Recent Advances in Chiral and Achiral Metasurfaces Under Symmetry Preservation and Breaking

Abstract

Structural symmetry preservation and breaking play important roles in optical manipulation at subwavelength scales. By precisely engineering the symmetry of the nanostructures, metasurfaces can effectively realize various optical functions such as polarization control, wavefront shaping, and on-chip optical integration, with promising applications in information photonics, bio-detection, and flexible devices. In this article, we review the recent advances in chiral and achiral metasurfaces based on symmetry manipulation. We first introduce the fundamental principles of chiral and achiral metasurfaces, including methods for characterizing chirality and mechanisms for phase modulation. Then, we review the research on chiral metasurfaces based on material type and structural dimensions and related applications in high-sensitivity chiral sensing, reconfigurable chiral modulation, and polarization-selective imaging. We then describe the developments in the application of achiral metasurfaces, particularly in polarization-multiplexed holography, phase-gradient imaging, and polarization-insensitive metalenses. Finally, we provide an outlook on the future development of chiral and achiral metasurfaces.

1. Introduction

In optoelectronics, structural symmetry is one of the key factors influencing the interaction between light and matter [1]. This influence becomes particularly significant at subwavelength scales, where symmetry plays a vital role in governing the propagation direction, polarization evolution, and energy distribution of the optical field [2]. Highly symmetric structures typically support degenerate optical modes and exhibit characteristics such as directional uniformity and polarization insensitivity [3,4,5]. Once the symmetry is broken, the system may introduce effects such as modal splitting, anisotropic propagation, and polarization-selective coupling, which in turn may stimulate a variety of novel optical phenomena, such as asymmetric transmission, chiral response, and momentum non-conservation processes [6,7].

Recently, metasurfaces have emerged as an ideal platform for investigating the intrinsic link between structural symmetry and optical field manipulation, owing to their programmable and tailorable geometric properties [8,9,10,11,12,13]. By precisely designing the unit cell geometry of a metasurface, researchers can independently control multiple degrees of freedom, including phase, amplitude, and polarization at the nanoscale [14,15,16,17,18,19]. This capability enables the development of highly integrated and high-performance optical devices.

Among various mechanisms of optical field manipulation enabled by metasurfaces, the chiral effect, as a typical phenomenon caused by symmetry breaking, has received extensive attention [20]. Chirality refers to a geometric property whereby a structure cannot be superimposed on its mirror image [21]. It was first widely used in the fields of chemical and biological molecular recognition [22,23,24]. In recent years, this concept has been gradually extended to optical systems, manifested in the different responses of materials to left-handed circularly polarized light (LCP) and right-handed circularly polarized light (RCP), collectively referred to as optical chirality [25,26,27,28,29]. When circularly polarized light impinges on a chiral metasurface element, its electromagnetic field components couple asymmetrically within the structure, leading to pronounced differences between LCP and RCP in terms of propagation direction, energy dissipation, and momentum exchange [30,31,32,33,34,35]. Currently known optical chiral response mechanisms can be classified into two categories: intrinsic chirality arising from geometrically asymmetric architectures and extrinsic chirality induced in symmetric structures under tailored electromagnetic excitation conditions [36,37].

In contrast to chiral metasurfaces, achiral metasurface elements often possess certain degrees of mirror or rotational symmetry, and their polarization response depends critically on the specific symmetry order [38]. Only when a structure exhibits high-order rotational symmetry (such as ${\mathrm{C}}_4$ or higher) can it maintain polarization-insensitive behavior under varying incident polarization states [39,40,41]. For structures lacking high rotational symmetry, the phase response to LCP and RCP light may differ, resulting in residual polarization selectivity [42,43]. Therefore, achiral metasurfaces are not inherently polarization-insensitive, and their actual polarization response must be evaluated based on their specific symmetry characteristics.

The concept of “symmetry control” discussed in this paper encompasses not only optical field manipulation strategies based on symmetry preservation but also novel optical phenomena induced by deliberate symmetry breaking. To systematically explore the role of symmetry control in optical responses, this review focuses on the intrinsic relationship between structural symmetry and the optical behavior of metasurfaces, as illustrated in Figure 1. First, based on the principles of polarization manipulation, we systematically examine the manipulation mechanisms of propagation and geometric phases in achiral structures and introduce their applications in wavefront shaping, beam splitting, and focusing. Second, considering the distinct material properties of metals and dielectrics, we classify recent research on chiral metasurfaces, with an emphasis on advances in enhancing chiral optical responses, dynamically tunable devices, and applications in imaging and sensing. Finally, we discuss the challenges and future directions in the structural symmetry design of metasurfaces, aiming to provide theoretical insights and technical guidance for the development of high-performance photonic devices. Compared with existing review articles on metasurfaces, which often focus on specific materials, fabrication techniques, or application domains such as holography or sensing, this work offers a more distinctive perspective by emphasizing the central role of structural symmetry in metasurface design and functionality. We systematically review recent advances in both chiral and achiral metasurfaces under symmetry-preserving and symmetry-breaking strategies and further discuss how different optical modulation mechanisms are influenced by symmetry considerations and, in turn, affect device performance. We hope that this symmetry-centered analytical framework can serve as a valuable reference for the continued development of metasurface-based photonic devices.

2. Fundamental Principles of Chiral and Achiral Metasurfaces

2.1. Principles of Chiral Metasurfaces

The optical response of chiral metasurfaces is primarily characterized by two distinct phenomena: circular birefringence (CB) and circular dichroism (CD) [44]. CB manifests as the phase difference accumulated between LCP and RCP as light propagates through a medium. This phenomenon is conventionally termed optical rotation (OR). CD, on the other hand, describes the differential absorption, scattering, or transmission exhibited by LCP and RCP, serving as a critical parameter for evaluating the chiral optical properties of materials.

