Plotting the Density of States (DOS) from Vienna Ab initio Simulation Package (VASP) calculations using the Atomic Simulation Environment (ASE) is a crucial task in computational materials science. This process involves extracting electronic structure data from VASP outputs and visualizing it to gain insights into material properties. The workflow typically includes running VASP calculations, processing the output files, and using ASE’s tools to plot the DOS. Key considerations include proper structure relaxation, performing static self-consistent field (SCF) calculations, and conducting non-SCF calculations with denser k-point meshes for accurate DOS representation. Best practices involve using the IBZKPT file for consistency, setting appropriate VASP parameters like ISTART and ICHARG, and leveraging ASE’s capabilities for efficient data extraction and visualization. This approach enables researchers to effectively analyze and present electronic structure information from their VASP simulations.
Workflow for Plotting DOS
The workflow for plotting DOS from VASP outputs involves three main steps: structure relaxation, static SCF calculation, and non-SCF DOS calculation. Initially, a geometry optimization is performed to obtain the ground state structure. This is followed by a static SCF calculation on the relaxed structure to generate converged charge density and wavefunctions. Finally, a non-SCF calculation with a denser k-point mesh is conducted for accurate DOS representation. This approach ensures consistency between structural, electronic, and DOS calculations while maintaining computational efficiency. The IBZKPT file from the static SCF calculation can be used as the KPOINTS file for the non-SCF DOS calculation, promoting consistency and potentially reducing computational cost
Non-SCF Calculation Settings
For non-SCF DOS calculations, key VASP settings include ISTART = 1 to read wavefunctions from the WAVECAR file, and ICHARG = 11 or 12 to fix the charge density. These settings ensure consistency with previous SCF calculations while allowing for changes in calculation parameters. The IBZKPT file from the static SCF run can be used as the KPOINTS file for the non-SCF calculation, promoting efficiency and compatibility. However, for highly accurate DOS representations, a denser k-point mesh may be necessary. It’s crucial to include enough bands (NBANDS) to cover the desired energy range for DOS analysis.
Using ASE for DOS
The Atomic Simulation Environment (ASE) provides powerful tools for plotting Density of States (DOS) from VASP calculations. Key steps include:
• Reading the vasprun.xml file using ASE’s read function
• Creating a DOS object with the VASP calculator
• Extracting DOS values and energies
• Plotting the data using matplotlib
ASE offers flexibility in customizing DOS plots, allowing for energy window adjustments, spin-polarized representations, and efficient handling of large datasets. The DOS object’s get_site_dos() method enables plotting projected DOS for specific atoms or orbitals, enhancing the analysis capabilities. For more advanced analysis, ASE’s get_distribution_moment method can be used to calculate moments of the DOS distribution.
Best Practices for VASP
Best practices for VASP DOS calculations include using the tetrahedron method with Blöchl corrections (ISMEAR = -5) for accurate results. Setting ISTART = 1 allows reading wavefunctions from a previous SCF calculation, maintaining consistency while allowing for changes in calculation type. ICHARG = 11 or 12 should be used to fix the charge density in non-SCF runs. For hybrid functional calculations like HSE06, starting from a converged wavefunction is particularly important. Ensuring sufficient bands (NBANDS) to cover the desired energy range for DOS is crucial. These settings optimize accuracy and efficiency in VASP-based electronic structure analysis.
References:
- Tutorials – VASP
- mukkelian/VASP-DOS-plots – GitHub
- calculation – VASP
- File input and output — ASE documentation – CAMD Wiki pages
- DOS and Band Structure Calculation using VASP – YouTube
- Density of states – JDFTx
- Part 1: Introduction to VASP
- DeepMind’s AlphaFold: Revolutionizing Protein Folding Solutions
- AI Hardware: GPUs, TPUs, and NPUs Explained
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Theoretical & Computational Materials Science 