Metasurfaces: Two-Dimensional Marvels
In the quest for overcoming the limitations of bulk metamaterials, researchers turned to metasurfaces – two-dimensional counterparts offering compactness, flexibility, and ease of integration. Composed of arrays of subwavelength nanostructures, metasurfaces enable precise control over light waves with unprecedented efficiency. These thin films serve as versatile platforms for encoding complex optical functionalities, including beam steering, focusing, and polarization manipulation.
The beauty of metasurfaces lies in their simplicity and versatility. By tailoring the arrangement of nanoantennas, metasurfaces can modulate the phase and amplitude of light waves with remarkable precision. This capability enables a myriad of applications, ranging from high-resolution imaging to optical communication and sensing.
The main advantages of metasurfaces with respect to the existing conventional technology include their low cost, low level of absorption in comparison with bulky metamaterials, and easy integration due to their thin profile. Part of the promise of these devices lies in their ability to perform these complex optical functions using metasurfaces that are manufactured using standard lithographic techniques common in the semiconductor industry.
Among the most promising applications of metamaterials are optical metasurfaces, which are rapidly redefining the way light is controlled. These ultrathin, nanostructured materials are being used to revolutionize key technologies like optical modulators, beam steering systems for LiDAR, radiation management devices, and even optical chips designed for AI inference in data centers. The precision and scalability of optical metasurfaces enable the creation of compact, high-performance systems that can outperform traditional optical devices in both size and functionality. These breakthroughs are ushering in a new era of communication technologies, advanced data processing, and autonomous systems, unlocking previously unattainable levels of efficiency and capability.
Metasurfaces: The Core of Meta-Optics
Metasurfaces can be classified as either active or passive. Passive metasurfaces offer fixed functionalities, acting as static elements in optical systems. In contrast, active metasurfaces represent a significant leap forward by enabling tunable, reconfigurable, and time-varying functionalities. This critical innovation allows the integration of multiple optical functions into a single layer, facilitating precise control over light’s polarization and trajectory. As a result, conventional bulky optical systems can now be transformed into exceedingly compact formats, unlocking new properties and applications that were once beyond the reach of traditional optics.
Despite their groundbreaking capabilities, metamaterials and metasurfaces currently operate effectively over narrow spectral bands. While progress is being made toward achieving broadband performance, significant challenges remain in terms of efficiency, scalability, and manufacturing cost. However, even in their narrowband applications, metasurfaces are already proving transformative in fields such as LiDAR for autonomous vehicles, 3D sensing, and biometrics.
New Material Platforms for Metasurfaces
The development of next-generation metasurfaces is increasingly driven by breakthroughs in advanced materials that support tunability, low optical losses, and CMOS compatibility. Transition-metal nitrides like titanium nitride (TiN) combine plasmonic behavior similar to gold with high thermal stability, making them ideal for high-intensity photonic applications. Transparent conducting oxides (TCOs), such as indium tin oxide, offer precise spectral control through epsilon-near-zero properties, enabling enhanced light manipulation for nonlinear and active optical devices.
Equally compelling are phase-change materials like chalcogenide alloys and perovskite nickelates (e.g., SmNiO₃), which switch between optical states using electrical or optical stimuli. These materials promise dynamic, non-volatile metasurface functionalities for display technologies and optical switches. Their broad spectral tunability—from visible to mid-IR—ushers in the potential for novel photonic systems based on strongly correlated electronic behavior, enabling a new class of active metasurfaces.
Fundamental Advances in Metasurface Design
A 2019 breakthrough by University of Delaware researchers introduced a compact and low-loss integrated photonics platform utilizing metasurface-enabled metalenses. Their design, based on high-contrast transmitarray (HCTA) technology, achieved a high numerical aperture (up to 2.14) and less than 1 dB insertion loss, allowing sub-10 µm light focusing and enabling key computational operations like Fourier transforms and spatial differentiation on-chip.
The design of metasurfaces has undergone a renaissance with the emergence of bilayer architectures. Researchers at Harvard’s School of Engineering and Applied Sciences (SEAS) have pushed past the limitations of single-layer designs with a titanium dioxide-based bilayer structure, which resembles stepped skyscrapers when viewed under high-resolution microscopy. This configuration addresses longstanding polarization constraints by allowing independent manipulation of wavelength, phase, and polarization. Manufactured through a CMOS-compatible process, the freestanding dual layers maintain chemical stability, opening the door to a range of applications, from multidirectional displays that project different images on opposite surfaces to broadband operation that spans the visible to near-infrared spectrum. Moreover, these advances pave the way for generating entangled photon states in quantum systems through gradient metasurfaces.
Hybrid Photonic Integrated Chip for Multifunctional Optical Systems
At Penn State, researchers led by Prof. Xingjie Ni developed a photonic integrated chip that unites the best of metasurfaces and photonic integrated circuits (PICs). This hybrid chip architecture allows guided light within PICs to dynamically control multiple metasurfaces, vastly improving light routing and enabling the execution of several optical functions on a single, compact chip.
