pyCOFBuilder is a simple and powerful python package to automatically assembly COF structures with specifics building blocks, topologies, and functionalizations following the reticular approach to build and represent COF structures. The project was developed to address the need for the generation of COFs structures in a high-throughput style, based on a nomenclature tha allows direct sctructural feature interpretation from a simple name. The package uses pymatgen to create the structures.

This package is still under development, but it is already possible to create a large number of COF structures.

Learn more at the Documentation

0. Python >= 3.10
1. pymatgen >= 2022.0.0
2. numpy >= 1.2
3. scipy >= 1.6.3
4. simplejson
5. ase
6. gemmi

The Python dependencies are most easily satisfied using a conda (anaconda/miniconda) installation by running

conda env create --file environment.yml

You can install pyCOFBuilder using pip:

pip install pycofbuilder

Alternatively, you can use pyCOFBuilder by manually importing it using the sys module, as exemplified below:

# importing module
import sys
 
# appending a path
sys.path.append('{PATH_TO_PYCOFBUILDER}/pyCOFBuilder/src')

import pycofbuilder as pcb

Just remember to change the {PATH_TO_PYCOFBUILDER} to the directory where you downloaded the pyCOFBuilder package.

To create a specific COF, such as T3_BENZ_NH2_OH-L2_BENZ_CHO_H-HCB_A-AA:

# importing module
import sys
 
# appending a path
sys.path.append('{PATH_TO_PYCOFBUILDER}/pyCOFBuilder/src')

import pycofbuilder as pcb

cof = pcb.Framework('T3_BENZ_CHO_OH-L2_BENZ_NH2_H-HCB_A-AA')
cof.save(fmt='cif', supercell = [1, 1, 2], save_dir = '.')

You should see an output such as:

T3_BENZ_NH2_OH-L2_BENZ_CHO_H_H-HCB_A-AA                       hexagonal   P    P6/m # 175    12 sym. op.

A .cif file (the default save format is CIF, but it can be easily changed by setting another value on the fmt option) will be created in the out folder. The code will print out some information about the structure created.

Currently, it is possible to select the following formats:

• cif
• xsf
• pdb
• cjson
• vasp
• turbomole
• pqr
• qe
• gjf
• xyz

Besides, the variable structure is now a Framework object. This object has some attributes that can be accessed:

>>> cof.name
'T3_BENZ_NH2_OH-L2_BENZ_CHO_H-HCB_A-AA'
>>> cof.smiles
'(N)C1=C(O)C((N))=C(O)C((N))=C1O.(C([H])=O)C1=C([H])C([H])=C((C([H])=O))C([H])=C1[H]'
>>> cof.lattice
array([[ 22.49540055,   0.        ,   0.        ],
       [-11.24770028,  19.48158835,   0.        ],
       [  0.        ,   0.        ,   3.6       ]])
>>> cof.n_atoms
72
>>> cof.space_group
'P6/m'

To ensure greater reproducibility as well as quick and easy access to relevant information from the COFs, I've developed a simple nomenclature to name the structure. Generally speaking, a COF can be described as

where:

• Building_Block_1: The building block with the greater connectivity.
• Building_Block_2: The building block with the smaller connectivity.
• Net: The net describing the reticular structure.
• Stacking: The stacking (for 2D structures) or interpenetrating degree (for 3D structures)

To name the building blocks, I also developed a set of rules. The building block can be described as

where:

• Symmetry: The general symmetry of the building block. Also represents the connectivity of the building block. For 2D building blocks can be L2, T3, or S4, and for 3D building blocks can be D4.
• Core: The 4-letter code referring to the building block core.
• Connector: The type of functional group that will be used to assembly the COF structure. Ex.: NH2, CHO, CONHNH2, etc.
• RadicalGroupRN: The Nth radical group in the structure. The number of Radical groups will change according to the availability of the core.

Note that every "card" for the building block name is separated by an underline (_) and every "card" for the COF name is separated by a dash (-). This makes it easy to split the COF name into useful information.

If you find pyCOFBuilder useful in your research, please consider citing the following paper:

F. L. Oliveira and P. M. Esteves, pyCOFBuilder: A Python package for automated creation of Covalent Organic Framework models based on the reticular approach J. Chem. Inf. Model. 2024, 64, 8, 3278–3289 10.1021/acs.jcim.3c01918 DOI