Infrared light, although crucial to numerous applications such as night vision, medical diagnostics, and industrial monitoring, remains largely inaccessible to conventional silicon-based imaging sensors. Traditional infrared detectors rely heavily on complex, costly, and cryogenically cooled semiconductor materials, which limit their deployment in portable or consumer devices. An alternative strategy has been to “upconvert” infrared photons into visible wavelengths, thus enabling the use of standard silicon cameras known for their sensitivity and compactness. Nevertheless, achieving efficient nonlinear optical conversion in miniaturized, chip-scale platforms has proven a formidable challenge—until now.

The research team led by Professor Tingting Liu has developed an ultra-thin silicon chip patterned with a meticulous array of nanoscale silicon disks, forming a so-called metasurface. This metasurface exploits a unique optical resonance phenomenon to trap and intensify incoming infrared light within nanoscale volumes. By amplifying the local electromagnetic field, the device dramatically boosts silicon’s intrinsic third-order nonlinear optical response, enabling effective third-harmonic generation (THG)—a frequency tripling process that converts IR light into bright visible green light.

At the heart of this achievement is the mastery of a sophisticated optical mode known as a quasi-bound state in the continuum (quasi-BIC). Bound states in the continuum are modes that remain confined without radiating energy despite existing within the spectrum of free-space waves. By introducing a delicate spatial asymmetry to the silicon nanodisks, the team transformed a perfectly trapped, non-radiative mode into a leaky, yet high-quality (high-Q) resonance. This quasi-BIC resonance exhibits an impressive quality factor on the order of 4000, indicating exceptionally low energy loss and prolonged photon confinement within the structure.

The enhanced light-matter interaction facilitated by this quasi-BIC resonance leads to a third-harmonic generation efficiency of approximately 3×10^-5—an unprecedented performance for CMOS-compatible silicon metasurfaces. This breakthrough marks a transformative advancement over previous nonlinear silicon devices, which suffered from limited conversion efficiencies and significant fabrication challenges. Importantly, the use of silicon, a widely available and CMOS-friendly material, ensures compatibility with established semiconductor manufacturing processes, paving the way for scalable and cost-effective production of these metasurface chips.

Beyond raw efficiency, the metasurface device demonstrates its true versatility through direct infrared imaging capabilities. Acting as a dense parallel array of nanoscale converters, the chip can faithfully translate intricate IR images projected onto it into corresponding visible images. The research team has showcased high-fidelity upconversion imaging of standard resolution targets, such as the Siemens star pattern, as well as custom 3D-printed test objects. Optical resolution reaches finely detailed imaging at approximately 6 micrometers spatial scale, all achieved with a single continuous-wave IR pump laser under ambient room temperature conditions.

This advance addresses the critical bottleneck that has long impeded the practical adoption of nonlinear metasurface-based upconversion technology: balancing compactness with conversion efficiency and operational simplicity. By circumventing the bulky, alignment-sensitive nonlinear crystals traditionally used, the silicon metasurface platform offers an elegant and integrable solution, ideal for next-generation infrared sensing and imaging. The device’s room-temperature operation negates the need for complex cooling systems, substantially reducing power consumption and system complexity.

The implications of this work extend broadly across sectors reliant on infrared detection and imaging. Security and surveillance systems stand to gain from enhanced night-vision capabilities embedded into compact, robust hardware. Industrial automation and quality control processes can benefit from high-resolution IR imaging with straightforward silicon sensor integration, enabling real-time defect detection and process monitoring. In consumer electronics, the potential for embedding efficient IR cameras into smartphones and wearable devices becomes increasingly tangible.

Scientifically, this study underscores the power of tailoring light confinement and resonance properties at the nanoscale to elevate nonlinear optical phenomena. The strategic employment of quasi-BIC modes to create ultra-high-Q resonances within accessible materials like silicon highlights a promising paradigm for enhancing other nonlinear optical processes, such as frequency conversion, all-optical switching, and quantum photonics. This could stimulate future research into metasurface-enabled photonic devices with multifunctional capabilities embedded within ultra-compact footprints.

Professor Tingting Liu’s interdisciplinary expertise in micro/nano-photonics and signal processing has been instrumental in realizing this milestone. Under her leadership, the team has combined rigorous theoretical modeling with cutting-edge nanofabrication techniques to optimize the metasurface architecture. The result is a scalable and reproducible platform firmly grounded in CMOS technology while circumventing the limitations imposed by traditional infrared detector materials.

As the field of photonics increasingly gravitates towards integrated and miniaturized solutions, this work sets a new standard for how nonlinear optical metasurfaces can be harnessed for practical applications. The synergy between high-quality resonances, material engineering, and nanoscale device fabrication demonstrated here suggests a bright future for compact infrared imaging technologies that are accessible, efficient, and versatile.

In conclusion, the ultra-thin silicon metasurface developed by Nanchang University’s team represents a transformative advance in infrared photonics. By achieving record third-harmonic generation efficiency through high-Q quasi-BIC resonances, the device opens up new avenues for low-cost, high-performance infrared imaging at room temperature. Its compatibility with existing silicon technology and demonstrated imaging capabilities herald a new era in nonlinear optical devices poised to impact diverse areas including defense, healthcare, industry, and consumer electronics. This breakthrough underscores the growing potential of metasurfaces to revolutionize how light manipulation enables next-generation sensing and imaging systems.

Subject of Research: Infrared upconversion imaging using nonlinear silicon metasurfaces empowered by quasi-bound states in the continuum (quasi-BIC).

Article Title: High-efficiency infrared upconversion imaging with nonlinear silicon metasurfaces empowered by quasi-bound states in the continuum.

News Publication Date: Not explicitly stated; article DOI suggests 2026.

