Supercomputer Adds Efficiency to Light-Matter Simulations

Researchers at the University of Tsukuba developed a unified method for simulating light-matter interaction at the atomic scale. The method is considered to be highly efficient by the researchers, as it simultaneously calculates electromagnetic analysis for light electromagnetic fields: ab initio time-dependent density functional theory (TDDFT) for electron dynamics, and molecular dynamics for ionic motion.

The simulation method addresses the three physical laws that apply to light-matter interaction: electromagnetism for light fields, quantum mechanics for electrons, and Newtonian mechanics for ionic motion.

It also addresses the multiscale nature of light-matter interaction; a wavelength of light is typically 1 μm, whereas the motion of electrons and ions is measured in nanometers or less. Time periods of light and electronic excitations are usually measured in femtoseconds, while atomic motion is measured in picoseconds.

Researchers led by the University of Tsukuba developed a computational approach for simulating interactions between matter and light at the atomic scale. The team tested its method by modeling light-matter interactions in a thin film of amorphous silicon dioxide, composed of more than 10,000 atoms, using the world’s fastest supercomputer, Fugaku. The proposed approach is highly efficient and could be used to study a wide range of phenomena in nanoscale optics and photonics. Courtesy of the University of Tsukuba.

Because it is multifaceted, light-matter interaction is typically modeled using two separate computational methods. The first method is an electromagnetic analysis of the light fields. The second is an ab initio quantum-mechanical calculation of the optical properties of the material. This approach assumes that the electromagnetic fields are weak and that there is a difference in the length scale.

It is of limited use in current research, according to the Tsukuba team.

“Our approach provides a unified and improved way to simulate light-matter interactions,” professor Kazuhiro Yabana said. “We achieve this feat by simultaneously solving three key physics equations: the Maxwell equation for the electromagnetic fields, the time-dependent Kohn-Sham equation for the electrons, and the Newton equation for the ions.”

The new method could be used to examine the size and timescales encountered in practical problems. For example, when pulsed light irradiates a bulk material, the emission of electrons and ions can occur on the surface, and the electronic structure at the surface may differ from the bulk system. The simulation method could be used to calculate materials that are several nanometers thick (several tens of layers of atoms) from the surface to investigate light-matter interaction in the bulk material.

The time step used in the simulation was compared with physical timescales for typical laser pulses, electron excitations, and ion motion, and with the time it takes to transfer energy from light to electrons and from electrons to ions.

To achieve this approach to simulating light-matter interaction, the code needed to be scalable and efficient. The researchers implemented the method in their in-house software, SALMON (scalable ab initio light-matter simulator for optics and nanoscience), and optimized the simulation computer code on a Fujitsu A64FX processor to maximize its performance.

They tested the code by modeling light-matter interactions in a thin film of amorphous silicon dioxide composed of more than 10,000 atoms. This simulation was carried out using almost 28,000 nodes of the Fugaku supercomputer at the RIKEN Center for Computational Science in Kobe, Japan. The simulation test results demonstrated excellent time to solution.

“We found that our code is extremely efficient, achieving the goal of one second per time step of the calculation that is needed for practical applications,” Yabana said. “The performance is close to its maximum possible value, set by the bandwidth of the computer memory, and the code has the desirable property of excellent weak scalability.”

The researchers expect to achieve greater flexibility in parallelization using the entire Fugaku system, which could lead to even better performance.

Although the researchers focused on simulating light-matter interactions in a thin-film material that was several nanometers thick, a variety of phenomena in nanoscale optics and photonics could be examined by large-scale calculations using SALMON. It is important to model not only film materials but a prototype of bulk surfaces, the researchers believe. They also think that the interaction of pulsed light with nanomaterials of various 3D structures, either isolated or placed on surfaces, could be extremely interesting to model.

The research was published in The International Journal of High-Performance Computer Applications (www.doi.org/10.1177/10943420211065723).