Ab Initio Molecular Dynamics for Molten Salt Structure Calculations
Matthew Louis / May 2025 (352 Words, 2 Minutes)
Ionic lquids are interesting in their own right due to the interesting chemical properties, which allow them to be used for carbon capture, among other things. In the energy production industry, inorganic ionic liquids: particularly molten salts have proven a very useful heat transfer fluid. Their high heat capacity and high boiling point at low pressure have made them useful in concentrated solarpower for thermal storage. These useful fluids have also found use in novel Generation IV nuclear reactors. Famously, a series of experiments at Oak Ridge National Lab proofed the use of molten salts as a simultaneous coolant and fuel carrier culminating with the construction of the Molten Salt Reactor Experiment: the as yet only operating molten salt reactor in the United States (though hopefully not for long). Due to the then current administration’s preference for the liquid metal fast breeder reactor, this technology was shelved, but has recently had a great resurgence of interest, and companies like TerraPower, Moltex, Natura Resources, and several others are working to build commercial molten salt fueled reactors (and plenty of other companies are considering molten salt cooled reactors, like Kairos, who have already obtained a construction permit).
All that to say: molten salts are important fluids and they have been receiving a lot of attention as of late. Despite this, there is still a huge lack of high quality data on the thermophysical properties of these salts (especially actinide bearing salts necessary for nuclear applications). In short, due to the high melting points, and the huge variance of thermophysical properties in different multicomponent molten salt system (which have a intractably large composition space to adequately span), and cost, these salts are hard to measure experimentally. Thanks to this, these liquids fall into the rare category where it can actually be more cost efffective to simulate them using high fidelity quantum chemistry calculations rather than directly measuring them. Due to the complex ionic interactions of the elements that compose these salts, full electronic structure calculations are needed for adequate force calculations, and often Ab Initio Molecular Dynamics (AIMD) calculations are needed for thermophysical property simulation. As a nuclear engineer, I’ve always gawked at these calculations for how impossible they seem: simulating huge ensembles of atoms over large timesteps using quantum mechanics… seems pretty impossible.
Last semester, I took a computational chemistry class (CHEM 4485), that surveyed all of the clever techniques used to make many body quantum mechanics computationally tractable for many very large systems. For a class project, I decided to try my hand at some AIMD for molten salt calculations. I learned how to use CP2K (and how to compile it with GPU support using spack, but that’s a story for another day) and I performed a very basic structure calculation for molten UCl3 salt at a single temperature. This is a far cry from the long time (and at multiple temperature) AIMD simulations needed for determining useful thermophysical and transport properties, but it was a nice start, and, if nothing else, really gave me a greater appreciation for what a feat these calculations really are. Below is my writeup for anyone who wants to learn more!
Comments
If you're logged in, you may write a comment here. (markdown formatting is supported)
No comments yet.