Web References:
– DOI: http://dx.doi.org/10.29026/oea.2026.250257

Image Credits: OEA (Opto-Electronic Advances)

Traditional infrared imaging techniques have long faced limitations due to the intrinsic challenges of detecting and processing light at longer wavelengths. Infrared detectors generally suffer from lower resolution and higher noise compared to their visible-wavelength counterparts, making it difficult to extract detailed phase and amplitude information. The innovative upconversion entropy encoding method meticulously overcomes these hurdles by ingeniously transforming infrared light signals into visible light, where sophisticated and mature imaging technologies can be leveraged. This spectral translation not only enhances detection efficiency but also opens pathways to richer information content through complex-amplitude retrieval.

At the heart of this novel technique lies the concept of optical entropy encoding. Unlike conventional imaging methods that rely solely on intensity measurements, entropy encoding incorporates the spatial complexity and randomness inherent in optical wavefronts, allowing the capture of both amplitude and phase data with high fidelity. By applying advanced mathematical frameworks rooted in information theory, the team was able to develop an encoding protocol that effectively modulates the infrared wavefront’s entropy, embedding complex structural information within the upconverted visible light signal.

This methodological leap is facilitated through nonlinear optical processes, where the incident infrared photons interact within specially designed upconversion materials, generating photons at visible wavelengths. The precise control over this interaction enables the preservation of the complex-amplitude characteristics of the original infrared field during the wavelength conversion. As a result, the encoded visible light carries comprehensive optical data that can be decoded using tailored phase retrieval algorithms, reconstructing high-resolution images with quantitative phase information.

The implications of this breakthrough extend far beyond mere imaging clarity. Complex-amplitude imaging in the infrared spectrum is critical for numerous scientific investigations and technical applications where phase information reveals subtle variations in material properties, surface profiles, and biological tissues. For example, in biomedical optics, accessing complex-amplitude data in the infrared window facilitates non-invasive diagnostics of cellular structures beneath scattering layers, potentially revolutionizing early disease detection.

Moreover, this technique offers compelling advantages in remote sensing and environmental monitoring. Infrared complex-amplitude imaging can discern chemical compositions, temperature gradients, and moisture content with enhanced precision, providing more reliable data for climate modeling, agricultural management, and pollution tracking. Its integration with entropy-based data encoding also optimizes the information capacity and security of optical communication systems operating in challenging atmospheric conditions.

One of the remarkable aspects of this research is the harmonization between experimental optics and computational intelligence. The decoding process leverages sophisticated algorithms that interpret the entropy-encoded information, reconstructing complex-amplitude maps at superior resolutions unattainable by conventional detectors alone. This synergy amplifies the system’s adaptability across various imaging scenarios and forms a foundational strategy for next-generation optical sensors.

The team’s experimental demonstrations showcased the capability of their system to capture and reconstruct intricate complex-amplitude patterns with high sensitivity and spatial resolution. By employing a controlled test setup involving calibrated infrared sources and custom nonlinear materials, the researchers validated the robustness and accuracy of their approach. The comprehensive characterization included assessments of phase stability, signal-to-noise ratios, and resilience against environmental perturbations, cementing its applicability in real-world conditions.

From an engineering perspective, the upconversion optical entropy encoding framework embodies a scalable, integrable platform. The materials and optical components are compatible with existing photonic architectures, facilitating seamless integration into miniaturized devices suitable for portable spectroscopy, security scanners, and autonomous sensing instruments. This versatility promises accelerated adoption and innovation within diverse tech ecosystems.

The research also navigates the broader theoretical context of optical entropy and information encoding, providing insights into novel ways of structuring photonic data streams. By quantifying and manipulating entropy in complex fields, the study bridges physics, information theory, and materials science, stimulating interdisciplinary pursuits focused on optimizing optical information throughput and fidelity.

As the frontier of infrared imaging pushes forward, the contributions of Zhu, Pan, Tang, and colleagues mark a paradigm shift. Their technique not only overcomes persistent technical challenges but also enriches our conceptualization of how light-matter interactions can be harnessed for information-rich imaging. The transition from simply detecting photons to decoding their embedded entropy fundamentally transforms infrared optical measurement.

Looking ahead, this innovation is poised to unlock new capabilities in quantum imaging, adaptive optics, and multispectral sensing. The entropy-encoded upconversion strategy may further catalyze developments in encrypted optical communications, ultrafast imaging, and even astrophysical observations, where precious infrared signals need precise, high-information-content retrieval under noisy conditions.

The study’s pioneering integration of nonlinear photonics, entropy-based modulation, and computational decoding establishes a powerful methodological blueprint. It invites the optics community to rethink traditional paradigms and explore complex-amplitude imaging as a vital avenue for future research and technology development.

In sum, the discovery embodies the convergence of fundamental science and practical innovation. It exemplifies how nuanced manipulation of optical entropy can transcend conventional imaging limits, redefining infrared complex-amplitude acquisition. This milestone underscores the potential of interdisciplinary research to unlock transformative tools for understanding and interacting with the invisible domains of light.

Subject of Research: Infrared complex-amplitude imaging enabled by upconversion optical entropy encoding

Article Title: Upconversion optical entropy encoding for infrared complex-amplitude imaging

Article References:
Zhu, Sk., Pan, T., Tang, Cx. et al. Upconversion optical entropy encoding for infrared complex-amplitude imaging. Light Sci Appl 15, 158 (2026). https://doi.org/10.1038/s41377-026-02215-7

Image Credits: AI Generated

DOI: 10.1038/s41377-026-02215-7

Keywords: Upconversion, optical entropy encoding, infrared imaging, complex-amplitude retrieval, nonlinear optics, phase imaging, information theory