Pyrovskite

1. Extract structural features of 2D, 3D, and 2D/3D double perovskite systems.
2. Build and analyze bulk perovskites, double perovskites, 2D perovskites, and 2D double perovskites.
3. Generate elcetronic structure input files for xTB/GPAW.

🟩 If you use this code for your research, please cite: https://doi.org/10.1063/5.0159407 🟩

---

Goldschmidt and octahedral factors have been added for most systems, please see pyrovskite/data.py and pyrovskite/perovskite.py:337-363 if you need these experimental features before docs + examples are uploaded.

Requirements

• ASE https://wiki.fysik.dtu.dk/ase/
• pymatgen https://pymatgen.org/
• pytest to run pytest . in the tests directory.

Installation

The package can be installed via

pip install pyrovksite

or you can clone the repository locally and obtain an editable install with:

git clone https://github.com/r2stanton/pyrovskite.git .
pip install -e .

With this method you'll download the tests and example jupyter notebooks. Tests can be run with pytest by navigating to the tests/ folder and running pytest .

A small collection of organic A'-site and A-site are available in the res/ folder, see spacers.csv for charge info.

Usage

A brief overview of code usage is discussed here, however a much more detailed explanation is contained in the Jupyter notebooks found in /examples.

Example usage, from structure creation to analysis.

Example of building a bulk MAPbI3 perovskite:

from pyrovskite.builder import make_bulk
import ase.io
A = ase.io.read("path/to/methylammonium.xyz")
MAPbI3 = make_bulk(A, "Pb", "I", 3.1)

Similar such methods exist for 2D phases, and double perovskites. These are returned as an Atoms object, allowing for all the I/O methods supported by ASE.

One can write an input file, for example to do a variable cell xTB optimization with:

MAPbI3.write_xTB()

The optimized geometry could then be obtained and loaded back into a Perovksite object for further analysis:

from pyrovskite.perovskite import Perovskite
MAPbI3 = Perovskite("path/to/optimized_MAPbI3.cif")

Structural properties such as octahedral distortions can be computed with MAPbI3.compute_delta() MAPbI3.compute_sigma() MAPbI3.compute_lambda() and plots of relevant distributional quantities such as bond angle, bond length, and pRDFs can be computed with MAPbI3.plot_angles() MAPbI3.plot_distances() MAPbI3.plot_rdf()

Sensible defaults are selected for these functions, but please refer to the in-built documentation with help(function) or to the example notebooks to ensure the default behavior is what you want.

Note on xTB input file generation

xTB input files are for use with CP2K. Additionally, openbabel is required to convert files to a CP2K readable format. You can install both of these things with the command.

conda install -c conda-forge cp2k openbabel

• cp2k https://www.cp2k.org
• openbabel http://openbabel.org/wiki/Main_Page

Additionally, if you plan on using the xTB optimizations frequently, it may be worth setting

export XTB_D3_PATH=/path/to/dftd3.dat
export XTB_PARAMS_PATH=/path/to/xTB_parameters

to your shell profile, as the xTB calculation requires these parameters. If they are declared as environment variables, then they will be automatically parsed for you, otherwise you need to make a custom input file template, or pass the paths to these parameter files as keyword arguments to the relevant functions.