The forthcoming sixth generation (6G) communication network is envisioned to provide ultra-fast data transmission and ubiquitous wireless connectivity. The terahertz (THz) spectrum, with higher frequency and wider bandwidth, offers great potential for 6G wireless technologies. However, the THz links suffers from high loss and line-of-sight connectivity. To overcome these challenges, a cost-effective method to dynamically optimize the transmission path using reconfigurable intelligent surfaces (RISs) is widely proposed. RIS is constructed by embedding active elements into passive metasurfaces, which is an artificially designed periodic structure. However, the active elements (e.g., PIN diodes) used for 5G RIS are impractical for 6G RIS due to the cutoff frequency limitation and higher loss at THz frequencies. As such, various tuning elements have been explored to fill this THz gap between radio waves and infrared light. The focus of this review is on THz RISs with the potential to assist 6G communication functionalities including pixel-level amplitude modulation and dynamic beam manipulation. By reviewing a wide range of tuning mechanisms, including electronic approaches (complementary metal-oxide-semiconductor (CMOS) transistors, Schottky diodes, high electron mobility transistors (HEMTs), and graphene), optical approaches (photoactive semiconductor materials), phase-change materials (vanadium dioxide, chalcogenides, and liquid crystals), as well as microelectromechanical systems (MEMS), this review summarizes recent developments in THz RISs in support of 6G communication links and discusses future research directions in this field.

Terahertz (THz) waves, which fall between microwaves and infrared bands, possess intriguing electromagnetic properties of non-ionizing radiation, low photon energy, being highly sensitive to weak resonances, and non-polar material penetrability. Therefore, THz waves are extremely suitable for sensing and detecting chemical, pharmaceutical, and biological molecules. However, the relatively long wavelength of THz waves (30~3000 μm) compared to the size of analytes (1~100 nm for biomolecules, <10 μm for microorganisms) constrains the development of THz-based sensors. To circumvent this problem, metasurface technology, by engineering subwavelength periodic resonators, has gained a great deal of attention to enhance the resonance response of THz waves. Those metasurface-based THz sensors exhibit high sensitivity for label-free sensing, making them appealing for a variety of applications in security, medical applications, and detection. The performance of metasurface-based THz sensors is controlled by geometric structure and material parameters. The operating mechanism is divided into two main categories, passive and active. To have a profound understanding of these metasurface-assisted THz sensing technologies, we review and categorize those THz sensors, based on their operating mechanisms, including resonators for frequency shift sensing, nanogaps for enhanced field confinement, chirality for handedness detection, and active elements (such as graphene and MEMS) for advanced tunable sensing. This comprehensive review can serve as a guideline for future metasurfaces design to assist THz sensing and detection.

Terahertz (THz) part of the electromagnetic spectrum (0.1–10 THz) holds the key for next-generation high-speed wireless communication, non-destructive biosensing, fingerprint chemical detection and imaging for astronomy and security surveillance. The limited THz response of naturally occurring materials had left a technological gap in the THz region of the electromagnetic spectrum. Artificially engineered materials termed as ‘metamaterials’, have shown great potential in THz wave interaction and its active counterpart termed as ‘metadevices’ have been widely reported for on-demand manipulation of THz waves. One of the most efficient means of realizing metadevices is to reconfigure the shape of unit cells and hence the corresponding THz response. The 50+ years of development in microelectromechanical systems (MEMS) and the wide array of microactuator designs provide a perfect platform to achieve structural reconfiguration of microscale metamaterial unit cells in both in-plane and out-of-plane directions. In this review, we present a comprehensive overview of various MEMS approaches adopted for the demonstration of THz metadevices, their advantages and limitations. The future research directions of THz MEMS metadevices are also discussed. The seamless integration of matured MEMS technology with incipient THz metamaterials provides significant advantages in terms of enhanced performances, advanced functionalities and large scale manufacturability, that is critical for the development of future THz technologies.

In this letter, a group of laterally coupled alternating thickness (LCAT) mode resonators based on 12% scandium doped aluminium nitride (Sc0.12Al0.88N) are reported. The LCAT mode resonators are realized in-house by sandwiching a layer of 1μm-thick Sc0.12Al0.88N between the top two sets and the bottom two sets of interdigitated electrode (IDE) fingers, all made of molybdenum (Mo). Experimental results show that an effective electromechanical coupling coefficient (keff2) of more than 10% and loaded quality factors (QL) as high as 870 are obtained, while the motional resistance of the resonators is kept as low as 0.68Ω. The figure-of-merit (FoM) which is defined as keff2×Q and phase velocity (vp) is subsequently derived to be 87, and 7840m/s, respectively. By changing the pitch of the IDE fingers, the resonators can operate from 1.96GHz to 2.11GHz without significant perturbation to their performance, corresponding to a lithographic frequency tuning range of 7.5%. In addition, clean spectra without any spurious modes are obtained within a 1GHz frequency range for all fabricated devices of various pitches, making a significant step forward in the commercialization path of the reported resonators in the applications of the high-band selection filters integrated on a single chip for mobile communication.

In recent years, considerable research advancements have emerged in the application of inverse design methods to enhance the performance of electromagnetic (EM) metamaterials. Notably, the integration of deep learning (DL) technologies, with their robust capabilities in data analysis, categorization, and interpretation, has demonstrated revolutionary potential in optimization algorithms for improved efficiency. In this review, current inverse design methods for EM metamaterials are presented, including topology optimization (TO), evolutionary algorithms (EAs), and DL-based methods. Their application scopes, advantages and limitations, as well as the latest research developments are respectively discussed. The classical iterative inverse design methods categorized TO and EAs are discussed separately, for their fundamental role in solving inverse design problems. Also, attention is given on categories of DL-based inverse design methods, i.e. classifying into DL-assisted, direct DL, and physics-informed neural network methods. A variety of neural network architectures together accompanied by relevant application examples are highlighted, as well as the practical utility of these overviewed methods. Finally, this review provides perspectives on potential future research directions of EM metamaterials inverse design and integrated artificial intelligence methodologies.