Chemists Open Secrets Of Molten Salts

Chemists Open Secrets Of Molten Salts

University of Cincinnati computational chemist Yu Shi. Credit: Andrew Higley, University of Cincinnati

A chemist at the College of Cincinnati has come up with a unique means to examine the thermodynamic properties of molten salts that are utilized in numerous nuclear and solar power applications.

UC College of Arts and Sciences research join and computational chemist Yu Shi and his collaborators developed a brand-new simulation method to calculate free power using deep learning artificial intelligence.

Molten salt is salt heated up to enormous temperatures where it turns liquid. UC researchers examined sodium chloride, commonly known as common salt. Shi stated molten salt has properties which make it a valuable medium for cooling down systems in nuclear energy plants. In solar towers, they could be used to transfer warm or keep power.

Paradoxically, while salt is an insulator, molten salt conducts energy.

“Molten salts are stable at huge temperatures and can hold much energy in a liquid state”, Shi stated. “They have great thermodynamic properties. That makes them a good energy storage product for concentrated solar power plants. Moreover, they can be utilized as a coolant in nuclear reactors”.

Posted in the Royal Society of Chemistry journal Chemical Science, the study might help scientists analyze the corrosion that these salts could cause in metal containers like those located in the next generation of nuclear reactors.

The research provides a reliable strategy for converting dissolved gas to vapor in molten salts, helping engineers understand the effect of unusual impurities and solutes (the substance dissolved in a solution) on corrosion. She said it also would help researchers study the release of potentially toxic gas into the atmosphere, which will undoubtedly be highly beneficial for 4th-generation molten salt nuclear reactors.

“We utilized our quasi-chemical theory and our deep neural network, which we trained making use of information created by quantum simulations, modeling the solvation thermodynamics of molten salt with chemical accuracy”, Shi said.

Study co-author Thomas Beck is a former head of UC’s Department of Chemistry and right now works as section headman of science engagement for the Oak Ridge National Lab in Tennessee. Beck stated molten salts do not expand when heated, unlike water, that can develop extreme pressure at enormous temperatures.

“The power inside a nuclear reactor goes up a lot. That is the difficulty of reactor design– it leads to additional risks and higher costs”, he said.

Scientists turned to UC’s Advanced Research Computing Facility also the Ohio Supercomputer Center to run the simulations.

“At Oak Ridge, we have the world’s rapid supercomputer, so our experiment may take less time here”, Beck stated. “But on typical supercomputers, it could take weeks or months to run these quantum simulations”.

The study group also consisted of Stephen Lam at the University of Massachusetts Lowell.

“It is important to have accurate models of these salts. We were the first team to calculate free power of sodium chloride at huge temperature in liquid and compare it to former experiments”, Beck said. “So we proved it is a helpful technique”.

In 2020, Shi and Beck established a free-energy scale for single-ion hydration, utilizing quasi-chemical concept and quantum mechanical simulations of the salt ion in water in a study released in the journal PNAS. It was the 1st solvation free-energy calculation for the charged solute utilizing quantum mechanics, Shi said.

Beck said molten salts would undoubtedly be essential for developing brand-new sources of energy– even probably one-day combination power.

“They are proposing using molten salts as a coating coolant for the high-temperature activator”, he said. “But the combination is farther down the road”.


Reference:

Yu Shi et al, Deep neural network based quantum simulations and quasichemical theory for accurate modeling of molten salt thermodynamics, Chemical Science (2022). DOI: 10.1039/D2SC02227C

Read the original article on PHYS.

Share this post