The modularity of the platform offers a blueprint for creating a library of functional components usable across optical communication systems, LiDAR, VR/AR displays, and free-space optical interconnects. This integration of metasurfaces with PICs paves the way for multifunctional, reconfigurable photonic systems. Supported by funding from NASA, the Moore Foundation, and others, the work represents a scalable path toward next-generation photonic hardware with real-world applicability.
Optical AI Inference Chips: Accelerating Intelligent Computation
As artificial intelligence permeates every layer of modern computing, from autonomous systems to large-scale data centers, the need for energy-efficient and high-throughput hardware is growing rapidly. Optical metasurfaces are playing a pivotal role in this transformation by enabling AI inference chips that process data at the speed of light. These chips utilize the ability of metasurfaces to manipulate light at subwavelength scales to perform complex matrix operations essential for neural network computations. Unlike traditional electronic processors, these optical chips can handle massive parallel processing tasks while consuming significantly less energy and producing less heat.
Alongside hardware innovations, machine learning has begun to revolutionize the way metasurfaces are designed. Traditional inverse design approaches often struggled with the high-dimensional parameter spaces of metastructures. However, artificial intelligence now offers powerful optimization tools that can tune geometries for enhanced nonlinear optical effects, such as harmonic generation in gallium phosphide, or adapt resonance modes for high-sensitivity optomechanical sensors. These algorithms can even predict structures with minimal scattering and absorption losses, streamlining the development of metasurfaces for a wide array of performance-critical applications.
By leveraging machine learning, researchers can efficiently navigate the complex parameter space inherent in optimizing metasurfaces and metamaterials, ultimately achieving superior performance beyond the constraints of traditional optimization approaches. This novel methodology promises to accelerate advancements in photonics, paving the way for transformative innovations in various fields reliant on precise light-matter interactions.
Applications and Future Prospects
The applications of metamaterials and metasurfaces span a vast array of fields, promising transformative impacts on technology and society. From compact lenses with minimal aberrations to ultra-thin optical components for augmented reality devices, the potential uses are limitless. Metasurfaces, in particular, offer a pathway to lightweight, low-cost optical devices with unparalleled performance and integration capabilities.
Computing and Data Infrastructure Applications
The rise of metasurfaces is now reshaping data processing architectures, particularly in the era of artificial intelligence. A striking example comes from Neurophos, which has introduced a tensor core processor built on a metamaterial platform. Integrating silicon photonics with metasurface-based compute-in-memory architecture, the chip represents a fundamental shift away from traditional GPU constraints. This photonic processor offers a 1,000-fold reduction in size compared to conventional foundry-developed solutions while achieving throughput levels exceeding 1 exa-operation per second. The result is a highly energy-efficient platform, consuming 40% less power than NVIDIA’s flagship H100 GPUs when running transformer-based AI models.
This wave of innovation extends to neuromorphic computing platforms that blend photonic integrated circuits with metasurfaces. At the 2025 SPIE Photonic Computing Conference, researchers showcased advances including electro-optic microring neurons capable of operating at RF clock rates up to 12.5 GHz. Coherent Ising machines, which use degenerate optical parametric oscillators, are tackling combinatorial optimization problems with unmatched speed. Meanwhile, diffractive optical processors are demonstrating the ability to perform real-time Fourier transforms in sub-nanosecond latencies, establishing metasurfaces as foundational building blocks of next-generation computing.
Sensing, Imaging, and Automotive Technologies
The sensing landscape has also been transformed by metasurface-based innovations. Solid-state LiDAR systems, crucial for autonomous vehicles, are undergoing a revolution with the introduction of liquid crystal metasurfaces (LCMs).
Beam Steering LIDAR: Enhancing Sensing Capabilities
LIDAR (Light Detection and Ranging) is a cornerstone technology in applications ranging from autonomous vehicles and robotics to remote sensing and augmented reality. Optical metasurfaces are transforming LIDAR by enabling dynamic, precise, and fast beam steering without relying on bulky mechanical components. These metasurfaces control the phase and direction of incoming light at the nanoscale, allowing systems to scan environments in real time with high spatial resolution. The ability to perform rapid and accurate 3D mapping makes metasurface-enhanced LIDAR systems highly desirable for improving safety, navigation, and situational awareness in both commercial and defense sectors.
An important breakthrough in this field was reported in August 2021 by researchers at the Harvard John A. Paulson School of Engineering and Applied Sciences. They developed a metasurface-based laser system capable of manipulating various properties of laser light, such as shape, intensity, and wavelength, without requiring additional optical elements. Unlike traditional designs that use single nanopillars, their innovation involved “supercells”—groups of pillars engineered to collectively tune multiple aspects of the emitted light. This metasurface forms part of a laser cavity with a diode, splitting the beam and enabling precise, tunable laser output. Such advancements not only simplify and miniaturize optical setups but also open up exciting new applications in areas like quantum sensing, augmented and virtual reality, and advanced biomedical imaging.
A partnership between Lumotive and Himax has yielded beam-steering devices with 25×25 mm apertures fabricated at lithography-reticle scale. These systems deliver a 120-degree field of view with random-access beam steering speeds under 25 microseconds and have passed rigorous mechanical shock tests up to 15G—making them viable for deployment in rugged automotive environments.
