easyxtb 0.5.0

A Python API for xtb (and CREST).

easyxtb

easyxtb is an unofficial API for the xtb and CREST semi-empirical quantum chemistry programs with an emphasis on intuitive and straightforward usage.

It forms the basis for avo_xtb, a plugin for the 3D chemical visualization software Avogadro 2 that provides an in-app interface to the xtb program for quick and accurate calculations, as well as the CREST program for extended functionality.

The Python package easyxtb can be used independently of avo_xtb as an interface to launch calculations and process their results from Python.

xtb is developed by the Grimme group in Bonn and carries out semi-empirical quantum mechanical calculations using the group's extended Tight-Binding methods, referred to as "GFNn-xTB".

These methods provide fast and reasonably accurate calculation of Geometries, Frequencies, and Non-covalent interactions for molecular systems with up to roughly 1000 atoms, with broad coverage of the periodic table up to Z = 86 (radon).

crest (Conformer–Rotamer Ensemble Sampling Tool) adds a variety of sampling procedures for several interesting applications including conformer searches, thermochemistry, and solvation.

Usage

Full documentation is a work in progress, but to demonstrate basic usage:

To load a geometry (XYZ and CJSON files are supported currently):

>>> from pathlib import Path
>>> import py_xtb
>>> py_xtb.XTB_BIN = Path.home() / ".local/bin/xtb"
>>> input_geom = py_xtb.Geometry.from_file(Path.home() / "calcs/benzoic_acid.xyz")
>>> input_geom
<py_xtb.geometry.Geometry object at 0x7ff91fdb15d0>
>>> input_geom.atoms
[Atom(element='C', x=-3.61652, y=0.64945, z=-0.0), Atom(element='C', x=-3.59105, y=-1.15881, z=-0.0), Atom(element='C', x=-0.43296, y=-1.32436, z=-1e-05), ..., Atom(element='O', x=1.54084, y=3.42551, z=-1e-05), Atom(element='O', x=2.9034, y=0.39476, z=1e-05), Atom(element='H', x=2.60455, y=-0.50701, z=2e-05)]
>>> for atom in input_geom:
...     print(atom)
... 
Atom(element='C', x=-3.61652, y=0.64945, z=-0.0)
Atom(element='C', x=-3.59105, y=-1.15881, z=-0.0)
... # Truncated for brevity
Atom(element='O', x=2.9034, y=0.39476, z=1e-05)
Atom(element='H', x=2.60455, y=-0.50701, z=2e-05)

The package provides a function API for basic xtb calculation types (energy, optimize, frequencies, opt_freq, orbitals):

>>> optimized = py_xtb.calc.optimize(input_geom, charge=0, multiplicity=1, solvation="water", method=2, level="normal")
>>> for atom in optimized:
...     print(atom)
... 
Atom(element='C', x=-2.94841377173243, y=0.53595421697827, z=-0.00205114575446)
Atom(element='C', x=-2.96353116164446, y=-0.85021326429608, z=-0.00028605784754)
... # Truncated for brevity
Atom(element='O', x=1.83119245847571, y=0.4663499792285, z=-0.00384872174663)
Atom(element='H', x=1.55896002888703, y=-0.46876579604809, z=-0.00948378184114)

For greater control and for runtypes or command line options that don't yet have support in the API the Calculation object can be used:

