Official Resources
- Homepage: https://ape.gitlab.io/ape/
- Source Repository: https://gitlab.com/ape/ape
- Documentation: https://ape.gitlab.io/ape/
- License: GNU General Public License v2.0
Overview
APE (Atomic Pseudopotentials Engine) is a computer package designed to generate and test norm-conserving pseudopotentials within density functional theory. Developed by Micael Oliveira and Fernando Nogueira at the Centre for Computational Physics, University of Coimbra, APE produces pseudopotential files suitable for use with SIESTA, OCTOPUS, ABINIT, and PWscf. The generated pseudopotentials can be non-relativistic, scalar relativistic, or fully relativistic, and can explicitly include semi-core states.
APE leverages the Libxc library for a wide range of exchange-correlation functionals, making it versatile for different chemical systems. The code has also been modified for orbital-free DFT (OF-DFT) calculations, enabling the study of kinetic energy functionals through decomposition of the Kohn-Sham equation. APE is written in Fortran and distributed under GPLv2.
Scientific domain: Electronic structure theory, pseudopotential generation
Target user community: Computational physicists and chemists requiring custom pseudopotentials
Theoretical Methods
- Norm-conserving pseudopotential generation
- Non-relativistic, scalar-relativistic, and fully relativistic treatments
- Semi-core state inclusion
- All-electron atomic solve using ODE integration (Runge-Kutta Prince-Dormand)
- DFT with wide range of XC functionals via Libxc
- Orbital-free DFT modifications (Pauli potential decomposition)
Capabilities (CRITICAL)
- Generate norm-conserving pseudopotentials for multiple DFT codes
- Non-relativistic, scalar relativistic, and fully relativistic pseudopotentials
- Explicit semi-core state support
- Output formats for SIESTA, OCTOPUS, ABINIT, and PWscf
- Pseudopotential testing and validation
- All-electron reference calculations
- Libxc integration for extensive XC functional support
- Modified version for orbital-free DFT calculations
Inputs & Outputs
Input formats:
- APE input files (inp.ape) specifying atom, theory level, and configuration
- XC functional selection
- Mesh and convergence parameters
Output data types:
- Pseudopotential files in multiple formats (SIESTA, OCTOPUS, ABINIT, PWscf)
- All-electron wavefunctions and eigenvalues
- Pseudopotential wavefunctions
- Convergence and testing reports
Interfaces & Ecosystem
- Programming language: Fortran
- Compatible codes: SIESTA, OCTOPUS, ABINIT, PWscf
- XC functionals: Via Libxc library
- Modified versions: OF-DFT version for orbital-free calculations
Limitations & Known Constraints
- Focused on norm-conserving pseudopotentials (not PAW or ultrasoft)
- Requires Fortran compiler for building
- Documentation primarily in code examples
- Limited to atomic calculations
Performance Characteristics
- Single-atom calculations, very fast (seconds)
- No parallelization required
- Minimal memory requirements
- Testing multiple configurations increases runtime modestly
Comparison with Other Codes
- vs OPIUM: APE supports fewer construction methods but has cleaner Libxc integration; OPIUM offers RRKJ, Kerker, and TM methods with ghost state checking
- vs ATOMPAW: ATOMPAW generates PAW datasets; APE focuses on norm-conserving pseudopotentials
- vs PSEUDODOJO: PSEUDODOJO provides pre-generated validated potentials; APE is a generation tool
Best Practices
- Test pseudopotential transferability across multiple configurations
- Use scalar-relativistic or fully relativistic modes for heavy elements
- Include semi-core states when core-valence overlap is significant
- Verify all-electron convergence before pseudopotential generation
- Cross-check generated potentials with the target DFT code
Verification & Sources
Primary sources:
- Official website: https://ape.gitlab.io/ape/
- GitLab repository: https://gitlab.com/ape/ape
- M. Oliveira and F. Nogueira, Comput. Phys. Commun. 178, 524 (2008)
Confidence: VERIFIED - Official website and GitLab repository confirmed accessible