Mat Xrd Calculator

XRD Spectrum Calculation

This skill calculates the X-ray Diffraction (XRD) pattern of a crystal structure using pymatgen. It identifies diffraction peaks, their intensities, and associated (hkl) indices.

Requirements

• Conda environment: base-agent
• pymatgen
• matplotlib

Usage

The primary script for this skill is calculate_xrd.py. It takes a structure file as input and generates a JSON file with the diffraction data and a plot of the intensities versus $2\theta$.

Command Line Interface

python .agents/skills/mat-xrd-calculator/scripts/calculate_xrd.py <structure_file> --output_dir <output_dir> --wavelength <wavelength>

Arguments

• structure: Path to the input structure file (e.g., POSCAR, CIF).
• --output_dir: (Optional) Directory to save the results. Defaults to the current directory.
• --wavelength: (Optional) Radiation wavelength or source name (e.g., CuKa, MoKa, CrKa). Defaults to CuKa ($1.54184$ Å).
• --symprec: (Optional) Symmetry precision for identifying equivalent peaks. Defaults to 0.1.

Output Files

1. <filename>_xrd.json: Contains $2\theta$ positions, intensities, d-spacings, and (hkl) indices.
2. <filename>_PV_xrd.png: A plot of the simulated XRD spectrum (Pseudo-Voigt model).

Example

To calculate the XRD pattern for LiFePO4:

```bash
conda activate base-agent
python .agents/skills/mat-xrd-calculator/scripts/calculate_xrd.py .agents/skills/mat-xrd-calculator/examples/LiFePO4/LiFePO4.cif --output_dir .agents/skills/mat-xrd-calculator/examples/LiFePO4

## Foundation Potential Recommendations

Since XRD is a purely geometric property of the crystal structure, it does not require a machine learning interatomic potential (MLIP) for the calculation itself. However, it is **highly recommended** to perform a structure relaxation using a high-quality MLIP (e.g., MACE, CHGNet) before calculating the XRD pattern to ensure the structure is at its energy minimum.

For recommendations on relaxation models, see the [ml-foundation-potentials](file:///home/bdeng/projects/AtomisticSkills/.agents/skills/ml-foundation-potentials/SKILL.md) skill.
---

**Author:** Bowen Deng
**Contact:** [GitHub @learningmatter-mit](https://github.com/learningmatter-mit)