The magnitude of CD can be quantitatively expressed in two ways. Specifically, the absolute difference represents the intensity contrast between LCP and RCP light in terms of absorption, transmission, or reflection directly. Another way is to normalize the intensity difference by the total intensity, thereby enabling meaningful comparisons across samples with varying concentrations or thicknesses. The normalized CD is defined as

$$CD={\displaystyle \frac{K_{LCP}-K_{RCP}}{K_{LCP}+K_{RCP}}}$$

(1)

where K represents the absorption, transmission, or reflection coefficient. This normalization effectively eliminates the influence of background intensity, enabling standardized comparisons across different platforms and material systems. Normalized CD has found widespread application in areas such as chiral metasurface optimization, molecular chirality recognition, and pharmaceutical stereoisomer identification. For instance, in drug screening, the position and intensity of peaks in the CD spectrum can be used to accurately distinguish molecular configurations. Notably, the definition of CD and the method of normalization may vary across the literature. Therefore, when comparing data or evaluating performance, it is essential to ensure consistency in the calculation of CD values to avoid misleading conclusions.

The strength of the optical response in chiral structures depends on the matching between structural dimensions and the wavelength of the incident light. In naturally occurring organic molecules, the characteristic structural dimensions are typically much smaller than the wavelength of visible light, so their interaction with the optical field exhibits weak chirality, resulting in low CD values. When the structural feature size approaches the wavelength of the incident circularly polarized light, the chiral response can be effectively enhanced by exciting the resonance of the local surface to form a highly localized chiral light field. The intensity of such localized chiral fields can be quantified by the optical chirality density, C, which is defined as [45,46]:

$$C=-{\displaystyle \frac{\omega }{2c^2}}\cdot \mathrm{Im}({\overrightarrow{E}}^{*}\cdot \overrightarrow{H})$$

(2)

where $\overrightarrow{E}$ and $\overrightarrow{H}$ denote the complex electric and magnetic field vectors, respectively, ω represents the angular frequency of the incident light, and C is the speed of light in a vacuum. In chiral metasurfaces, the C value can be further improved by precisely adjusting the structural parameters of the metasurface.

In addition, the dissymmetry factor $g$ is commonly used to characterize the chiral properties of a material and is defined as [47]:

$${\displaystyle \frac{g}{g_{CPL}}}=-{\displaystyle \frac{|\overrightarrow{E}||\overrightarrow{H}|\cos ({\beta }_{i\overrightarrow{E},\overrightarrow{H}})}{\omega c|\overrightarrow{E}{|}^2}}$$

(3)

where $g_{CPL}$ denotes the dissymmetry factor for circularly polarized light, and ${\beta }_{i\overrightarrow{E},\overrightarrow{H}}$ represents the phase difference between the electric field $\overrightarrow{E}$ multiplied by the imaginary unit i and the magnetic field $\overrightarrow{H}$.

Therefore, a higher value of $g$ indicates stronger sensitivity and selectivity of the structure with respect to different optical chiral states and is widely used to analyze the spatial characteristics of near-field chirality distributions in metasurfaces.

To achieve a strong chiral optical response, it is essential to efficiently design nanostructures that excite and manipulate localized chiral optical fields. The different response of chiral metasurfaces to LCP and RCP is primarily attributed to the selective coupling and field enhancement associated with specific electromagnetic resonance modes excited within the structure. Among the various resonance mechanisms, localized surface plasmon resonances (LSPRs) and Mie resonances represent the two most extensively studied mechanisms, respectively governing the chiral optical behavior of metallic and dielectric metasurfaces. In metallic chiral metasurfaces, subwavelength metallic nanostructures interact with the incident light field, inducing collective oscillations of free electrons driven by electromagnetic excitation and resulting in strong localized resonant electric fields [48]. These fields significantly enhance the interaction between light and the nanostructure, as shown in Figure 2a. When the geometry of the structure exhibits broken mirror symmetry, LCP and RCP light can couple to plasmonic modes with different efficiencies and radiative losses, thereby leading to a pronounced chiral optical response. This asymmetry can be physically interpreted using the Born–Kuhn coupled oscillator model [49], as shown in Figure 2b. The model envisions two orthogonally arranged and mutually coupled electronic oscillators that respond to different components of the incident electric field. When linearly or circularly polarized light is incident, the differences in relative phase and coupling strength between the oscillators give rise to oscillation modes with specific polarization selectivity, thereby resulting in optical activity.

In dielectric chiral metasurfaces, when the frequency of the incident light matches the Mie resonance condition of the nanostructure, strong electric and magnetic dipole resonances are excited within the structure, thereby significantly enhancing the local electric and magnetic field distributions [50,51], as shown in Figure 2c. This enhancement of the localized electromagnetic fields greatly strengthens the interaction between light and the structure. In chiral geometries, LCP and RCP excite different electromagnetic modes. These modes exhibit distinct coupling strengths and phase characteristics, thereby amplifying the chiral optical response. Moreover, Mie resonances are often accompanied by multimodal interference effects. In structures lacking mirror symmetry, the interference between multiple resonance modes enhances polarization selectivity, thereby further strengthening the chiral optical response.

