Siesta: desenvolvimento de um utilitário para automação e análise de cálculos DFT

Abstract

This work presents the development of a Python-based utility designed to automate the processing of input and output files from the SIESTA software, widely used for simulations based on Density Functional Theory (DFT) (HOHENBERG; KOHN, 1964; KOHN; SHAM, 1965; SOLER et al., 2002). The main objective was to create a tool capable of automatically generating input files (.fdf) from user-defined parameters and converting SIESTA structural output files into .CIF format, compatible with crystallographic databases and visualization programs (MOMMA; IZUMI, 2011; KOKALJ, 1999). The implementation was carried out using the ASE (Atomic Simulation Environment), NumPy, Pandas, and Matplotlib libraries, enabling efficient data handling, atomic structure manipulation, and graphical representation (LARSEN et al., 2017; HARRIS et al., 2020; MCKINNEY, 2010; HUNTER, 2007). The chosen material for validation was CuWO4 , a semiconductor oxide of scientific interest due to its relevant structural and electronic properties. The utility automates the extraction of relaxation, optimization, and final geometry data, converting them into standardized formats for subsequent analysis. The results demonstrated that the tool simplifies the SIESTA workflow, reducing manual steps and conversion errors while ensuring greater reproducibility of DFT simulations. Therefore, the developed application represents a significant contribution to data processing in computational materials science, combining computational efficiency and scientific accessibility.