2019 Fiscal Year Research-status Report
Testing and enhancement of computational DFT+U method with magnetic exchange correction
| Project/Area Number |
18K05031
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
Shishkin Maxim 京都大学, 実験と理論計算科学のインタープレイによる触媒・電池の元素戦略研究拠点ユニット, 特定研究員 (20793011)
|
|---|
| Project Period (FY) |
2018-04-01 – 2021-03-31
|
| Keywords | Hubbard corrections / Redox potentials / Double counting / magnetic moments |
|---|
| Outline of Annual Research Achievements |
I have completed implementation of extended DFT+U functional, which contains corrections for interactions between the electrons with opposite spins (formerly called magnetic exchange corrections). This approach properly treats double counting corrections due to additional term and also allows evaluation of forces on atoms. The Hamiltonian has been derived and corrections of too narrow band gaps have been introduced using scissor operator approach, implemented in PAW formalism. Then, the tests have been performed on several Fe-based sulfates and Fe oxides. Additionally, it has been demonstrated that the approach of Lichtenstein does not provide higher accuracy as compared to proposed extension of DFT+U method, when applied to calculation of cation magnetic moments and energetics of studied materials.
Subsequently, I have also applied parameterized DFT+U approach to evaluation of voltage profile of NaMnO2 cathode in conjunction with cluster expansion method. The cluster expansion has been extended for treatment of antiferromagentic ordering, which however requires larger cells to sustain translational symmetry. This work have shown that parameterized DFT+U calculations at least in some cases (e.g. NaMnO2) provide more accurate results as compared to calculations with a fixed U parameter. More recently, I have also applied extended functional to Mn oxides, testing its accuracy for evaluation of energies of formation and magnetic moments. The proposed extension of DFT+U method was found necessary to provide accurate magnetic moments, for instance in MnO2.
|
| Current Status of Research Progress |
Current Status of Research Progress
2: Research has progressed on the whole more than it was originally planned.
Reason
Two milestones, namely implementation of proposed extended DFT+U functional and comparison with the method of Lichtenstein have been addressed. In addition, application of DFT+U method, parameterized by the linear response in conjunction with cluster expansion method has been accomplished. Application of extended functional to other materials, e.g. Mn oxides is being studied at present.
|
| Strategy for Future Research Activity |
The future work will include application of formulated extended DFT+U functional to the analysis of energy of formation of strongly correlated materials, evaluation of magnetic moments of their transition metal cations and possibly evaluation of redox potentials. Mn oxides will be studied, although oxides of other transition metals could be studied too. Poor ability of linear response approach in case of analysis of conducting materials will be also addressed and attempts to resolve this deficiency will be made.
|
| Causes of Carryover |
The cluster needs to be purchased at this point in time as the necessary implementations and testings have been completed. Now these are ready to be applied for the analysis of materials properties. For this reason the cluster machine is required.
In the future it is planned to: 1) purchase the cluster computer; 2) perform 1 international visit to the other scientific group; 3) attend international conference (e.g. PSI-k conference). It should be noted that due to travel restrictions, caused by a corona virus, travels might be also postponed to the next year.
|
Research Products
(4 results)
