Understanding how electrons move inside materials is essential for revealing the properties of functional and quantum materials. In this study, we developed a method for reconstructing the electron momentum distribution and Fermi surface with high accuracy from far fewer measurements than previously required by introducing a technique known as compressed sensing into the analysis of electron-state experiments.
Researcher from the Research Center
Yoshinori Nakanishi-Ohno (Faculty of Culture and Information Science)
Research Background
The motion of electrons in a material determines many of its physical properties, including electrical conductivity, magnetism, and superconductivity. In metallic materials, understanding the Fermi surface—the boundary that separates occupied and unoccupied electron states—is particularly important because it governs how electrons move through the material.
One experimental technique used to investigate electron motion is Compton scattering. However, reconstructing a three-dimensional Fermi surface from Compton scattering data typically requires measurements from many different directions, making experiments time-consuming and resource-intensive.
Research Overview
In this study, we applied the concept of compressed sensing, a technique widely used in medical image reconstruction and signal processing, to the analysis of Compton scattering data.
Our approach takes advantage of the fact that the electron momentum distribution is smooth over most regions of momentum space and changes significantly only near the Fermi surface. Based on this characteristic, we introduced a sparse optimization technique known as generalized LASSO and developed an algorithm capable of estimating the most plausible three-dimensional electronic structure from a limited number of measurements.
Key Findings
We evaluated the proposed method using body-centered cubic (bcc) lithium metal and demonstrated that the three-dimensional electron momentum distribution and Fermi surface can be reconstructed with high accuracy using Compton scattering data from only 14 measurement directions.
Furthermore, the method remained robust even when measurement noise was added to the data, successfully preserving key features of the Fermi surface. These results suggest that Compton scattering experiments, which have traditionally required long measurement times, can be performed much more efficiently. The proposed approach is expected to facilitate future studies of a wide range of electronic and quantum materials.
Publication Information
Otsuki, J., Yoshimi, K., Nakanishi-Ohno, Y., Sekania, M., Chioncel, L., & Mizumaki, M. (2026). Compressed Sensing of Compton Profiles for Fermi Surface Reconstruction: Concept and Implementation. Journal of the Physical Society of Japan, 95(7), 074707.