The Science
The inner crust of a neutron star is characterized by the presence of a neutron superfluid. A superfluid is a fluid that has no viscosity. In a neutron star, this means the superfluid allows neutrons to flow without resistance. To accurately predict the properties of neutron matter at its lowest energy levels in this low-density form, researchers make theoretical calculations that typically assume that neutrons join together to form “Cooper pairs.” This study used artificial neural networks to make accurate predictions without relying on this assumption. The study modified the standard “single-particle” approach by introducing “hidden” neutrons that facilitate interactions among the “real” neutrons and encode quantum many-body correlations. This allows Cooper pairs to emerge naturally during the calculation.
The Impact
Understanding neutron superfluidity provides important insights into neutron stars. It sheds light on their cooling mechanisms, their rotation, and phenomena such as glitches—sudden changes in their rotation rate. While scientists cannot directly access neutron star matter experimentally, the fundamental interactions that govern this matter’s behavior are the same as those that govern atomic nuclei on Earth. Researchers are working to construct nuclear interactions that are simple, yet predictive. Accurately solving the quantum many-body problem is a crucial part of assessing the quality of these interactions. This work uses simple interactions that agree well with previous calculations that assume much more complex interactions.
Summary
Low-density neutron matter is characterized by fascinating emergent quantum phenomena, such as the formation of Cooper pairs and the onset of superfluidity. Researchers used artificial neural networks alongside advanced optimization techniques to study this density regime. Using a simplified model of the interactions between neutrons, the researchers calculated the energy per particle and compared the results to those obtained from highly realistic interactions. This approach is competitive with other computational methods at a fraction of the cost.
Funding
This work is supported by the Department of Energy (DOE) Office of Science, Office of Nuclear Physics, the DOE Office of Science, Office of Advanced Scientific Computing Research, Scientific Discovery through Advanced Computing (SciDAC) NUCLEI program, and the U.S. National Science Foundation. Numerical calculations used the Laboratory Computing Resource Center at Argonne National Laboratory and the computers of the Argonne Leadership Computing Facility, a DOE Office of Science user facility.
