural relaxation i am seeing that the structure does not remains in cubic anymore. But i have seen in other papers that Co doing in cubic Ni2MnGa they are giving lattice parameters of doped cubic structure. So i was wondering how to relax the doped system?do i need to relax the structure with ionic position unchanged?
Not an expert in the field of Ga alloys, though, are you sure you have one phase system? You do not specify how much Co you are doping with. Also, please provide XRD patterns with and without doping (raw data). How can anyone help you without seeing data?
Please also specify cell parameter and atomic positions for your atoms.
Hi Tapasendra Adhikary
It somewhat depends on precisely what you are trying to look at / quantities you are looking to extract.
It is very common when calculating properties such as defect formation energies to compare the energies between the reference system and the defected system with the cell fixed to that of the reference system. That is, you keep the lattice vectors of the reference system and relax only the atomic positions. There are then methods you can use to approximately correct the elastic energy from finite size effects (as an example by a collaborator of mine which uses a 1st order correction based on elasticity theory: P.W.Ma & S.L.Dudarev, Comp. Phys. Comms. 107130 (2020)). In that case it was for self interstitial types and additional care must be taken if using foreign elements depending on which method you use.
The inclusion of a dopant into your structure will cause the neighbouring atoms to move (this may be quantified by the relaxation volume for example). Depending on the chemistry and symmetry, this may be isotropic or anisotropic. In the dilute limit, there would be negligible change to the average lattice parameter at larger scales but of course in DFT calculations we expect interactions between defects and their periodic images (ie the concentration of the dopant is not dilute). As an example, adding a single additional Fe atom into BCC Fe in a crowdion configuration has a defect core with a spatial extent close to 10 atomic units which is already quite considerable even for O(N) DFT methods. If you also consider the environment surrounding the dopant, the symmetry will affect cell relaxations. For example, if the local environment of the dopant is asymmetric then I expect this would be reflected by changes in the cell parameters you found (since the forces would be imbalanced upon substitution).
Hope this is helpful,
Jacob
Tapasendra Adhikary,
In general, any dopant atom, regardless of concentration, will deform the local environment by changing bond lengths and maybe bond angles. Therefore, you should relax both the atomic positions and cell volume. Depending on the size of your supercell, it might be practical to break the relaxation into parts and cycle the relaxation procedure a few times. I recommend first relaxing the internal positions, then the volume in a separate run. These two relaxations should be repeated a few times.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Metals
- Alloys