python-ase

Files (6)

• QUICK_REFERENCE.md
• README.md
• SKILL.md
• adsorption_energy.py · examples/
• lattice_optimization.py · examples/
• neb_barrier.py · examples/

SKILL.md (first lines)

# Python ASE (Atomic Simulation Environment) Skill

This skill provides expert guidance for working with the Atomic Simulation Environment (ASE), a Python library for setting up, manipulating, running, visualizing, and analyzing atomistic simulations.

## When to Use This Skill

Use this skill when:

- Building atomic structures (molecules, surfaces, bulk crystals, nanoparticles)
- Setting up and running DFT calculations (VASP, GPAW, Quantum ESPRESSO, etc.)
- Performing geometry optimizations or molecular dynamics
- Analyzing trajectories and simulation results
- Working with ASE Atoms objects and calculators
- Converting between different structure formats
- Calculating properties (energies, forces, stress, bands, DOS)
- Writing automation scripts for materials simulations

## Core ASE Concepts

### 1. Atoms Objects

The fundamental ASE object representing atomic structures:

- Contains atomic positions, cell parameters, periodic boundary conditions
- Can be modified, visualized, and used with calculators
- Supports constraints for selective dynamics

### 2. Calculators

Interface to quantum chemistry and materials science codes:

- EMT (Effective Medium Theory) - fast, for testing
- VASP - plane-wave DFT
- GPAW - real-space/plane-wave DFT
- Quantum ESPRESSO, CASTEP, NWChem, etc.
- Must be attached to Atoms objects: `atoms.calc = calculator`

### 3. Optimizers

Geometry optimization algorithms:

- BFGS, LBFGS - quasi-Newton methods
- FIRE - fast inertial relaxation
- MDMin - molecular dynamics minimization

### 4. Dynamics

Molecular dynamics simulations:

- VelocityVerlet - NVE ensemble
- Langevin - NVT with friction
- NPT - constant pressure and temperature

## Common Patterns and Workflows

### Building Structures

```python
from ase import Atoms
from ase.build import bulk, molecule, surface, add_adsorbate

# Bulk crystals
atoms = bulk('Cu', 'fcc', a=3.6)
atoms = bulk('Si', 'diamond', a=5.43)

# Molecules
atoms = molecule('H2O')
atoms = molecule('CH4')

# Surfaces
slab = surface('Au', (1, 1, 1), size=(4, 4, 4), vacuum=10.0)

# Add adsorbate to surface
add_adsorbate(slab, 'O', height=1.5, position='fcc')
```

### Setting Up Calculators

```python
# EMT calculator (for testing)
from ase.calculators.emt import EMT
atoms.calc = EMT()

# VASP calculator
from ase.calculators.vasp import Vasp
calc = Vasp(
    xc='PBE',
    encut=400,
    kpts=(4, 4, 4),
    ismear=1,
    sigma=0.1,
    ibrion=2,
    nsw=100,
    ediff=1e-5,
)
atoms.calc = calc

# GPAW calculator
from gpaw import GPAW, PW
calc = GPAW(
    mode=PW(400),
    xc='PBE',
    kpts=(4, 4, 4),
    txt='output.txt'
)
atoms.calc = calc
```

### Geometry Optimization

```python
from ase.optimize import BFGS, LBFGS, FIRE

# Basic optimization
opt = BFGS(atoms, trajectory='opt.traj')
opt.run(fmax=0.05)  # converge until forces < 0.05 eV/Å

# With constraints (fix bottom layers)
from ase.constraints import FixAtoms

From woernfl/ai-lab, MIT.

More from woernfl/ai-lab