Meta-Optics for AR Waveguides

Overview: This project focuses on overcoming the limitations of traditional refractive optics in Augmented Reality (AR) systems. By leveraging metasurfaces—planar arrays of sub-wavelength nanostructures—we can manipulate the phase, amplitude, and polarization of light with unprecedented control.

Key Research Areas:

• Field of View (FOV) Expansion: Using high-index materials and optimized grating designs to push the FOV limit beyond 50 degrees.
• Efficiency Optimization: Addressing the 'rainbow effect' and uniform light extraction through precise nano-imprint lithography.
• Full-Color Display: Developing multiplexed metasurfaces that can diffract RGB wavelengths simultaneously without crosstalk.

Our approach integrates theoretical simulation (RCWA/FDTD) with extensive cleanroom fabrication (E-beam lithography, RIE) to create lightweight, high-performance AR glasses.

Metalens & Large-Area Meta-Optics

Overview: Metalenses are planar optical elements composed of sub-wavelength nanostructures that can manipulate light wavefronts with high precision. Our lab pioneers large-area metalens fabrication using deep-UV KrF photolithography, enabling scalable manufacturing beyond traditional e-beam lithography limitations.

Key Research Areas:

• Large-Area Fabrication: Achieving 2cm x 2cm metalens arrays with consistent optical performance using KrF stepper lithography and optimized process parameters.
• AI-Assisted Design: Implementing Intelligent Proximity Correction (IPC) algorithms to compensate for optical proximity effects and achieve precise nanostructure patterning.
• High Focusing Efficiency: Demonstrating 60-64% focusing efficiency for NIR metalenses through Jacobian phase error correction and optimized MTF performance.
• Multifocal Applications: Developing metalens systems for Vergence-Accommodation Conflict (VAC) mitigation in AR/VR displays through precise depth-plane generation.

This research bridges the gap between laboratory-scale metasurface demonstrations and industrial-scale production for consumer optics applications.

Resolution Enhancement Techniques (RET)

Overview: As semiconductor nodes shrink below 3nm, traditional imaging limits pose a massive challenge. Our lab develops advanced computational algorithms to extend the life of current lithography equipment.

Methodologies:

• Source-Mask Optimization (SMO): Simultaneously optimizing the illumination source shape and the mask pattern to maximize the process window.
• Inverse Lithography Technology (ILT): Using pixel-based optimization to create complex, curvilinear mask shapes that print perfectly square wafers features.
• Optical Proximity Correction (OPC): Machine-learning enhanced OPC models to predict and correct layout hotspots faster than conventional rigorous simulation.

This research has direct applications in high-volume manufacturing, aiming to reduce Edge Placement Error (EPE) and improve yield.

---

OPC+ Cloud Service Implementation

Overview: The OPC+ Cloud Service is a general-purpose OPC platform designed with one goal: making advanced mask correction accessible to everyone. With 'easy access, easy use' as our core philosophy, the platform leverages GPU acceleration and cloud-native architecture to deliver a streamlined freeform mask correction workflow.

Key Features:

• Easy Access: Cloud-based platform accessible from anywhere, eliminating the need for expensive local infrastructure.
• Easy Use: Intuitive 'OPC for Everyone' interface - no scripting or deep technical knowledge required. Simply upload your layout and configure parameters.
• General-Purpose Freeform OPC: Native support for curvilinear masks, essential for photonic integrated circuits and advanced semiconductor designs.
• GPU-Accelerated Performance: CUDA Multi-Stream technology delivers significant speedup compared to traditional CPU-based methods.
• Cloud-Native Architecture: Kubernetes-based elastic scaling ensures optimal resource utilization and predictable performance.
• Industry Standards: Seamless integration with existing workflows through GDS/OASIS format support.

Platform Goals:

• Democratize access to advanced OPC technology
• Remove complexity barriers in mask correction
• Support both semiconductor and photonic manufacturing

Quantum Annealing for Lithography

Overview: Optimization problems in lithography (like SMO and ILT) are NP-hard and computationally expensive. We explore Quantum Annealing (QA) as a specialized solver to find global minima more efficiently than classical simulated annealing.

Research Highlights:

• Ising Model Formulation: Mapping the lithography mask optimization problem onto the Ising model suitable for D-Wave quantum annealers.
• Hybrid Algorithms: Combining classical gradient descent with quantum sampling to overcome the qubit connectivity limitations of current hardware.
• Benchmarking: Demonstrating significant speedup in convergence time for complex 2D mask patterns compared to traditional CPU/GPU clusters.