>>> freq_calc = py_xtb.Calculation(program="xtb", runtype="hess", options={"charge": 0, "multiplicity": 1, "solvation": "water"})
>>> freq_calc.input_geometry = input_geom
>>> freq_calc.run()
>>> freq_calc.energy
-24.418716794336
>>> freq_calc.output_geometry
<py_xtb.geometry.Geometry object at 0x7ff91fdb15d0>
>>> freq_calc.frequencies
[{'mode': 1, 'symmetry': 'a', 'frequency': -816.7902, 'reduced_mass': 12.1428, 'ir_intensity': 7.2604, 'raman_scattering_activity': 0.0, 'eigenvectors': [[-0.26, -0.46, 0.0], [-0.24, 0.43, -0.0], [-0.05, -0.05, 0.0], [0.08, 0.3, -0.0], [0.42, 0.02, -0.0], [-0.02, 0.03, -0.0], [0.24, -0.09, -0.0], [0.0, -0.02, 0.0], [-0.26, 0.1, 0.0], [0.05, 0.01, -0.0], [-0.02, -0.03, 0.0], [0.0, -0.21, 0.0], [0.05, 0.0, -0.0], [-0.02, 0.01, -0.0]]}, {'mode': 2, 'symmetry': 'a', 'frequency': -759.3794, 'reduced_mass': 12.9124, 'ir_intensity': 17.3638, 'raman_scattering_activity': 0.0, 'eigenvectors': [[0.12, 0.19, -0.0], [0.15, -0.32, 0.0], [0.07, 0.18, -0.0], [0.12, 0.58, -0.0], [0.01, -0.11, 0.0], [0.01, -0.03, 0.0], [-0.29, 0.05, 0.0], [-0.02, 0.01, 0.0], [-0.32, -0.02, -0.0], [0.02, -0.03, -0.0], [0.01, 0.01, -0.0], [-0.01, -0.47, 0.0], [0.13, -0.0, -0.0], [-0.03, 0.02, 0.0]]}, ..., {'mode': 36, 'symmetry': 'a', 'frequency': 3752.5636, 'reduced_mass': 1.893, 'ir_intensity': 79.5584, 'raman_scattering_activity': 0.0, 'eigenvectors': [[-0.0, -0.0, 0.0], [0.0, -0.0, 0.0], [0.0, 0.0, -0.0], [-0.0, 0.0, 0.0], [-0.0, 0.0, -0.0], [-0.0, 0.0, -0.0], [0.0, 0.0, 0.0], [-0.0, 0.0, 0.0], [0.0, 0.0, -0.0], [0.0, 0.0, -0.0], [-0.0, 0.0, -0.0], [-0.0, -0.0, 0.0], [0.08, 0.23, -0.0], [-0.31, -0.92, 0.0]]}]

We were returned some negative frequencies! The xtb team recommend generally using the ohess (optimization followed by a frequency calculation) as, in the case of negative frequencies, xtb will produce a distorted geometry along the imaginary mode. The opt_freq() calculation function takes advantage of this and will continue reoptimizing using these distorted geometries until a minimum is reached, so is more reliable in getting what we want:

>>> output_geom, output_freqs, calc = py_xtb.calc.opt_freq(input_geom, solvation="water", return_calc=True)
>>> output_freqs[0]
{'mode': 1, 'symmetry': 'a', 'frequency': 70.0622, 'reduced_mass': 13.4154, 'ir_intensity': 6.422, 'raman_scattering_activity': 0.0, 'eigenvectors': [[0.0, 0.0, -0.28], [-0.0, 0.0, -0.0], [-0.0, -0.0, 0.25], [0.0, -0.0, 0.04], [0.0, -0.0, -0.24], [-0.0, 0.0, -0.0], [-0.0, 0.0, 0.29], [-0.0, 0.0, 0.15], [0.0, -0.0, 0.02], [0.0, -0.0, -0.12], [0.0, 0.0, -0.15], [0.0, -0.0, 0.55], [0.0, 0.01, -0.56], [0.0, 0.0, -0.19]]}

The geometries can be converted back to XYZ or CJSON formats, or saved to file:

