Now a novel automated system will allow the discovery of rare earth compounds


By: Dr. Franyi Sarmiento, Ph.D., Inspenet, March 25, 2022

A group of researchers from scientific institutions in the United States has developed a computer system based on a machine learning model that will help search for new rare earth chemical compounds, the Ames National Laboratory reported.

The machine learning model was built using a database of rare earths from the Ames Laboratory itself, made up of around 600 chemical elements, as well as the use of the quantum calculation method known as density functional theory, used for the calculation. analysis of the thermodynamic and electronic properties of rare earth compounds.

robot mas pequeno del mundo
agua en la superficie de 2 asteroides
robot enviado a la EEI
robot submarino 1
mision lunar Odie
Students built the world’s smallest robot
Scientists find water on the surface of two asteroids
Robot sent to the ISS can perform operations in space
Autonomous underwater robot developed to combat biofouling on ships
The “Odie” lunar mission successfully lifted off in the U.S.
previous arrowprevious arrow
next arrownext arrow

In addition, this model uses regression learning based on the SISSO method (from the English acronym for Safe Independence Detection and Dispersion Operator), which allows evaluating the energy stability of the phases of the compounds, that is, the computer system will be able to determine whether or not a combination of rare earths will be viable.

“Machine learning is really important here because when we talk about new compositions, all the ordered materials are very well known in the rare earth community,” said Prashant Singh, a scientist at Ames Laboratory and lead author of the research, published in the journal Material Act.

“However, when you add disorder to known materials, it’s very different,” Singh said, explaining that this is because sometimes you can’t “investigate all possible combinations using theory or experiments,” because the amount of compositions may be higher. Order and disorder is a reference to how the particles are distributed in the material, directly influencing its properties.

The scientist mentioned that the experimental data can be fed back to the machine learning system in order to reduce errors, since they will be able to find compounds that would not work in real life.

“It’s not really meant to discover a particular compound,” said Yaroslav Mudryk, supervisor of the research, adding that there were questions at the time of the design and construction of the computer system used for “rare earth discovery and prediction.”

Mudryk reiterated that this is the beginning, as the team of researchers is studying the potential of the machine learning model to be used in a wide range of applications in the future.

Source RT news in Spanish :

Source Ames Laboratory :

Share this news on your social networks

Rate this post
1 star2 stars3 stars4 stars5 stars (No rating yet)