2.2. Principles of Achiral Metasurfaces

In metasurface-based optoelectronics, structural symmetry plays an important role in phase modulation. The control of optical phase by metasurfaces is generally achieved through the combination of geometric phase and propagation phase [52]. These two mechanisms are based on different physical principles, as shown in Figure 3, and their superposition enables precise wavefront engineering. The geometric phase originates from the spatial variation of the polarization state. When polarized light passes through anisotropic meta-atoms possessing rotational symmetry or a continuously varying orientation angle, the acquired phase is linearly proportional to the rotation angle of the structural unit. This phase depends solely on the evolution of polarization and is independent of the physical thickness or propagation path of the structure. The propagation phase, on the other hand, arises from the conventional phase delay accumulated during light propagation within the metasurface structure. It is determined by factors such as the refractive index of the material, the structural thickness, and the optical path difference.

When the metasurface structure exhibits achiral characteristics, the interaction between light and the metasurface imposes constraints on the evolution of the geometric phase. For achiral nanostructures, which do not possess optical activity, the Jones matrix can be specified in a particular form when the principal axis of the nanostructure is aligned with the coordinate system [53]:

$$J_g=\left[\begin{array}{cc}{t}_u& 0\\ 0& t_v\end{array}\right]$$

(4)

where t_u and t_v are the transmission coefficients along the u- and v-directions, respectively. When the nanostructure is rotated by an angle θ relative to the x-axis, the Jones matrix transformation is given by

$$J_{\varsigma }=R(-\theta )\cdot J_g\cdot R(\theta )=\left[\begin{array}{cc}{t}_u{\cos }^2\theta +t_v{\sin }^2\theta & (t_u-t_v)\sin \theta \cos \theta \\ (t_u-t_v)\sin \theta \cos \theta & t_u{\sin }^2\theta +t_v{\cos }^2\theta \end{array}\right]$$

(5)

When the incident light is circularly polarized, it can be represented as ${\left[1,i\delta \right]}^{\prime }$, where δ = ±1 corresponds to LCP and RCP, respectively. The transmitted light can be given by

$$\left[\begin{array}{c}{E}_{xout}\\ E_{yout}\end{array}\right]={\left(1/2\right)}^{1/2}\cdot J_{\zeta }\left[\begin{array}{l}1\\ i\delta \end{array}\right]={\displaystyle \frac{1}{2\sqrt{2}}}((t_u+t_v)\left[\begin{array}{c}1\\ i\delta \end{array}\right]+(t_u-t_v)e^{2i\delta \varsigma })\left[\begin{array}{c}1\\ -i\delta \end{array}\right]$$

(6)

From the above expression, it is clear that when circularly polarized light passes through an achiral nanostructure, the transmitted light contains both LCP and RCP. The polarization component orthogonal to the incident polarization acquires an additional geometric phase $2\delta \varsigma$. Specifically, when LCP and RCP are incident upon a mirror-symmetric nanostructure, the geometric phase imparted to the orthogonal polarization components exhibits a symmetric reversal. This distinctive optical property enables metasurfaces to independently manipulate orthogonal polarization states, thereby enabling the development of polarization multiplexing technology. This technology not only fully exploits the polarization degree of freedom of light but also significantly enhances the functional density of optical components. It has demonstrated great potential in advanced fields such as wavefront engineering, three-dimensional holography, broadband achromatic focusing, and image edge detection.

In particular, for achiral structures with ${\mathrm{C}}_4$ or higher rotational symmetry (such as square or octagonal silicon pillar arrays), the structure remains invariant after rotation by a certain angle around the symmetry axis, exhibiting quasi-isotropic optical responses. In such structures, the Jones matrix simplifies to an approximate identity matrix, and the geometric phase effect is effectively suppressed. The phase of the output light is entirely governed by the propagation phase induced by local resonances in the structure. Since the transmission phase mainly depends on the structure size and material dispersion, it is insensitive to the polarization state. This symmetry endows the metasurface with polarization robustness, making it widely applicable in broadband lenses, non-polarizing optical filters, and polarization-insensitive beam-shaping devices.

3. Chiral Metasurfaces

With the continuous advancement of metamaterial design principles and nanofabrication techniques, chiral metasurfaces have emerged as a key platform for the artificial manipulation of optical chirality. In recent years, chiral metasurfaces have attracted significant attention for their capabilities in polarization-selective control, chiral field enhancement, nonlinear optical response, and biosensing. The fundamental working mechanism relies on the differential coupling and enhancement of LCP and RCP light within the structure at subwavelength scales, thereby enabling the generation of intrinsic or extrinsic chiral optical responses. Depending on geometric dimensionality and material composition, chiral metasurfaces can be broadly categorized into metallic and dielectric types, as well as into three-dimensional (3D) and two-dimensional (2D) configurations. This section provides a systematic overview of the structural design and current development of representative metallic and dielectric chiral metasurfaces from the perspective of structural dimensionality.

3.1. Metallic Chiral Metasurfaces

3.1.1. 3D Structures of Metallic Chiral Metasurfaces

During the early development of chiral optics, 3D metallic chiral nanostructures garnered significant attention due to their unique optical activity. These structures typically utilize spatially asymmetric metallic nanounits to excite LSPRs, resulting in strong intrinsic chiral optical responses.

In 2012, Hentschel et al. [54] successfully fabricated an ordered array of bilayer gold nanodisks using multiple processes of electron beam lithography, as shown in Figure 4a. Their study demonstrated that the strength of the chiral response and the associated CD spectra could be effectively tuned by adjusting the geometric parameters and spatial arrangements of the nanostructures. This work was the first to reveal the dominant role of near-field coupling in plasmonic chirality, providing a foundational model for understanding chiral plasmonic interactions.

In 2013, Oh et al. [55] employed DNA origami techniques to construct highly controllable 3D helical metal structures, as shown in Figure 4b. This study provided the first observation of the coupling mechanism between metallic plasmons and DNA-based dielectric scaffolds. Through theoretical modeling, the study further demonstrated the effect of 3D structures on CD responses and identified the geometric modulation mechanism responsible for chirality sign reversal. This work laid the foundation for a deeper understanding of chiral light–matter interactions at the nanoscale.