>>> output_geom.to_xyz()
['14', ' energy: -25.558538770724 gnorm: 0.000926226194 xtb: 6.7.1 (edcfbbe)', 'C        -2.94841     0.53597    -0.00204', 'C        -2.96354    -0.85020    -0.00027', 'C        -0.61077    -0.84138     0.00148', 'C         0.75088     1.26262     0.00405', 'C        -1.75577     1.23036    -0.00190', 'H        -3.89917    -1.39047     0.00050', 'C        -1.76452    -1.56433     0.00117', 'H        -1.76882    -2.64357     0.00283', 'C        -0.54492     0.52708     0.00053', 'H        -1.70901     2.30959    -0.00318', 'H        -3.88278     1.07775    -0.00326', 'O         0.84510     2.45868     0.01349', 'O         1.83119     0.46634    -0.00386', 'H         1.55896    -0.46877    -0.00949']
>>> output_geom.to_cjson()
{'atoms': {'coords': {'3d': [-2.94841097258746, 0.53596965124863, -0.00204178926542, -2.96353561428538, -0.85019774776951, -0.0002742292702, -0.61076575425764, -0.84137850926872, 0.00147639902781, 0.75088171422694, 1.2626220330939, 0.00405051293617, -1.75577241044775, 1.23035899741552, -0.00189846159875, -3.89916811270068, -1.39047034080804, 0.00049537269594, -1.76451828574256, -1.56433051119015, 0.00116984833688, -1.76882023929805, -2.64357304665774, 0.00283010066276, -0.54491722564374, 0.52708246765258, 0.00052577760818, -1.70901066150243, 2.30958748867245, -0.00317692528336, -3.88277644809706, 1.07774511697242, -0.00326302099431, 0.84510262816245, 2.45867627843098, 0.01348592446867, 1.83119488680479, 0.46634038440289, -0.00385592774839, 1.55895754130232, -0.46877397528872, -0.00948840548749]}, 'elements': {'number': [6, 6, 6, 6, 6, 1, 6, 1, 6, 1, 1, 8, 8, 1]}}}
>>> output_geom.to_file(Path.home() / "calcs/optimized_benzoic_acid.xyz")

Requirements

xtb

Only tested for xtb >= 6.7.

The xtb binary is not bundled with the package. Instead, it must be obtained separately.

The location of xtb can be set from Python code simply by setting easyxtb.XTB_BIN to an appropriate pathlib.Path object.

An xtb binary will also be picked up automatically by easyxtb if located in one of the following locations:

1. The system or user PATH
2. Within the easyxtb binary directory at <user data>/easyxtb/bin/xtb (see below for more information on where this is on your system)
3. Within the folder it is distributed in under the easyxtb binary directory, which would thus currently be at <user data>/easyxtb/bin/xtb-dist/bin/xtb
4. Any other location but with a link to it from <user data>/easyxtb/bin/xtb

CREST

Only tested for crest >= 3.0.

While xtb is cross-platform, crest is currently distributed only for Linux/UNIX systems.

crest can be made visible to the plugin in the same ways as for xtb listed above. If it is not in $PATH, the crest binary, or link to it, should be located at <user data>/easyxtb/bin/crest.

Data location

easyxtb uses a central location to run its calculations, store its configuration, and save its log file. This location is <user data>/easyxtb, where <user data> is OS-dependent:

• Windows: $USER_HOME\AppData\Local\easyxtb
• macOS: ~/Library/Application Support/easyxtb
• Linux: ~/.local/share/easyxtb

Additionally, if the environment variable XDG_DATA_HOME is set its value will be respected and takes precedence over the above paths (on all OSes).

Disclaimer

xtb and crest are distributed by the Grimme group under the LGPL license v3. The authors of easyxtb, avo_xtb, and Avogadro bear no responsibility for xtb or CREST or the contents of the respective repositories. Source code for the programs is available at the repositories linked above.

Cite

General reference to xtb and the implemented GFN methods:

• C. Bannwarth, E. Caldeweyher, S. Ehlert, A. Hansen, P. Pracht, J. Seibert, S. Spicher, S. Grimme WIREs Comput. Mol. Sci., 2020, 11, e01493. DOI: 10.1002/wcms.1493

For GFN2-xTB (default method):

• C. Bannwarth, S. Ehlert and S. Grimme., J. Chem. Theory Comput., 2019, 15, 1652-1671. DOI: 10.1021/acs.jctc.8b01176

For CREST:

• P. Pracht, S. Grimme, C. Bannwarth, F. Bohle, S. Ehlert, G. Feldmann, J. Gorges, M. Müller, T. Neudecker, C. Plett, S. Spicher, P. Steinbach, P. Wesołowski, F. Zeller, J. Chem. Phys., 2024, 160, 114110. DOI: 10.1063/5.0197592
• P. Pracht, F. Bohle, S. Grimme, Phys. Chem. Chem. Phys., 2020, 22, 7169-7192. DOI: 10.1039/C9CP06869D

See the xtb and CREST GitHub repositories for other citations.