"2 1/2th" generation Calphad databases: Extrapolating heat capacities of elements and compounds to 0K. (December 2021)

• Record Type: Journal Article
• Authors: Obaied, A.; Tang, F.; Roslyakova, I.; to Baben, M.
• Is Part Of: Calphad. Volume 75(2021)
• Publication Date: 2021-12
• Subjects: Third generation CALPHAD databases -- SGTE -- Extension to 0 Kelvin -- Statistical analysis -- Ridge regression; Phase diagrams -- Data processing -- Periodicals; Thermochemistry -- Data processing -- Periodicals; Diagrammes de phases -- Informatique -- Périodiques; Thermochimie -- Informatique -- Périodiques; Thermodynamica; Electronic journals; 541.363
• DOI: 10.1016/j.calphad.2021.102352
• Languages: English
• ISSNs: 0364-5916

Abstract

Abstract: The extrapolation of thermodynamic properties below the room temperature is still, in practical terms, an unsolved issue. The majority of the already established Calphad databases are based on the 2nd generation Calphad models, which are available only at ( T > 298 . 15 K), there exist a few assessments relying on 3rd generation Calphad models that include more physical descriptions of heat capacity and extend the range of applicability down to 0K. To combine these slowly developing assessments with huge number of existing assessments, a compatible solution that can bridge the gap between 0K and 298.15K in an efficient and fully automated process is needed. Here, an automated modelling approach is proposed, that can be used to produce a reliable thermodynamic descriptions for a wide range of pure elements and compounds. The heat capacity and entropy values at room temperature are the only inputs required. Two fitting parameters are derived, which are, depending on the values of C P and S at 298.15K, either the Debye temperature and its low-temperature dependence or the Debye temperature and a C P contribution changing linearly with temperature. The results are produced as a simple polynomial expression that can be universally utilized in different computational thermodynamic software. Moreover, they are more accurate when compared with experimental or ab-initio results than solutions that rely on pure Einstein or pure Debye models. The applicability is shown on a huge dataset of solid compounds from the material classes of elements, metallic compounds, oxides, semiconductors and salts, thus covering all material classes relevant in Calphad assessments.