Owing to the intrinsic ability to regulate the polarization state of optical fields, chiral nanostructures exhibit unique optical characteristics that are highly valuable for applications in miniaturized polarization-based communication and high-density polarization-encoded information storage. In 2009, Gansel et al. [56] designed and fabricated a uniaxial photonic metamaterial composed of a 3D gold helical array, as shown in Figure 4c. This nanostructure was fabricated by direct laser writing in positive photoresist, followed by electrochemical gold deposition. Experimental results showed that, when light propagated along the helical axis, the structure could completely block circularly polarized light matching the handedness of the helix. Meanwhile, it efficiently transmitted the opposite polarization over a bandwidth exceeding one octave. In addition, the structure can be geometrically tuned to operate across different frequency regimes, making it suitable for application as a compact broadband circular polarizer.

With advancements in nanofabrication, reconfigurable 3D structures based on the “nanokirigami” strategy have been developed. In 2020, Tang et al. [57] fabricated an ${\mathrm{C}}_3$-symmetric 3D kirigami structure using an FIB/SEM system on Au/SiN thin films, forming a reconfigurable plasmonic metasurface, as shown in Figure 4d. This structure exhibited a nonlinear CD response of up to 97% in second-harmonic generation (SHG). It demonstrated broadband and highly tunable nonlinear optical behavior and provided a new scheme for applications in chiral nonlinear optics.

In addition to physical microfabrication approaches, chemical synthesis strategies have shown unique advantages in constructing complex 3D chiral structures. In 2018, Lee et al. [58] used amino acids or peptides as templates to guide the growth of gold nanoparticles into helically twisted chiral structures, as shown in Figure 5a. Although the resulting particles were randomly distributed in solution, they still exhibited strong CD responses, confirming the effectiveness of their chiral structure at the macroscopic spectral level. In 2025, Li et al. [59] achieved atomically precise structural control using a supramolecular self-assembly strategy by tuning the non-covalent interactions between silver nanoclusters and organic ions, as shown in Figure 5b. The co-crystalline system exhibited more than three times higher chiral selectivity compared to conventional designs, providing a tunable structural platform for exploring chirality-dependent photonic devices, solid-state catalysis, and negative-index materials.

In recent years, dynamically tunable chiral structures have also garnered increasing attention. In 2025, Lao et al. [60] proposed a strategy combining femtosecond laser direct writing and capillary-driven self-assembly on thermoresponsive hydrogel substrates. This approach enabled the construction of 3D metallic structures with reversible chiral responses, as shown in Figure 5c. The structure enables chirality switching by modulating hydrophilic–hydrophobic interactions and adjusting geometric configurations under thermal stimulation. Results showed that the structure exhibited pronounced topology-related vortex dichroism (VD), providing a new approach for reconfigurable chiral photonic devices, intelligent sensing, and microfluidic manipulation.

3.1.2. 2D Structures of Metallic Chiral Metasurfaces

With the rapid development of nanofabrication technologies, researchers have gradually compressed 3D metallic chiral structures into subwavelength-thick 2D metasurfaces, achieving efficient control of chiral responses through planar metasurface design. This reduction in structural dimensionality not only significantly reduces fabrication complexity and optical losses but also utilizes physical mechanisms such as surface plasmon polaritons (SPPs), LSPRs, and geometric phase to realize precise control over the phase, polarization, and amplitude of light. The 2D metasurface provides a new platform for polarization optics, quantum control, and on-chip integrated photonics.

In 2016, Wang et al. [61] designed and fabricated a 2D metallic chiral metasurface, as shown in Figure 6a. A thin gold film was deposited on a sapphire substrate and patterned into a specific array using nanolithography. By coupling LSPRs and SPPs, the structure exhibited a highly tunable chiral optical response. In 2017, Ye et al. [62] designed a 2D chiral metasurface composed of arrays of L-shaped gold nanoantennas, as shown in Figure 6b. By tailoring the geometric parameters, they achieved synergistic excitation of LSPRs and propagating magnetic dipole resonances, resulting in a CD value of up to 0.68 in the near-infrared region.

Utilizing Fano resonance to further strengthen light–matter coupling is another key strategy for improving the performance of metallic 2D chiral structures. Zu et al. [63] proposed a metasurface structure based on gold heptamers, as shown in Figure 6c. This structure supports Fano resonance, exhibiting strong narrowband asymmetric responses under LCP and RCP. The CD value achieved a maximum of 30% near the resonance, demonstrating the potential of resonant interference in enhancing optical chirality.

