That is,
- How many intermediate compounds would be possible, if any?
- would the phase diagram be Eutectic, Peritectic, Monotectic, Eutectoid, Peritectoid, Syntectic, Catatectic.... (of course, isomorphous phase diagrams are quite well predictable by Hume-rothery rules
- What would be the critical points and compositions?
Of course, one may calculate gibbs free energy of individual phases if entropy and specific heat data of any phase is known, but specific heat calculation from Density Functional Theory or Electronic Band Structure is not easy. If even possible for elements, what about the crystal structure of intermetallics which might be completely unknown? Then you cannot even predict the vibrational modes as it is required for statistical mechanics calculation for specific heat.
Dear Sumit Bhowmick,
Sometimes, you can get a good idea of what a binary phase diagram is like by referencing known phase diagrams of similar systems, that is, binaries between elements from the same periodic table groups.
You can find phase diagrams compilations in books (such as Binary Alloy Phase Diagrams by Massalski et al., ISBN-13: 978-0-87170-403-0) and online (such as https://cpddb.nims.go.jp/en/).
Our research group is working on developing AI tools to predict phase diagrams based on this strategy. We are currently considering different approaches to predict binary phase diagrams. If you are interested, you can follow our updates here :
https://www.researchgate.net/project/Phase-diagram-prediction-using-machine-learning
Best regards,
Guillaume
- Badges
- Science topic