Jian-Wen Dong

On‐chip lasers that emit wavelength‐multiplexed structured light are important in integrated photonics. Bound states in the continuum (BICs) provide high‐quality resonances with radiation topology, enabling vortex lasing at multiple wavelengths. However, the topological charges of the symmetry‐protected BICs at the Γ‐point are constrained by lattice symmetry, whereas creating Dirac‐point degeneracies between photonic bands can redistribute polarization singularities. The connection between momentum‐space topology reconstruction and topological charge exchange of dual‐wavelength vortex emission remains underexplored. Here, we demonstrate a dual‐wavelength BIC vortex microlaser in a Lieb‐lattice photonic crystal. Numerical calculations reveal that, during continuous structural parameter tuning, the topological reconstruction is accompanied by the emergence of Dirac points, while the topological charges associated with the Γ‐point BIC modes are exchanged. Experimentally, room‐temperature dual‐wavelength lasing operation is realized, with far‐field polarization and self‐interference measurements revealing the transformation of topological singularities between two BIC modes. This work connects momentum‐space BIC topology reconstruction with coherent vortex lasing, offering an active route toward compact and tunable structured‐light sources.

Integral imaging (InIm) enables 3D displays with continuous parallax and authentic depth cues, representing a highly promising technical solution for true 3D extended reality (XR) displays. However, it suffers from a fundamental trade-off between spatial resolution and field of view (FOV). This study verifies that a heterogeneous metalens array (MeLA)–based InIm display system, while achieving a wide FOV, possesses inherent subpixel shifts among its elemental images, which hold the potential to recover information beyond the system Nyquist limit. We propose an adaptive super-resolution (SR) method based on local uniform-pattern matching, which delivers up to a two-fold resolution improvement in such a display system without additional dynamic hardware and with no degradation to the FOV. Numerical and optical simulations performed on a system equipped with a 4×16 MeLA and a high-resolution display panel demonstrate that this scheme achieves a wide FOV of 50°, while the peak resolution is boosted from 21 pixels per degree (PPD) to 42 PPD. This method maintains stable resolution enhancement across the full FOV and the entire depth of field (DOF), offering a viable technical pathway for static XR 3D display systems featuring high resolution and a wide FOV.

Photonic crystals are typically fabricated with a fixed topology, freezing their edge and corner states in space and frequency. Here we introduce a kirigami-actuated reconfigurable topological photonic crystal in which a single in-plane rotation reshapes the unit cell, drives a bandgap closure and reopening, and thereby relocates the Wannier center and switches the associated Zak phase—the topological invariant that determines whether boundary states appear. We engineer a symmetry-preserving square-lattice deformation path that programs the intracell rod configuration into a compact, discrete topological phase library. Using this library, we demonstrate reconfigurable control of topological modes in both resonance frequency and spatial localization from the same mechanical degree of freedom. A higher-order corner mode is switched between two distinct resonance frequencies while its localization moves between different geometric corners. Meanwhile, at a fixed operating frequency, the propagation path of reciprocal crystalline-symmetry-protected edge modes is mechanically rewritten, enabling on/off gating and two-path beam routing. These results establish a route toward reconfigurable topological photonic devices, showing that quantized band topology can be encoded and switched by shape.

Extended reality with an immersive visual experience needs to achieve vergence-accommodation-conflict-free (VAC-free) and human-level field of view (FOV) simultaneously. Holography is a dreamlike solution but seriously suffers from both an ultra-narrow diffraction angle and uneliminated coherent laser speckle. Incoherent-based light-field display is an alternative solution, but it still has an inherent issue with off-axis aberration when using a homogeneous microlens array for lightweight applications. Here, we propose a heterogeneous metalens array (MeLA) that integrates VAC-free operation and aberration correction into a slim monolithic component with expanded FOV. Such a large etendue is due to a unique phase profile in the heterogeneous MeLA, each lenslet of which has an identical hyperbolic term cascading a varying linear-tiling term. Nanoimprint lithography is employed to achieve a centimeter-scale MeLA with a feature aspect ratio of 7:1. In this way, we developed a light-field 3D display prototype with monocular focus cues and a FOV of 50°, approximately four times that of the microlens-array case, and fully exploiting the commercial micro-displays. The proposed MeLA system can even be scaled to human stereoscopic vision, with a maximum value of 80°, without compromising other performance metrics such as depth of field, resolution, and compactness. The characteristics of FOV expansion with monocular 3D display capability are demonstrated through experimental videos generated using a real-time elemental image array algorithm dedicated to the proposed MeLA system. This work opens new opportunities for extended reality through VAC-free 3D display and a wide FOV, poised to transform consumer electronics, precision nanofabrication, and healthcare.