The above focuses on the intrinsic optical chiral effects determined by the geometric asymmetry of the metasurface nanocells themselves. Beyond intrinsic chirality, some studies have explored the concept of “extrinsic chirality.” Distinct from this intrinsic chirality, which stems from the nanocells’ inherent geometric properties, the extrinsic chiral response arises through a completely different mechanism: it is dictated by the relative orientation between the incident-light direction and the overall lattice symmetry of the metasurface. Even when the metasurface elements themselves are mirror-symmetric (for example, simple nanodisks or squares), oblique illumination causes the incident wave vector, the lattice periodicity vector, and the surface normal to form a chiral reference frame that breaks the structure’s overall mirror symmetry. This externally imposed symmetry breaking can trigger near-field interference between plasmonic modes, thereby producing pronounced chiro-optical effects. Therefore, the extrinsic chiral response is induced and controlled entirely by external optical conditions, providing a new degree of freedom for tailoring chiral optical behavior. near-field interference between plasmonic modes. In 2016, Cao et al. [64] designed a metal–dielectric–metal (MDM) trilayer metasurface that exhibited dual-band extrinsic chirality in the terahertz regime, as shown in Figure 6d. By combining oblique illumination with an asymmetric aperture array, the structure exhibited strong circular polarization CD. In contrast, conventional single-layer metallic structures, lacking magnetic dipole enhancement mechanisms, failed to demonstrate similar behavior. In 2019, Horrer et al. [65] designed a non-chiral plasmonic supramolecule structure consisting of three gold nanodisks arranged in an equilateral triangle, as shown in Figure 6e. This structure exhibited locally induced optical chirality at a wavelength of 770 nm, which originated from near-field interference between plasmonic modes. Although the structure itself lacks chirality and shows no far-field circular polarization selectivity, its near-field exhibits distinct intensity distributions under LCP and RCP illumination. This phenomenon originates from dipolar mode interference and the phase correlation of circular polarization components at the excitation wavelength, revealing polarization-dependent features in the near-field states of non-chiral structures.

3.2. Dielectric Chiral Metasurfaces

Dielectric chiral metasurfaces have attracted significant research attention due to their low optical losses, strong magnetic dipole resonances, and excellent material compatibility. These structures not only enable the excitation of electric and magnetic multipolar interactions through Mie resonances but also facilitate strong near-field optical chirality enhancement. Moreover, they are fully compatible with standard CMOS fabrication techniques.

3.2.1. 3D Structures of Dielectric Chiral Metasurfaces

3D dielectric chiral metasurfaces offer greater structural degrees of freedom and enable more complex symmetry-breaking and electromagnetic mode-control mechanisms. These features provide significant advantages in enhancing chiral response strength and expanding operational bandwidth. In 2020, Kang et al. [66] proposed a nonlinear chiral metasurface composed of amorphous silicon (α-Si) split-ring resonator arrays integrated with a silver backplane, as shown in Figure 7a. By leveraging Fabry–Pérot cavity enhancement of high-Q Mie resonances, this design achieved picosecond-scale polarization switching at 1530 nm in the near-infrared region. Under LCP illumination, the structure exhibited an absorption rate exceeding 90%, while the reflectance under RCP light reached up to 80%. This single-layer, low-power, and fast-response metasurface offers a promising platform for on-chip optical communication, chiral sensing, and quantum information modulation. In 2024, Kilic et al. [67] developed an all-dielectric L-shaped metamaterial composed of tilted nanopillar arrays, as shown in Figure 7b. The structure was fabricated via glancing angle deposition (GLAD) and exhibited significantly enhanced chirality over a broad spectral range from the near-infrared to the ultraviolet. Its physical mechanism is attributed to the cooperative resonance between electric dipole and magnetic dipole. The study further demonstrated that by adjusting parameters such as nanopillar tilt angle and height, the direction, strength, and spectral position of the chiral response can be flexibly tuned, providing a design paradigm for large-area, low-loss chiral photonic devices.

Based on conventional Mie resonances, dielectric structures often face the conflicting issue of low Q-factors and limited bandwidth, which severely restricts their performance improvement in areas such as nonlinear manipulation and high-sensitivity detection. To overcome this bottleneck, the concept of bound states in the continuum (BICs) has been introduced. Ideal BICs are non-radiative states that do not couple to radiation modes and, in theory, can achieve infinitely high Q-factors. In practice, quasi-BICs (q-BICs) are realized by breaking structural symmetry. This approach allows the state to maintain high Q-factors while enabling coupling to external optical fields. In 2023, Kühner et al. [68] achieved the first realization of a quasi-BIC with maximal optical chirality in the visible regime by tuning the out-of-plane symmetry of dielectric resonators, as shown in Figure 7c. The structure showed narrowband resonant absorption (Q ≈ 80) under incident light of one circular polarization while remaining highly transmissive to the opposite handedness. Its corresponding transmission contrast (ΔT) reached ±0.7, which is close to the theoretical chiral limit. In 2024, Li et al. [69] designed a dielectric chiral metasurface based on a sector-shaped unit cell with ${\mathrm{C}}_6$ symmetry, in which the topological features of BICs were exploited to simultaneously achieve high Q-factors and strong CD responses, as shown in Figure 7d. By breaking either in-plane or out-of-plane symmetry, the integer topological charge V point of the BIC was split into two half-integer C points, inducing both intrinsic and extrinsic chirality within the structure. The q-BIC states, excited via toroidal dipole (TD) modes, yielded exceptionally strong chiral responses with CD values of approximately 0.92 and 0.98 at 938.86 nm and 939.85 nm, respectively, further advancing the application of dielectric chiral structures in topological photonics and precision sensing.

3.2.2. 2D Structures of Dielectric Chiral Metasurfaces

In the development of dielectric chiral metasurfaces, 2D planar configurations have increasingly emerged as a focal point due to their unique integration advantages and simplified fabrication processes. Compared to 3D structures that often rely on multi-step lithography or self-assembly techniques, 2D designs enable high-precision patterning on standard nanofabrication platforms, making them more suitable for the low-cost, large-area, and scalable production of chiral functional devices. Additionally, benefiting from their subwavelength thickness and compact layout, 2D dielectric chiral metasurfaces offer a superior balance among optical performance, response speed, and system compatibility. These characteristics position them as a critical enabling technology for on-chip photonics and multifunctional integration.

