Density Functional Theory for Nuclear Science

"Physics & Astronomy"
21 followers
Aug 27, 2026
What are the limits of nuclear existence? How can we use observed nuclei to expand our incomplete picture of the nuclear landscape? This video highlights how researchers at the Facility for Rare Isotope Beams at Michigan State University address these, and other fundamental questions by using nuclear Density Functional Theory (DFT).
AI Generated·May contain errors
Video Summary
AI GeneratedThis video explores how the Facility for Rare Isotope Beams (FRIB) at Michigan State University uses Density Functional Theory (DFT) and advanced computational methods to study the nuclear landscape. By modeling the interactions of protons and neutrons, researchers can predict the properties of rare isotopes and understand the cosmic origins of heavy elements.
How were heavy elements like gold and uranium created in our universe?
What are the limits of nuclear existence?
How can we use observed nuclei to expand our incomplete picture of the nuclear landscape?
What can nuclei that we study here on Earth
tell us about processes that are happening within neutron stars?
And how can we study the ways in which nuclei decay and transform?
These are all just a few of the fundamental questions that we study here at EFRIB,
the Facility for Rare Isotope Beams at Michigan State University.
The many-body wave function of a nucleus is an incredibly complex object.
It contains all the correlations between protons and neutrons.
Information grows rapidly with every additional particle.
Solving this problem computationally becomes extremely difficult.
Density functional theory, or DFT for short, provides a powerful alternative
by replacing the countless interactions between individual protons and neutrons
with effective dynamics driven by the nucleonic densities.
The calculation becomes a conversation.
The mean field sets the densities,
and the densities in turn modify the mean field until the nucleus settles.
To make it all work, we need an object called the energy density functional,
which is a recipe that ties together energy and density.
Its couplings are finely tuned to reproduce available experimental data,
such as binding energies or charged radii.
Each tuning gives a distinct model of the nucleus.
Once calibrated, these DFT models allow us to extrapolate
to the unseen parts of the nuclear charge.
DFT allows us to connect phenomena that takes place across vast scales of time.
For example, when nuclei collide, they exchange protons and neutrons at timescales
on the order of a billionth of a trillionth of a second.
DFT lets us explore the dynamics of these near-instantaneous events
following the evolution of a nascent compound as it forms.
Today, time-dependent DFT is helping us predict fission yields,
understand the formation of super-heavy elements,
and gain insight into fundamental properties of atomic nuclei.
The heaviest elements in nature are forced by the rapid neutron capture process,
or the R-process, in extreme environments like neutron star mergers.
DFT predicts the masses, decay rates,
and fission properties of short-lived nuclei that we cannot yet measure.
These predictions are essential for R-process simulations,
while the same theoretical framework helps describe the structure of neutron stars.
With DFT, we can open a window from our laboratories to the stars and our cosmic origins.
While there are many different effective nuclear DFT models,
each of them have physical differences that make them disagree
when we predict nuclei that we have never observed before.
Combining different flavors of nuclear physics
tells us how confident to be in our predictions of nuclei that we don't know.
Bayesian model combination lets us blend what we have
to make something even better than a single flavor.
Today's functionals are remarkable approximations, but they don't capture everything.
Instead of handcrafting every improvement,
we can use machine learning methods like symbolic regression and genetic programming
to evolve candidate equations, revealing new ways to describe the nuclei.
However, doing so requires evaluating many, many possibilities, which becomes quite expensive.
So, we have to come up with a way to mitigate these expenses
without sacrificing the physics precision.
We can combat this problem by building emulators.
Consider this piano chord.
It produces a rich and complicated response,
yet much of what we hear can be described by just a few dominant resonances.
We can use a similar idea to compress expensive nuclear calculations,
reproducing them with sub-percent accuracy while running them millions of times faster.
This opens the door to physics I was previously out of reach,
bringing us closer to answering the fundamental questions that we started with.
If you'd like to be a part of the science, we invite you to connect with our community
and explore more of the resources we've collected here.
