Machine learning of phase diagrams

Lund, John E.; Wang, Haoyue; Garcı́a, R. Edwin; Braatz, Richard D.

doi:10.1039/d2ma00524g

Cited by 4 publications

(4 citation statements)

References 121 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…used the modified quasichemical model for the liquid solution for the Nd 2 O 3 −B 2 O 3 system, however, they noted that there were no studies on thermal properties for Nd 4 B 2 O 9 causing discrepancies between the optimized phase diagram and the phase diagram produced by Ji et al [32] . Hence, validation of the phase diagram and thermal properties for Nd 4 B 2 O 9 is required before Lund et al's [40] . model could be applied for this system.…”

Section: Resultsmentioning

confidence: 99%

“…Despite thermodynamic predictive software not containing equilibrium data on the above systems, there is a method in literature where real solution reduction predictions (based on the activity of the simple oxides present) could be made. Lund et al [40] devised a machine learning model using thermodynamic solution models of phase diagrams and line-compound thermal properties to determine properties of phases. Properties could include Gibbs free energy and activity information of Ta 2 O 5 À K 2 O, Nd 2 O 3 À B 2 O 3 and Nd 2 O 3 À CuO.…”

Section: Methodsmentioning

confidence: 99%

“…Jakobsson et al [42] used the modified quasichemical model for the liquid solution for the Nd 2 O 3 À B 2 O 3 system, however, they noted that there were no studies on thermal properties for Nd 4 B 2 O 9 causing discrepancies between the optimized phase diagram and the phase diagram produced by Ji et al [32] Hence, validation of the phase diagram and thermal properties for Nd 4 B 2 O 9 is required before Lund et al's. [40] model could be applied for this system. Experimentally validating and obtaining thermodynamic solution models for the molten phases of the identified phase diagrams, in addition to gathering thermal properties of line compounds (e. g. melting temperature, enthalpy, entropy, heat capacity etc.…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Evaluating Phase Relations in Molten Oxides for Electrochemical Operating Windows for Selective Metal Reduction

Ford,

Newport,

Marshall

et al. 2024

ChemElectroChem

View full text Add to dashboard Cite

Molten oxide electrolysis is a promising pathway to decarbonize primary metal production. The oxide electrolyte is less hazardous and eco‐toxic than halides (molten salt electrolysis) and, with the use of renewably generated electricity and an inert anode, oxygen is produced as a by‐product instead of CO2. Building fundamental understanding of electrolytic operating windows requires starting with simple chemistries. Here we investigate a simplified binary oxide system as an electrolyte for titanium oxide reduction to metal. The TiO2−Na2O binary system forms a eutectic, reducing the liquidus temperature over a wide composition range. FactSage 8.1 predictions suggested Ti reduction would become favorable over Na reduction for compositions greater than 0.49 mole fraction TiO2. The prediction was validated by the detection of metallic Ti after electrolysis experiments. However, the reduction efficiency was too low (0.24±0.08 % at −0.1 V vs Ti reference electrode) for the TiO2−Na2O system to be a viable industrial electrolyte for Ti production. Scoping for other binary oxide systems was performed for electrolytic production of two critical metals, tantalum and neodymium. Based on TiO2−Na2O system's predicted behavior, candidate binary oxide systems were identified that contained congruently melting line compounds flanked by eutectic reactions from the ACerS‐NIST Phase Equilibria diagrams database.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Evaluating Phase Relations in Molten Oxides for Electrochemical Operating Windows for Selective Metal Reduction

Ford,

Newport,

Marshall

et al. 2024

ChemElectroChem

View full text Add to dashboard Cite

show abstract

“…Another option is to incorporate information related to changes and instability in the system phases into the models. There has been work on the prediction of phase diagrams and phase stability using ML methods, which could pave the way for the creation of phase change-informed models for heat capacity prediction, , which will be explored in the future. Nonetheless, the current ML approach already predicts C p ( T , x ) with an error that is less than the experimental uncertainty and state-of-the-art models (ideal, semiempirical), presenting a significant advance and useful tool for the screening and C p prediction for mixed oxides, which can be generalized to higher-order systems.…”

mentioning

confidence: 99%

Chemistry Informed Machine Learning-Based Heat Capacity Prediction of Solid Mixed Oxides

Barra,

Chahal,

Audesse

et al. 2024

J. Phys. Chem. Lett.

View full text Add to dashboard Cite

Knowing heat capacity is crucial for modeling temperature changes with the absorption and release of heat and for calculating the thermal energy storage capacity of oxide mixtures with energy applications. The current prediction methods (ab initio simulations, computational thermodynamics, and the Neumann−Kopp rule) are computationally expensive, not fully generalizable, or inaccurate. Machine learning has the potential of being fast, accurate, and generalizable, but it has been scarcely used to predict mixture properties, particularly for mixed oxides. Here, we demonstrate a method for the generalizable prediction of heat capacity of solid oxide pseudobinary mixtures using heat capacity data obtained from computational thermodynamics and descriptors from ab initio databases. Models trained through this workflow achieved an error (mean absolute error of 0.43 J mol −1 K −1 ) lower than the uncertainty in differential scanning calorimetry measurements, and the workflow can be extended to predict other properties derived from the Gibbs free energy and for higher-order oxide mixtures.

show abstract

Topological data analysis assisted machine learning for polar topological structures in oxide superlattices

Du,

Zhou,

Huang

et al. 2025

Acta Materialia

View full text Add to dashboard Cite

Machine learning of phase diagrams

Abstract: By starting from experimental- and ab initio-determined phase diagrams (PDs) of materials, a machine learning (ML) method is developed to infer the free energy function for each phase. The ML...

Cited by 4 publications

References 121 publications

Evaluating Phase Relations in Molten Oxides for Electrochemical Operating Windows for Selective Metal Reduction

Evaluating Phase Relations in Molten Oxides for Electrochemical Operating Windows for Selective Metal Reduction

Chemistry Informed Machine Learning-Based Heat Capacity Prediction of Solid Mixed Oxides

Topological data analysis assisted machine learning for polar topological structures in oxide superlattices

Contact Info

Product

Resources

About