Light field displays resolve the vergence-accommodation conflict in near-eye systems. Metalens arrays offer an ultra-thin alternative to refractive microlens arrays, but their diffractive nature causes severe chromatic aberrations under full-color illumination, leading to color deviation and spatial blurring. Correcting these aberrations through meta-atom design increases fabrication complexity. In this paper, a computational pre-compensation scheme is proposed, with a closed-loop training framework consisting of a light field display surrogate network (LFDSNet) and a pre-compensation network (PreCompNet). To drive this framework, a cross-scale ray tracing model based on the windowed Fourier transform is developed to simulate the meta-optics system. Validated through simulations of a designed all-dielectric metalens array, the proposed scheme achieves a full-color 3D display with natural depth-of-field effects, with an improval of the mean PSNR from 13.7 to 22.7 dB, the mean SSIM from 0.600 to 0.842, as well as a reduction of the mean SAM from 0.338 to 0.061 rad. These results may serve as a generalizable paradigm for pre-compensation in metalens-based AR/VR display systems.

Synthetic dimensions have emerged as a powerful framework for exploring topological phenomena beyond real space. Here, we investigate the evolution of surface modes in three-dimensional periodic crystals by introducing a synthetic translation dimension associated with unit-cell translation. By combining the surface-parallel Bloch wave vector with the cyclic translation parameter, we construct a minimal two-dimensional synthetic parameter space that fully captures boundary spectral evolution. Remarkably, this surface-mode evolution persists even when multiple bulk bands are involved, extending beyond the single-band scenarios. We demonstrate that the existence and spectral flow of surface modes are independent of microscopic details and rely only on the topology of bulk bands. Our results are validated in both three-dimensional acoustic and photonic crystal platforms, underscoring the universality of the mechanism across classical-wave systems. These findings establish synthetic translation dimensions as a general framework for predicting and controlling surface-mode evolution in three-dimensional periodic media.

Topological corner states with wavelength‐selective excitation and directional routing enable on‐chip wavelength‐dependent beam splitters (WDBSs) for multiplexing technology. In this work, we employ two types of topological corner states to implement WDBS in kagome lattice photonic crystals. A tight‐binding model incorporating both nearest‐neighbor and next‐nearest‐neighbor couplings is applied to describe the bulk and corner states in the second‐order topological photonic crystal. A 120‐degree‐turning interface is constructed that supports two types of corner states, where the type‐I corner state is even mode originating from nearest‐neighbor coupling and the type‐II state is odd mode originating from next‐nearest‐neighbor coupling. The spatial Fourier transformation of eigenstates indicates that the type‐I mode supports propagating waves along 30‐degree channel and is inhibited along the 90‐degree channel, and vice versa for type II. With different resonant wavelengths, the two types of corner states have direction‐wavelength locking mechanisms, enabling us to propose a 1 × 2 channel WDBS. Such WDBS design has achieved a low‐crosstalk multiplexing based on the directionality of corner states. This finding provides guidance for multiplexing different types of topological photonic states, which is conducive to promoting the design of on‐chip wavelength/mode‐division multiplexers with topological protection.

Pseudo-spin and valley are two significant degrees of freedom in topological photonics, which enable the realization of robust boundary states and consequently tremendous fabrication-tolerant photonic devices. As integration density increases in photonic chips, low-crosstalk waveguide crossings become critical for maintaining overall performance. With conventional waveguide crossings susceptible to defect-induced backscattering, topology can provide a viable solution to this issue. Here, we experimentally demonstrate a topological waveguide crossing based on pseudo-spin-valley-locked domain-wall states. By combining two unidirectional spin-polarized domain-wall channels supporting states with distinct valley degrees of freedom, our system can intrinsically suppress both backscattering and interchannel crosstalk (<−15 dB in simulation, <−6 dB in experiment). Moreover, the robustness of our topological waveguide crossing against a variety of structural defects is also confirmed via numerical simulations. Our results not only establish a robust strategy for waveguide crossing but also highlight that the interplay between pseudo-spin and valley degrees of freedom can offer expanded opportunities for functional photonic device design.