Research Fields: Shang-Hua Yang

Communication Group & Active Device Group

High-speed wireless communication has received great attentions in recent decades due to significant changes in the ways people produce and transfer information. However, the current wireless communication data rate is limited by the narrow bandwidth of existing sources and the extensive utilization of the EM spectrum up to 100 GHz. This trend has compelled researchers to raise carrier frequencies to operate in frequency bands that have not been assigned to any specific active services, which falls in the scope of the THz regime. In this research work, a heterogeneous communication system capable of delivering 5G/sub-THz signal carriers over an arbitrary long fiber and separated transmission links is presented by employing direct detection, multiplexing techniques, and advanced digital signal processing. The well-demonstrated heterogeneous system is compatible with radio-over-fiber technology, cost-effective, and easy to deploy, making it a promising candidate for indoor THz communication.

In this research, a statistical model is employed in the channel estimation section to enhance the signal-to-noise ratio of the received signal. Subsequently, a nonlinear equalizer based on machine learning is applied to the system to enhance the overall signal quality. To tackle the complexity of neural networks, regularization terms are employed to exploit insignificant paths within the neural network, thereby significantly reducing the overall system complexity. The proposed method demonstrates a significant improvement in system performance while minimizing signal delay.
˙Performance and complexity analysis using a sparse deep learning method for indoor terahertz transmission (Opt. Lett. 47, 4431-4434 (2022))
˙A 5G/Sub-Terahertz Heterogeneous Communication Network (IEEE. Access, 10, 65572-65584, (2022))

We propose a low-cost, mass-producible THz photoconductive emitter based on epitaxial GeSn alloy through all CMOS-compatible processes. Unlike conventional THz photoconductive emitters, they are mainly based on MBE-grown III-V materials requiring precise control epitaxial conditions to obtain a high-quality structure. This time-consuming and costly fabrication process limits the THz photoconductive emitters from mass production. We have comprehensively studied the alternative group IV material candidate, RPCVD-grown GeSn alloy, for high-performance THz emitter from material growth to material/device characterization compared with a conventional group III-V (InGaAs/InAlAs) THz emitter and a group IV (Ge). The demonstrated group IV THz optoelectronics shows its potential in realizing mass-producible, cost-effective THz system-on-chip (SoC) with CMOS technology, paving the way for rapidly visualizing multi-dimensional information of advanced material and packaging process for next-gen power devices and optoelectronics.
˙Group IV THz large area emitter based on GeSn alloy, (Opt. Lett. 47, 4411-4414 (2022))
˙Thermal evaporated group IV Ge(Sn)-on-Si terahertz photoconductive antenna, (Opt. Express, 30, 31742-31751 (2022))

Imaging Group

Sub-THz imaging has become increasingly essential for applications such as real-time non-invasive imaging, industrial inspection, material identification, and sensing. This surge in demand is attributed to the rapid progress in high-power THz coherent sources and the development of extensive THz detector arrays. Coherent sub-THz waves, inherently providing phase information and introducing diffraction phenomena, offer the potential for generating speckle-free images. Although deliberate control of wave coherency is a common practice in EM bands using diffusors, such solutions are not readily available commercially in the sub-THz band. In this study, we address this gap by designing and fabricating cost-effective sub-THz diffusors using polymer-based materials. The outcome is a real-time sub-THz imaging system capable of delivering spatial resolution, approaching the diffraction limit. This innovation holds promise for advancing applications in diverse fields requiring high-quality THz imaging and precise material analysis.

Visualizing information inside objects is an everlasting need to bridge the world from physics, chemistry, and biology to computation. Among all tomographic techniques, terahertz computational imaging has demonstrated its unique sensing features to digitalize multi-dimensional object information in a non-destructive, non-ionizing, and non-invasive way. However, there is still a gap between object information and signal properties that wait to be explored. Here, we opened a new research field – physics-guided terahertz deep learning computed tomography framework – to provide a crosslink between material digitalization, functional property extraction, and multi-dimensional imager utilization. The research demonstrates a new computer vision approach to utilize terahertz signals and move toward data-driven, multi-modal-fusion terahertz computational imaging. (OE 22’, IEEE SPM 23’, IJCV 23’)

Compressed sensing is a popular technique in digital signal processing society, where signal can be reconstructed through specific algorithm in under-sampled scheme. When this technique is applied to THz technology, one can perform THz imaging with much faster speed than conventional raster scanning scheme. Furthermore, this combination also allows one to perform multi-pixel imaging without moving the sample, realizing true “non-destructive” and “remote” sensing of given samples.

