Boundary condition
- Periodicity, Dirichlet, and von Neumann open boundaries
RESCU (Real Space Electronic Structure CalcUlator) is a KS-DFT calculation software that can study very large systems with just a small computer. It combines the respective advantages of plane waves, linear combinations of atomic orbitals, variable-precision meshes, and Chebyshev filtering algorithms. Through optimizations in multiple numerical algorithms, it significantly reduces the burden of traditional DFT calculations while ensuring the accuracy of DFT calculations, making DFT calculations for systems containing tens of thousands of atoms possible. It is a landmark new approach to solving the KS-DFT problem of very large systems. RESCU is currently widely applied in various fields such as metals, semiconductors, insulators, liquids, DNA, 1D, 2D, 3D, surfaces, molecules, magnetic, non-magnetic, impurities, solids, etc.
Main Functions
Exchange correlation functional
- LDA, GGA, mBJ, and hybrid functionals HSE support over a hundred functionals in the libxc library
DFT+U method for strongly correlated systems
- DFT+U method for strongly correlated systems
Electronic structure
- Band unfolding, band projection, band decomposition, projected density of states, local density of states, Mariken charge distribution
Magnetic properties
- Linear spin, nonlinear spin, spin-orbit coupling
Structural relaxation
- Steepest descent method, conjugate gradient method, quasi-Newton method
Mechanical properties
- Force analysis of equation of state and phonon spectrum (frozen phonon method)
DFPT
- Dielectric tensor, optical properties, phonon calculation, Raman tensor and intensity, electron-phonon coupling
Software Highlights
A larger system
- RESCU generates highly efficient initial Hilbert subspaces by using numerical atomic orbitals (NAO) and employs Chebyshev filtering to completely avoid the process of solving eigenvalues in the entire Hilbert space to precisely solve the KS equation.
- By optimizing the algorithm for solving the KS equation, RESCU can reduce the computational load that was originally proportional to the cube of the system size. In the test, RESCU's computational load was only 2.3 to the power of the system size when using the real-space discrete lattice basis set to calculate a system with more than 6,000 atoms.
- When calculating large systems, using real-space discrete lattice basis vectors can achieve the same level of accuracy as other standard plane wave basis vector KS-DFT software, while using NAO basis vectors maintains the same level of accuracy as other standard LCAO basis vector KS-DFT software.
Higher efficiency
- When the system contains tens of thousands of electrons, RESCU adopts a brand-new partial Rayleigh-Ritz algorithm to ensure extremely high computational efficiency. It features high parallel efficiency and excellent scalability, achieving a nearly linear speedup ratio even when there are a large number of cores.
Application Cases
Moire pattern simulation
1D Moire pattern simulation
ref. Physical Review Letters, 121(18): 186403(2018)
2D Moire pattern simulation
ref. Physical Review Materials, 1(6): 061003(2017)
Carrier mobility
Simulation of carrier mobility at different vacancy concentrations
ref. 2D Materials.4(4 045014(2017)
Simulation of carrier migration in twisted graphene
ret. Physical Review B. 96(19): 195406(2017)
Response theory
Phonon spectrum calculation
ref. Physical Review B,99(6): 064302 (2019)
Berry rate calculation
ref. Physica E: Low-dimensional Systems and Nanostructures, 126: 114390(2021)
Electronic structure calculation
Simulation of density of states and charge transfer
ref. Nature Communications, 9(1): 193(2018)
H2O Molecular dipole analysis and differential charge analysis
ref. npj Computational Materials,7(1): 1-9(2021)
Energy band simulation of carbon nanotubes
ref. Nano Letters, 19(6): 4146-4150(2019)
Removal band calculation of low-concentration doped dense Brillouin region
ref. ACS Applied Materials & Interfaces, 11(14): 13812-13821(2019)