Bright squeezed light from parametric down-conversion in the infrared (IR) frequency range has triggered the emergence of attosecond quantum optics, a new research field at the interface of quantum optics, strong-field physics, and attosecond technology. Two challenges arise at this interface: transferring quantum features of the IR light sources to the ultraviolet (UV) and extreme ultraviolet (XUV) frequency range via strong-field nonlinearities and exploiting quantum optical properties of the nonlinear optical response as a new probe in ultrafast dynamics. Here, we address both by driving high-harmonic generation (HHG) in solids with entangled photon pairs either in degenerate or nondegenerate frequency modes. In the degenerate mode, single-shot measurements of harmonics up to the 10th order reveal strong photon bunching whose g(2) first grows and then decreases with the harmonic order. We show that this behavior tracks different microscopic mechanisms responsible for harmonic emission, demonstrating the potential of ultrafast quantum optical spectroscopy. In the nondegenerate case, the harmonics retain quantum-induced correlations, verified by wavelength-resolved second-order cross-correlation maps. Our findings demonstrate the transfer of quantum photon correlations from bright squeezed vacuum into HHG and open a pathway toward quantum-enhanced ultrafast spectroscopy and control of ultrafast dynamics in solids.
Photonic parallel computing inspires important approaches to artificial intelligence, promising substantial gains in processing speed, energy efficiency, and latency. Nevertheless, the integration density and robustness of most optical computing architectures continue to pose significant challenges. Here, we report a waveguide-array optical neural network based on thin-film lithium niobate (TFLN). By virtue of the compact size of the waveguide arrays, this architecture yields a computational density of 9.48×104 TOPS/mm2. Moreover, the high-speed and energy-efficient characteristics of the TFLN platform enable a computing speed of 135.6 TOPS with an energy efficiency of 6.28 TOPS/W. Our experiments on the iris flower and handwritten digit recognition demonstrate accuracies of 100% and 89.7%, respectively. The device achieves over 90% accuracy across a 75-nm operating bandwidth in the iris flower recognition task, indicating the wavelength insensitivity and superior robustness of the waveguide array. As a compact paradigm for optical neural networks, we demonstrate the potential of waveguide arrays in performing optical computation, thereby offering solutions for the advancement of edge computing, optoelectronic integration, and intelligent automation.
Quantum mechanics predicts that measurement is inherently invasive, challenging the macrorealist view that physical properties exist independently of observation. The Leggett–Garg inequality (LGI) provides a test of this worldview, but its violation in systems in collective systems is often challenging due to technical measurement disturbance. By embedding an Elitzur–Vaidman interaction-free measurement (“quantum bomb-testing”) protocol into a microwave-driven Rydberg superatom (a mesoscopically coherent collective system) interferometer—we realize a high-fidelity, sequential measurement test. We observe a clear LGI violation that exceeds the classical bound by nearly 50 standard deviations in such a mesoscopic collective system. This result demonstrates the failure of noninvasive measurability for a collective quantum observable, establishing a scalable platform for probing temporal quantum correlations beyond the single-particle limit.
Janus metasurfaces have attracted considerable attention in encrypted communication, imaging, and display due to their unusual bidirectional asymmetric optical manipulation characteristics. Particularly, their multifunctionalization is of great significance for enhancing the compactness and integration of optical systems. However, the realization of multifunctional Janus metasurfaces in the optical band still faces enormous difficulties and challenges, which are essentially limited by the anti-error design and processing methods of multi-layer cascade metasurfaces. Here, a triple-layer Janus metasurface based on cascaded plasmonic nano-antenna arrays fully buried in SiO2 is constructed. Among them, the cascaded plasma nano-antenna consists of two types of enantiomers for each wavelength. Each enantiomer is fabricated with two layers of L-shaped nanostructures with phase modulation and one layer of dimer nanostructures with polarization selection stacked alternately along the optical axis. We experimentally show that the Janus metasurface achieves bidirectional asymmetric multi-channel holographic encryption, simultaneously using propagation direction, phase, polarization, and wavelength for the first time in the near-infrared band, which is expected to provide a frontier route for multifunctional optical displays, high-level optical information encryption, and large-capacity full-duplex communication.
Monolithic integration of electronic functionality into insulating ferroelectric photonic platforms remains a critical challenge. We introduce a paradigm in which ferroelectric domain walls (DWs) are engineered as active, programmable nanocircuits rather than passive boundaries. Using a single-step nanofabrication process, we directly encode a conductive DW network into thin-film lithium tantalate (TFLT), spatially coincident with a co-designed optical metasurface. This tailored superlattice enables quantum-tunneling-governed transport with a record-low resistivity of around 2.6×10−4 Ω·m. The synergy between the resonant metasurface and the embedded nanocircuit yields a high-performance, self-powered photodetector at the design resonant wavelength (785 nm), exhibiting a responsivity of 0.84 A/W and a response time as small as 160 μs. We transform domain engineering from a structural tool into a functional design platform, embedding tailored optoelectronic circuitry into passive photonics for next-generation multifunctional systems.