Our aim in this direction includes:
˙CMOS-compatible optoelectronic THz devices (OE 22’, OL 22’)
˙THz zoom optical system
˙Development of THz spatial light modulators
˙Channel modeling for image quality optimization
˙Signal reconstruction algorithm designated for FPGAs (IEEE OJCS 22’)
˙Table-top THz compressed sensing imager system

Passive Device Group

In this study we have developed a novel composite photoresin for 3D printing. This composite enhances the refractive index of commercial UV resins, boosting it from approximately 1.75 to 1.95, achieved by incorporating nanocomposites. We successfully 3D printed a plano-convex lens which is about 40% thiner than normal 3D printed material. Furthermore, we assessed its performance within a 6G communication system, revealing promising results. This advancement holds potential for the cost-effective production of compact THz passive devices, with applications in emerging communication, imaging, and spectroscopy applications.

This study presents a comprehensive exposition of Masked Stereolithography as applied to the 3D printing of THz diffractive lenses. Diffractive lenses hold considerable potential in diverse applications encompassing compact THz imaging, spectroscopy, and communication systems. Notably, the congruence observed between our simulation outcomes and experimental findings serves to substantiate the efficacy of the proposed methodology. Remarkably, our findings indicate an exceedingly modest cost structure, with an approximate expenditure of less than one US dollar per unit. This economic consideration underscores the pragmatic feasibility and scalability of the proposed 3D printing methodology for THz diffractive lenses, thereby contributing substantively to the discourse on advanced manufacturing techniques in the domain of THz technology.

Research Field: Wei-Chih Wang

Micro Technology Laboratory

˙Tunable THz filter using electro-optic Fishnet metamaterial
˙Various Wave manipulator
˙Mid-Infrared Broad band Radiation Energy Harvesting Device
˙Prism beam steering
˙THz Gas Detection Using Cellulose Nanoporous Polymer Enhanced Meta Structure
˙Planar Gradient Metamaterial Wave manipulator
˙Various high efficient absorber designs used in sensing and energy conversion


Current metamaterial research focusing on developing THz band wave manipulation, imaging, sensing, and energy harvesting devices using polymer based 2D or 3D active meta structure with on-demand structure-property or material modification.


Research Fields: Kuo-Ping Chen

Dielectric Metasurfaces

Lab: Institute of Photonics Tech.

˙Kerker Effects

Left: ACS Photonics (2018);    Right: ACS Nano (2022)

Achieving a high quantum efficiency at the wavelength near the absorptance edge is becoming more critical. With optical near-field enhancement in particles, the quantum efficiency of the photodetector can be improved. Because of the high-refractive-index (HRI) of Germanium, germanium nanoparticles can induce enhanced electric and magnetic resonances with a strong electromagnetic field inside particles and support Mie resonances in the NIR spectral range with the characteristic dimensions of 250-350 nm. There are large forward and backward scattering signals of Mie resonances, and controlling the directional scattering of nanoparticles has been actively investigated in the past few years. Using Mie theory for a spherical dielectric nanoparticle, the destructive interference between electric dipole (ED) and magnetic dipole (MD) resonances will result in scattering cancellation at the backward direction when the resonance wavelength of ED and MD are close to each other. This condition for the zero backward scattering has been derived by Kerker et al.

˙Lattice Resonance

Left: ACS Nano (2020);    Right: Lasers and Photonics Review (2021)

Bound states in the continuum (BICs) have attracted considerable attention due to their infinite quality factor (Q-factor) bound state and extremely localized field that drastically enhances light–matter interactions and offers great potential in topological photonics and quantum optics. In this study, we demonstrated directional emission normal to a bound state in the continuum (BIC) metasurface laser with hybrid surface lattice resonances (SLRs). Compared to the nanolaser, the BIC metasurface laser possesses directional radiation and a large emission volume, and the high Q-factor resonance overcomes the limitation of large mode volume in achieving thresholdless lasing. Interestingly, due to the high Q-factor resonance of BICs, the laser signals and images can be observed in almost transparent samples. The novel features of metasurfaces provide the way for engineering BICs. The developed device can be used in various applications, including novel light sources, optical sensing, nonlinear optics, and topological photonics.

Ultra Small Color Pixels


The works demonstrates the development of nanophotonic circuits by using graphene to detect the unidirectional surface plasmon polariton (SPP) propagation in a "nonscattering" approach. This device shows the strong impact of the nanoplasmonics integration system, as in nanolasers, SPP waveguides, and SPP modulation. (July 6, 2022, Volume 14, Issue 26)