Radiation Control and Wireless Communication: Nanoscale Solutions for Global Challenges
In the medical field, metasurfaces are contributing to a new era of lightweight and precise imaging technologies. At the University of Utah, engineers have developed multilevel diffractive lenses (MDLs) using polymer materials that drastically reduce the weight of thermal cameras—by as much as 200 times. This innovation addresses critical ergonomic challenges in military gear, particularly neck strain from night-vision headsets. In parallel, graphene oxide microlenses are enabling ultra-compact endoscopes with sub-diffraction imaging resolution, enhancing early detection of diseases like cancer through minimally invasive procedures.
In environments where managing electromagnetic radiation is critical—such as medical facilities, space systems, and nuclear power plants—metasurfaces offer revolutionary control over radiation behavior. By designing these nanostructures with precise geometries and material properties, scientists can manipulate absorption, reflection, and emission characteristics across a range of wavelengths. Radiative cooling metasurfaces, inspired by biological adaptations in extreme environments, exemplify this potential. These materials passively cool objects by radiating heat into space without consuming energy, offering game-changing possibilities for thermal management in electronics, energy-efficient buildings, and extreme-climate habitats.
Beyond thermal control, optical metasurfaces are redefining wireless communication by enabling new paradigms in signal modulation, beam shaping, and antenna design. As the world moves toward increasingly dense and fast wireless networks—such as 5G, 6G, and the Internet of Things (IoT)—traditional components are becoming bottlenecks. Metasurface-based antennas provide wider bandwidths, greater efficiency, and dynamic beam steering capabilities, ensuring higher data rates with lower interference. This positions metasurfaces as critical enablers for the next generation of high-performance, compact, and energy-efficient communication systems.
Quantum and Entanglement Technologies
Quantum technologies are perhaps the most promising frontier for metasurface applications. Researchers at Peking University have pioneered a gradient metasurface that simplifies the generation of entangled photon states. By exploiting single-surface interference, photons entering from divergent directions can become entangled without the need for complex bulk optical setups. These metasurfaces not only facilitate the creation of entangled clusters, such as GHZ states, but also lay the foundation for portable quantum devices that are compact enough to fit on a chip.
Beyond entanglement generation, metasurfaces are integral to the development of quantum-enabled metamaterials. The International Conference on Metamaterials (META 2025), aligning with the International Year of Quantum Science, has spotlighted breakthroughs in topological photonics, where silicon-boron nitride hybrids guide photons along defect-resistant pathways. Additionally, quantum dots embedded in hyperbolic metamaterials are showing promise in shielding qubits from environmental decoherence, a crucial step toward stable, scalable quantum computing.
Sustainability and Commercialization
As metasurfaces transition from research labs to industry, attention is turning to manufacturability and sustainability. Market forecasts indicate explosive growth, with the global metamaterials market projected to rise from $500 million in 2021 to $3.2 billion by 2026—a compound annual growth rate of 36.4%. This acceleration is driven in part by nanoimprint lithography (NIL), which enables mass production at wafer scale for costs as low as $0.50 per lens. Companies like STMicroelectronics are rapidly acquiring intellectual property, such as from Metalenz, to integrate metasurfaces into smartphone sensors.
Eco-conscious innovation is also gaining traction. Transition-metal nitrides like titanium nitride are replacing gold in high-temperature optical applications due to their environmental benefits and thermal stability. Metasurfaces have even found a role in passive radiative cooling systems. Inspired by the thermoregulation mechanisms of Saharan silver ants, engineered metasurfaces are achieving sub-ambient cooling of up to 10°C without requiring any external power source. Field tests in Arizona have demonstrated HVAC energy savings of up to 30% when applied to building envelopes.
Conclusion: The Path to Ubiquitous Meta-Optics
Metasurfaces are no longer a scientific novelty—they are becoming the photonic equivalent of integrated circuits. With continued advances in AI-driven design tools and scalable manufacturing techniques, these nanostructures are set to permeate consumer and industrial markets alike. In the near future, we can expect smartphones equipped with polarization-sensing cameras, augmented reality glasses using metasurface waveguides, and ultimately, exascale photonic AI data centers and quantum communication networks distributing entangled states across global infrastructure.
This transformation is being propelled by a confluence of academic breakthroughs, industrial investments, and international collaboration, as exemplified by events such as META 2025 and the OPTICA Winter College. As physicist Federico Capasso aptly puts it: “We’re not iterating—we’re reinventing light itself.”
References and Resources also include:
https://www.photonics.com/Articles/Metasurface_for_Nonlinear_Manipulation_Could/p4/a64253
http://iopscience.iop.org/article/10.1088/0034-4885/79/7/076401/ampdf
https://www.photonics.com/Article.aspx?AID=65278
https://eurekalert.org/pub_releases/2020-04/ssfo-mle043020.php
https://news.engr.psu.edu/2020/ni-xingjie-photonic-integrated-chip.aspx
https://www.photonics.com/Articles/Nanostructures_Allow_High_Harmonic_Generation/a67196