Computer prediction technology can be used with new magnetic materials

Although magnets are everywhere in daily life, magnetism is still very rare in metal. At present, humans have discovered that 5% of the known inorganic compounds show some magnetic force. Since some of the magnetic force is generated at a specific temperature, there are actually relatively few magnetic materials that can be actually used in reality. Materials scientists at Duke University have a new study. They use computer models to quickly predict the magnetic properties of new materials.

This research is documented in the recent "Science Advances". Scientists selected 55 elements and randomly selected 3 of them to combine them. According to the computer model, the results showed that among the 236,115 possible combinations, 14 have the potential to be magnetic substances. In order to confirm the accuracy of this screening, the scientists formulated two of the substances according to the computer model-one is a combination of cobalt, magnesium and titanium, and the other is magnesium, palladium and platinum. According to the test, these two substances The substance does have the magnetic properties and other properties simulated by the computer. In particular, the cobalt-magnesium-titanium alloy can still maintain magnetism at extremely high temperatures. It is a magnetic substance that does not contain rare earth metals but possesses this property. Perhaps this substance will be used in the future. It may be used in practice.

The research team stated, “Any of these new magnets may prove useful in the future, but this is not really important. The important thing is that rapid prediction technology is a major change for materials science. Promoting the advancement of materials science will be invaluable".

This prediction project was supported by the Irish Science Foundation, the European Commission and the National Science Foundation (DGF1106401). But what kind of surprises this technology can bring to us in the future, we can only wait and see.