Integração de cálculo CALPHAD com Python para seleção de ligas multielemento para armazenagem de hidrogênio

Abstract

The development of multielement alloys for solid-state hydrogen storage in the form of metal hydrides is essential for utilizing hydrogen as an energy vector. Alloys used for this application must have a good hydrogen-to-metal ratio, high volumetric and gravimetric capacities, acceptable thermodynamic and kinetic hydrogen absorption/desorption properties, and long-term cyclic stability. Within this category, monophase CCC and Laves_C14 alloys stand out. In this context, the objective of this work was to select compositions that meet the mentioned requirements by integrating the PanPython tool, present in the PANDAT™ software, which is based on the CALPHAD method, with the Python programming language, aiming to obtain a large volume of data in a short period of time. 3003 distinct compositions containing the elements Cr, Mn, Nb, Ti, V, and Zr were calculated using the PanHEAR 2022 database in 4102 seconds, of which 2970 yielded valid results, with 745 being monophase CCC alloys and 86 being monophase Laves_C14 alloys. The multiphase alloys composed of CCC and Laves_C14 phases were analyzed, and the composition of each phase allowed for the extraction of additional monophase alloys of interest. For the original monophase alloys, their compositions were analyzed using boxplots, thereby identifying the main elements composing them. Furthermore, the value of the φ parameter (an adimensional thermodynamic parameter used as an indicator of entropy) was calculated for the monophase CCC and Laves_C14 alloys, as well as the multiphase CCC + Laves_C14 alloys, and compared to the critical φ value (20). According to studies, alloys with a φ value higher than 20 tend to form monophase alloys through solid solution. However, when plotting configurational entropy against φ, it can be observed that this relationship is prone to failures. The results of this work open up the prospect of using integrated high-throughput methodologies that combine CALPHAD with Python programs, which can accelerate the development of new multicomponent alloys.