2025 Theses Doctoral

Visible Dielectric Metasurfaces for Scalable Quantum Control and Intelligent Sensing

Xu, Yuan

This dissertation traces a path from the physics of flat, subwavelength dielectric structures to wafer-scale systems that control and sense light with high efficiency and fidelity. The premise is that visible dielectric metasurfaces can replace stacks of bulk optics with lithographically defined elements that shape wavefronts, strengthen light–matter interactions, and encode information for downstream electronics or learning pipelines.

Across four studies—single-atom trapping, scalable quantum control, nonlinear IR-to-visible up-conversion, and metric depth sensing—the work shows how meta-optics moves advanced functions toward foundry-style manufacturing without sacrificing performance. Conventional hardware for atom arrays, nonlinear conversion, and depth perception depends on bulky relays, narrow-acceptance crystals, or multi-lens baselines that hinder integration and scale-up.

Metasurfaces write the required wavefronts directly at the device plane with subwavelength pixels. Two capabilities underpin the results: nonlocal resonances, especially quasi–bound states in the continuum, which provide large, tunable quality factors while preserving Bloch-mode momentum; and CMOS-compatible materials and process flows (TiO₂, silicon-rich SiN, GaN/InGaN; selective-area growth; removable masks) that support wafer-level replication and co-packaging.

The studies demonstrate: high-NA, phase-only metasurfaces that generate and focus tweezer arrays for single-atom trapping; a quantitative framework for scalable tweezer array generation, linking numerical aperture, field of view, array spacing, and pixel budget to array uniformity; a nonlinear GaN/InGaN metasurface platform for wavefront-preserving IR-to-visible up-conversion; and a compact 3D metric depth sensor consisting of a birefringent metasurface that encodes metric depth in orthogonally polarized image pairs and a lightweight neural network depth decoder.

Collectively, the contributions include device-and-system demonstrations that connect metasurface physics to atom-quality trapping, wafer-scale array control, wavefront-faithful up-conversion, and single-shot metric depth sensing; design rules and analytic mappings that make scaling predictable; and fabrication and integration flows aligned with foundry practice. The same principles extend to multi-band perception, on-chip wavefront correction for nonlinear modules, and hybrid PIC–metasurface architectures. By uniting nonlocal resonances, visible-band materials, and manufacturing-ready processes, metasurfaces emerge as a credible substrate for scalable quantum control and intelligent sensing.