Research

Optical Metamaterials

Metamaterials, artificially structured nanomaterials, have enabled the extraordinary optical properties and unprecedented phenomena such as invisibility cloaking and negative refraction. We are designing and conducting experimental realization of novel negative index metamaterials, chiral metamaterials, hyperbolic metamaterials and metasurface working with low loss and broad bandwidth. 



Acoustic and Elastic Metamaterials

Inspired by intriguing phenomena from solid-state physics and photonics, the classical waves such as acoustic and elastic waves can be much developed through coupling applied physics in a new way. In particular, we are investigating acoustic/elastic/seismic metamaterials/phononic crystals via effective parameter and bandstructure engineering in a long wavelength regime or relevant scale.


Plasmonics

Due to plasmonic mode excitation of high-k vector access in dispersion relation, light can be confined to deep sub-wavelength. We are not only investigating the energy transportation mechanisms and applications at the nanometer length scale, but also studying plasmon-phonon interaction for the non-linear properties of light. 


Topological Photonics

The recent advances in topological photonics have unveiled many peculiar phenomena such as Dirac/Weyl degeneracies and back-scattering immune surface waves. We are investigating topological metamaterials possessing topologically non-trivial branches in a momentum space by tailoring optical properties of metamaterials. Furthermore, a two-dimensional topological degeneracy, which can be also called as Weyl nodal surface, is observed in a metamaterial with nonsymmorphic symmetry.


Device Applications

With the properties of unnaturally high wavevectors access, super-resolution hyperlens imaging beyond diffraction limit has been demonstrated from our group. Its superior performances such as super-resolution, real-time and non-vacuum working environment open a new possibility for nanoscale biological imaging, which can be very practical. Such efforts to make metamaterials and plasmonics more practical continue to realize metadevices. Another interest is integrating metamaterials into the MEMS/NEMS devices to realize reconfigurable and actively controllable metadevices.

 

Nanofabrication and Nanomanufacturing

Metamaterials provided a huge potential to realize scientific fictions and change the world. Current metamaterials demonstration is relies on nanofabrication, so we are pursuing practical nanofabrication techniques useful for metamaterials structure realization such as ultra-high precision electron beam lithography overlay and ultra-thin/smooth thin film deposition. However, nanofabrication is very low-throughput and high-cost technologies. Experimental demonstration and applications of scalable nanomanufacutring methodologies include, but not limited to, hyperlens based 2D lithography and 3D printing (sub-50nm), plasmonic maskless flying head lithography (sub-22nm), single quantum dot patterning (sub-15nm), nano-cascade patterning (sub-10nm) and other scalable atomic/molecular level techniques such as mechanical reduction. Moreover, we are doing state-of-the-art bottom-up nanofabrication using self-assembly, nanoparticles, block co-polymer and etc to apply metamaterials and plasmonics applications. 




Design and Optimization (Artificial Intelligence, Machine Learning, Deep Learning)

In recent years, deep-learning, a subset of artificial intelligence, has emerged as a strong and efficient framework to solve many science or engineering problems. In nanophotonics, a deep-learning has been introduced either to predict optical properties or to design nanophotonic structures. Using vast amounts of simulation or experimental data, artificial neural networks can learn the mapping between various optical structures and their optical properties. This can be applied to inverse design problem of nanophotonics, which was mainly conducted by a time-consuming and laborious method of optimization. Once the network is trained, the trained network provides adequate structural designs without additional computational costs. This can greatly reduce the computational costs of multiple design tasks.


Comments