In 2014, Wu et al. [70] designed an all-dielectric metasurface based on CMOS-compatible silicon materials, as shown in Figure 8a. The structure integrated straight and curved silicon nanorods to form asymmetric unit cells, and by breaking in-plane symmetry, it triggered Fano resonances, thereby achieving high-Q resonant modes (Q > 100) on a low-loss silicon platform. The metasurface efficiently converted linearly polarized light into circularly polarized light, achieving a polarization conversion efficiency exceeding 50%. Its excellent integration compatibility provided a novel structure for infrared polarizers, integrated spectroscopic devices, and photonic chips. In 2018, Ma et al. [71] proposed a dielectric chiral metasurface based on high-index germanium (Ge) Z-shaped resonators, as shown in Figure 8b. The structure achieved a CD value as high as 0.8 and strong asymmetric transmission in a planar configuration, which significantly exceeds the theoretical limit of conventional metallic metasurfaces. It operated effectively under arbitrary linear polarization without the need for additional polarization components and further enabled 2D holography through geometric phase modulation, demonstrating strong potential for functional integration in photonic systems.

In recent years, researchers have introduced the BIC mechanism into planar structures to boost chiral optical responses. In 2023, Du et al. [72] designed a lattice metasurface based on H-shaped ${\mathrm{SiO}}_2$ meta-atoms, as shown in Figure 8c. By breaking in-plane mirror and inversion symmetries, the structure supported multiple chiral quasi-BIC modes and achieved high Q-factors and strong CD in the terahertz regime. The chiral resonant wavelength could be tuned by adjusting the structural parameters. In the same year, Li et al. [73] reported a fully silicon-based terahertz metasurface, as shown in Figure 8d. The design breaks both in-plane ${\mathrm{C}}_2$ and mirror symmetries, enabling the excitation of three high-Q quasi-BIC modes: magnetic dipole, electric dipole, and magnetic quadrupole. These modes exhibited linewidths of 0.3, 0.15, and 0.2 GHz, respectively, and collectively achieved Q-factors up to ${10}^4$, demonstrating excellent chiral selectivity and frequency resolution. In 2024, Zhang et al. [74] constructed a planar chiral structure composed of rectangular grating dimers integrated with a waveguide layer, as shown in Figure 8e. By tuning in-plane symmetry, high-Q quasi-BIC resonances were excited. The structure achieved 99.5% reflectance for LCP light and near-complete transmission for RCP light at a wavelength of 649.56 nm, corresponding to a CD value close to 0.995, exhibiting extreme polarization selectivity.

For dielectric metasurfaces, tuning the external excitation conditions can also induce extrinsic chirality, leading to unique resonant responses such as anti-crossing behavior in mode splitting. In 2024, Toftul et al. [75] introduced a novel approach for realizing chiral responses by breaking the symmetry of a monoclinic lattice, as shown in Figure 8f. Even though each individual meta-atom was achiral, the overall metasurface exhibited significant linear and nonlinear CD. Theoretically, the authors introduced a modal chirality parameter to quantify the coupling between the chiral eigenmodes of the lattice and circularly polarized light. They also revealed the mechanism by which the monoclinic lattice angle controls the modal chirality response. Experimentally, the silicon-based metasurface demonstrated pronounced linear and nonlinear CD effects in the near-infrared range, confirming the feasibility of symmetry-breaking-induced chiral responses. This approach enables efficient chiral light–matter interaction solely through lattice geometry, without requiring chiral meta-atoms, offering a new pathway for applications in nonlinear optics and chiral sensing.

3.3. Applications of Chiral Metasurfaces

3.3.1. Chiral Sensing

Accurate identification and quantitative detection of chiral molecules are essential in areas such as drug development, biomedicine, and food safety. However, conventional detection methods, such as liquid chromatography and CD spectroscopy, face challenges including insufficient sensitivity, complex and time-consuming labeling procedures, and inability to perform real-time analysis. In recent years, metasurface-based strategies leveraging localized field enhancement, multimode resonance coupling, and spin–orbit interaction have led to significant advances in both the sensitivity and versatility of chiral sensing.

In 2014, Zhao et al. [76] designed a plasmonic metasurface composed of multilayered twisted gold nanorods, achieving highly sensitive detection of chiral molecules down to the 10⁻²¹ mole scale, as shown in Figure 9a. The structure exhibited strong near-field chiral enhancement across the visible to near-infrared spectral range. This enhancement significantly improved the detectability of CD signals while suppressing background interference, providing a promising strategy for identifying trace amounts of enantiomers. In 2021, Zhang et al. [77] proposed a detection method based on a reflective terahertz time-domain polarization spectroscopy system integrated with a chiral metasurface, as shown in Figure 9b. Utilizing the artificial chirality of a helical metallic metasurface under oblique incidence, the system enabled high-sensitivity quantitative detection and D/L enantiomer discrimination of amino acid solutions, achieving detection limits as low as 10⁻⁵–10⁻⁴ g/mL, with the signal of tyrosine detectable down to 1 × 10⁻⁵ g/mL. By analyzing differences in frequency-dependent responses of polarization ellipse angle, rotation angle, CD, and OR, the system effectively distinguished D/L enantiomers at the same concentration. Notably, for proline at a concentration of 0.6 g/mL, a PRA difference of up to −42° was observed between the D- and L-forms, demonstrating practical potential in biological sample analysis.

Compared to conventional resonance modes, high-Q BICs can significantly enhance sensor sensitivity. In 2020, Chen et al. [78] presented an integrated molar chiral sensing technique based on high-Q metasurfaces, as shown in Figure 9c. By simultaneously exciting quasi-BIC modes under both x- and y-polarized illumination, the metasurface generated a highly superchiral near field, leading to a 59-fold enhancement in the CD reflection signal from the chiral analyte. The system achieved a high refractive index figure of merit (FOM) of 80.6 and a sensitivity of 80.6 nm/RIU, corresponding to a molar concentration resolution of approximately 9.4 × 10⁻⁴ mol/cm³. This platform enabled simultaneous quantification of enantiomer ratios and molar concentrations within a single measurement, thereby streamlining complex chiral analysis and significantly enhancing detection efficiency.

