Search
Menu
Photonics HandbookResearch & Technology

Nanoresonator for 6G Enhances THz Waves 30,000-Fold

Facebook X LinkedIn Email
ULSAN, South Korea, Jan. 4, 2024 — The swift development of 6G communications using terahertz (THz) electromagnetic waves has intensified the demand for highly sensitive nanoresonators that can detect these waves.

Responding to this need, researchers at Ulsan National Institute of Science and Technology (UNIST), in collaboration with the University of Tennessee and Oak Ridge National Laboratory, optimized a THz nanoresonator specifically for 6G communications using artificial intelligence.

The optimized nanoresonator is versatile, with potential applications for ultra-precise detectors, ultrasmall molecular detection sensors, and bolometer studies, according to researcher Young-Taek Lee.

The team’s analytical, model-based approach significantly reduces the computational resources required to optimize THz nanoresonators and offers a practical alternative to numerical, simulation-based inverse designs for THz nanodevices. To put the advancement into perspective, the process enables efficient design of THz nanoresonators on personal computers, a process which was previously time-consuming and demanding even with supercomputers.
A research team led by professor Hyong-Ryeol Park at UNIST developed a technology capable of amplifying THz electromagnetic waves by more than 30,000 times. The new nanoresonator, which was developed using a rapid inverse design method combined with AI based on physical models, could catalyze the commercialization of 6G communication frequencies. Courtesy of Nano Letters (2023). DOI: 10.1021/acs.nanolett.3c03572.
A research team led by professor Hyong-Ryeol Park at UNIST developed a technology capable of amplifying terahertz electromagnetic waves by >30,000×. The new nanoresonator, which was developed using a rapid inverse design method combined with AI based on physical models, could catalyze the commercialization of 6G communication frequencies. Courtesy of Nano Letters (www.doi.org/10.1021/acs.nanolett.3c03572).

In about 39 hours using a mid-level PC, the researchers identified the optimal structure through 200,000 iterations and achieved an experimental electric field enhancement of 32,000 at 0.2 THz. The electric field generated by the THz nanoresonator, which surpassed general electromagnetic waves by >30,000×, demonstrated an efficiency improvement of >300% compared to previously reported THz nanoresonators.


AI-based inverse design technology has typically been used for optical device structures in the visible and infrared areas, which are only a fraction of the wavelength. The use of AI-based inverse design for the 6G communications frequency range (0.075 to 0.3 THz) presented the researchers with significant challenges, due to the small scale — approximately one-millionth the size of the wavelength — of the range.

The researchers worked with nanogap loop arrays, a type of resonator that has demonstrated the potential to detect THz electromagnetic waves. Because the unit cells of these arrays are 10× smaller than millimeter wavelengths, with nanogap regions that are 1,000,000× smaller, they require significant computational resources for accurate simulation.

To improve the efficiency of nanogap loop arrays, the researchers combined the nanoresonators with a rapid inverse design method based on physics-informed machine learning. Specifically, the inverse design approach used double deep Q-learning with an analytical model of the THz nanogap loop array.

Professor Hyong-Ryeol Park, who led the research, stressed the need to understand physical phenomena in conjunction with AI technology. “While AI may appear to be the solution to all problems, comprehending physical phenomena remains crucial,” he said.

The team evaluated the efficiency of the new nanoresonator through a series of THz electromagnetic wave transmission experiments conducted in simulation with the help of physics-informed machine learning.

“The methodology employed in this study is not limited to specific nanostructures, but can be extended to various studies using physical theoretical models of different wavelengths or structures,” Lee said.

The research was published in Nano Letters (www.doi.org/10.1021/acs.nanolett.3c03572).

Published: January 2024
Glossary
artificial intelligence
The ability of a machine to perform certain complex functions normally associated with human intelligence, such as judgment, pattern recognition, understanding, learning, planning, and problem solving.
deep learning
Deep learning is a subset of machine learning that involves the use of artificial neural networks to model and solve complex problems. The term "deep" in deep learning refers to the use of deep neural networks, which are neural networks with multiple layers (deep architectures). These networks, often called deep neural networks or deep neural architectures, have the ability to automatically learn hierarchical representations of data. Key concepts and components of deep learning include: ...
nano
An SI prefix meaning one billionth (10-9). Nano can also be used to indicate the study of atoms, molecules and other structures and particles on the nanometer scale. Nano-optics (also referred to as nanophotonics), for example, is the study of how light and light-matter interactions behave on the nanometer scale. See nanophotonics.
terahertz
Terahertz (THz) refers to a unit of frequency in the electromagnetic spectrum, denoting waves with frequencies between 0.1 and 10 terahertz. One terahertz is equivalent to one trillion hertz, or cycles per second. The terahertz frequency range falls between the microwave and infrared regions of the electromagnetic spectrum. Key points about terahertz include: Frequency range: The terahertz range spans from approximately 0.1 terahertz (100 gigahertz) to 10 terahertz. This corresponds to...
Research & TechnologyeducationAsia-PacificLight SourcesOpticsSensors & Detectorsartificial intelligencedeep learningCommunications6Gnanoresonatorsoptical resonatorsnanoterahertzterahertz nanoresonatorphysics-informed machine learninginverse designdouble deep Q-learningnanogap loop arrayelectromagnetic waves

We use cookies to improve user experience and analyze our website traffic as stated in our Privacy Policy. By using this website, you agree to the use of cookies unless you have disabled them.