Applications of Python in Computational Chemistry and Material Design
Owain T. Beynon, Alun Owens, Andrew J. Logsdail
Cardiff Catalysis Institute, School of Chemistry, Cardiff University, Cardiff, Wales.

Computational methods afford an insight into the behaviours and properties of materials and recently Python packages have increasingly become a powerful tool for the study of chemical systems. Computational chemistry is the branch of chemistry that studies chemical systems through simulation and involves HPC architecture and software packages. In general, there are two categories of simulation, dynamic and static and of these types, atoms in the system may be described by varying levels of theory: molecular (classical) mechanics (MM), quantum mechanics (QM) or a combination of the two (QM/MM). Static calculations obtain the property of the system at a fixed geometry, whereas dynamic calculations study the evolution of a system over a given timeframe.

Python has become an integral part of computational modelling of materials in recent years, with development of packages such as the Atomic Simulation Environment (ASE) [1], which is a set of modules for manipulating, running and visualising atomic simulation. Furthermore, ASE integrates seamlessly with many electronic structure software packages, used for calculating the energy and properties of systems based on some level of theory, e.g Density Functional Theory (DFT), such as FHI-aims and VASP. [2,3] Moreover, the combination with other Python packages that integrate with ASE provide an ecosystem for atomic simulations (figure 1). Packages such as CatLearn, [4] a machine-learning approach used for determining transition states and energies needed for reactions, along with Phonopy and FHI-vibes, [5,6] both are for studying lattice dynamics of materials, and Py-ChemShell,[7] used for QM/MM calculations, to name a few, provide a comprehensive toolkit for the computational study of materials and chemical systems.

In our research, such approaches as outlined above are essential to further our understanding of materials and chemical processes, and of particular interest are materials for green and sustainable processes such as catalysts used for biofuel production, and battery materials, namely, zeolite Tin-BETA and Prussian Blue, [8-10] respectively. In collaboration with experimentalists, we seek to understand the synthetic methods and properties of these materials, with the aim to further establish them as viable alternatives to fossil fuels. In this regard, as Python software becomes increasingly popular for the simulation and study of materials, it also provides the tools and methods needed for tackling some of the grand challenges of today.

References
[1] A. Hjorth Larsen, et al. J. Phys. Condens. Matter 29 273002 (2017) [2] V. Blum, et al. Comput. Phys. Commun., 180, 2175–2196 (2009) [3] G. Kresse and J. Hafner, Phys. Rev. B 47 , 558 (1993) [4] A. Garrido Torres, et al. Phys. Rev. Lett., 122, (2019)[5]A. Togo, I. Tanaka, Scr. Mater, 108 (2015). [6] F. Knoop et al. J. Open Source Softw. (2020) [7] Y. Lu, et al. J. Chem. Theory Comput., 15, 1317–1328 (2019) [8] A. Corma et al. Nature, 412, 423–426 (2001) [9] C. Hammond, S. Conrad, I. Hermans, Angew. Chem. Int., 51, 11736–11739 (2012)[10] C. Ling. et al. J. Phys. Chem C 117 (2013)