To further enhance the specific recognition ability of chiral molecules, Liu et al. [79] introduced a Pancharatnam–Berry (PB) phase metasurface that significantly enhanced CD signals in the terahertz regime through spatial spin-state separation, as shown in Figure 9d. The PB metasurface achieved efficient separation of LCP and RCP beams via spatial gradient phase modulation. When integrated with an angle-resolved terahertz time-domain polarization spectroscopy system, the polarization rotation angle difference of d-tyrosine reached 16.4°, and that of l-tyrosine reached −11.6° at a beam deflection angle of ±33°, representing enhancements of approximately 9.3-fold and 11.9-fold, respectively, compared to the values obtained without the metasurface (1.6° and −0.9°). This approach utilized the frequency-angle-dependent beam deflection properties of the metasurface to significantly enhance the interaction between chiral molecules and spin-polarized terahertz beams at specific frequencies.

Hybrid material platforms have also emerged as a promising direction for sensor performance enhancement. Luo et al. [80] developed a hybrid circular tetramer metasurface (HCTM) consisting of gold (Au) and silicon nitride (${\mathrm{Si}}_3{\mathrm{N}}_4$), as shown in Figure 9e. By combining the high sensitivity of metals with the low-loss properties of dielectrics, they constructed a dual-channel qBIC system. Under normal incidence, the dielectric-dominated qBIC mode (d-qBIC) achieved a high Q-factor of 412, a bulk refractive index sensitivity of 492.7 nm/RIU, and an FOM of 266.3 RIU⁻¹. Meanwhile, the metal-dominated qBIC mode (m-qBIC) demonstrated strong surface affinity in biotin–streptavidin biosensing, with a maximum resonance shift of 1.8 nm. Furthermore, detection limit analysis based on the standard deviation (3σ) of resonance wavelength fluctuations revealed that the d-qBIC mode could detect biomolecules at concentrations as low as 2.4 nm, while the m-qBIC mode exhibited a detection limit of 2.8 nm. For comparison, the surface plasmon resonance (SPR) mode only demonstrated a detection limit of 3.0 nm under identical conditions. This hybrid architecture overcomes the performance limitations of conventional sensors by simultaneously leveraging the high sensitivity of metals and the low-loss advantages of dielectric materials.

3.3.2. Dynamic Chirality Tuning

In recent years, dynamically tunable chiral metasurfaces based on phase-change materials, mechanical deformation, and active medium integration have undergone rapid development. By applying optical, thermal, electrical, or mechanical stimuli, real-time reconfiguration of chiral responses can be achieved, including reversible modulation of CD sign, intensity, and resonance wavelength. These structures transcend the functional limitations of traditional static chiral devices, offering dynamic and multifunctional approaches for intelligent photonic systems, multimodal sensing, and information encoding.

In 2021, Probst et al. [81] proposed a mechanically reconfigurable chiral metasurface combining colloidal self-assembly with a flexible substrate, as shown in Figure 10a. The structure consists of crossed gold nanoparticle chains deposited on an elastic substrate. By adjusting the stacking angle (±45°) and applying external compression, dynamic control of CD signals from the visible to near-infrared spectral range was successfully demonstrated. In the experiment, the adjustable range of CD value reached 11°, and continuous and reversible adjustment of its sign and peak position was achieved. This low-cost and mechanically flexible approach offers a promising solution for flexible optoelectronics, adaptive sensing, and multi-channel recognition.

Beyond mechanical tuning, phase-change materials provide an efficient and reversible mechanism for dynamic chiral control. In 2024, Sha et al. [82] introduced a resonance-phase-change-based method for dynamically tuning chirality, as shown in Figure 10b. By integrating quasi-BICs with the phase-change material ${\mathrm{Sb}}_2{\mathrm{S}}_3$, continuous and reversible inversion of the CD sign was achieved. The designed metasurface consisted of obliquely aligned silicon nanorods. During the phase transition from amorphous to crystalline states, refractive index variations enabled precise tuning of the resonance wavelengths of two chiral quasi-BIC modes. Around 1505 nm, the structure successfully achieved a wide range of continuous tuning of CD values from −0.947 to +0.958. Experimental results further confirmed the light-pumped phase-change response, with a minimum switching time of 800 ns. Moreover, by controlling intermediate states of the phase-change material, the research team demonstrated the chirality continuum effect for the first time, providing a viable strategy for achieving high-speed, high-contrast dynamic chiral manipulation.

In 2025, Zhang et al. [83] developed a dynamic chiral metasurface operating in the infrared regime using the phase-change material GST-225, as shown in Figure 10c. The structure incorporated vertical tilt angles and trapezoidal nanopores to break structural symmetry and employed a BIC resonance mechanism to achieve highly selective chiral responses with high CD (>0.8) and narrow bandwidth (~1 nm). By varying the heating temperature (20–200 °C), dynamic tuning of the resonance wavelength was realized within the 4650–4660 nm range while maintaining stable CD performance. The study also revealed that the response was driven by the electric dipole and magnetic dipole coupling mechanism, and analyzed the effects of structural period, nanopore size, and GST thickness on the performance.

In addition to intrinsic chirality control, in 2024, Lv et al. [84] proposed an extrinsic chirality manipulation method based on BIC resonance, as shown in Figure 10d. The silicon-based metasurface achieved a high CD response by breaking both in-plane and out-of-plane symmetries, and further utilized obliquely incident circularly polarized light to excite extrinsic chirality, enabling selective absorption manipulation of LCP and RCP. It exhibited significant intensity tunability with variation in the incident angle.

3.3.3. Chiral Imaging

As a frontier direction in optics, by regulating the chiral interaction between light and matter, chiral imaging offers new paradigms for polarization-sensitive imaging, information encryption, and the design of nanophotonic devices. In 2017, Zhang et al. [85] proposed a monolayer all-dielectric chiral metasurface based on the mechanism of asymmetric spin–orbit interaction (SOI), enabling simultaneous control over highly asymmetric transmission (AT) of circularly polarized light and arbitrary wavefront shaping, as shown in Figure 11a. The structure consists of two asymmetrically shaped silicon nanofin pairs arranged with a relative rotation angle of π/4. This configuration breaks the inversion symmetry of the phase gradient, thereby inducing spin-selective interference effects. At the wavelength of 9.6 μm, the metasurface achieved an AT parameter of 0.69 and an extinction ratio of approximately 10:1. Building upon this design, the study further demonstrated three types of imaging devices based on the proposed metasurface. The first is a polarization beam deflector capable of effectively separating LCP and RCP components, achieving a +1 diffraction efficiency of up to 67.4% and a polarization contrast ratio of 16:1. The second is a vortex beam generator that produces annular high-contrast beams with a topological charge of +2. The third is a phase hologram that generates a single goldfish image in both the transmission and reflection fields, effectively eliminating interference from conventional conjugate images. Moreover, broadband testing in the 9.3–10.6 μm range showed that the ±1 diffraction contrast remained stable between 11.1 and 18.2, further confirming the metasurface’s potential for broadband chiral imaging and optical communication applications.

Researchers have further explored the integration of multilayer chiral metamaterials with novel fabrication techniques to overcome material and process limitations and advance chiral imaging toward integrated multifunctionality. In 2023, Zhang et al. [86] proposed a Moiré double-layer nanoimprinted chiral metasurface device based on a low-refractive-index contrast material (SU-8), as shown in Figure 11b. The device consists of two hexagonally arranged nanopillar layers. By adjusting the relative rotation angle between the layers, a Moiré interference pattern is introduced, which effectively breaks the structural symmetry and induces a strong CD response. Under different twist angles and interlayer distances, the device exhibits energy flow singularities and phase vortex structures in the near field and generates a CD signal exceeding 10% in the far field. Building on this effect, the team demonstrated polarization-state visualization by exploiting the difference in transmittance between LCP and RCP light. This approach transformed otherwise invisible polarization information into observable intensity contrast in the image, achieving micrometer-level spatial resolution and producing opposite CD distributions for LCP and RCP channels.

In 2024, Yuan et al. [87] proposed a fully polarized holographic display scheme based on a chirality-assisted metasurface, enabling both information encryption and secure imaging through independent phase modulation of four circular polarization components, as shown in Figure 11c. By co-engineering chirality-assisted phase, propagation phase, and geometric phase, the design decouples the diagonal and off-diagonal elements of the transmission matrix under circular polarization, forming a multilayer stacked metasurface structure. The hologram is decomposed into four sub-images, each mapped to the co-polarized and cross-polarized channels of LCP and RCP light, thereby enabling multi-channel encryption control. Experimental results in the 10 GHz microwave regime demonstrated successful holographic reconstruction of the “HIT” pattern. When the input linear polarization matches the pre-defined key, the system achieves high imaging fidelity, with a reconstruction efficiency of 96.5% and a signal-to-noise ratio of 8.1 dB. The focal spot resolution reached 1.064λ₀/NA, approaching the theoretical Airy limit of 1.22λ₀/NA. Moreover, under arbitrary elliptical polarization input, only specific polarization states yield clear image recovery, further validating the high polarization selectivity and robust encryption capability of the approach.

3.4. Performance Comparison and Summary

Chiral metasurfaces constructed from either metallic or dielectric materials exhibit distinct design principles and optical characteristics due to fundamental differences in their resonance mechanisms, material properties, and fabrication processes. A systematic comparison of these two categories across several performance dimensions is presented in

Table 1. Comparison of metallic and dielectric chiral metasurfaces.

Structure Type	Typical Materials	Resonance Mechanism	CD Value (Typical)	Q-Factor	Bandwidth	Fabrication Complexity	CMOS Compatibility	Surface Affinity
Metallic	Au, Ag, Al	LSPR	0.1–0.6	Low (10–50)	Narrow	Medium (etching, self-assembly)	Poor	Strong (near-field)
Dielectric	Si, Ge, TiO2, GaN	Mie, q-BIC	0.6–0.85	High (100–1000+)	Tunable/Wide	High (EBL, GLAD)	Good	Moderate (internal field)

.

Metallic chiral metasurfaces primarily rely on LSPRs, which enable moderate to strong near-field enhancement and optical activity. These structures are particularly effective in the visible and near-infrared spectral regions. Reported CD values typically range from 0.1 to 0.6, depending on the specific nanostructure geometry, excitation polarization, and illumination conditions. However, their Q-factors and spectral selectivity are inherently limited by Ohmic losses associated with metals. In contrast, dielectric chiral metasurfaces—especially those based on Mie resonances or quasi-BICs—offer high-Q resonances (Q > 100) with minimal absorption. Optimized dielectric designs have demonstrated CD values as high as 0.85, particularly in the near-infrared and terahertz regions. These metasurfaces are advantageous for applications requiring narrowband response and high spectral resolution. Nevertheless, their internal field confinement may limit surface-based sensing efficiency, and their fabrication often involves more sophisticated processes such as electron-beam lithography (EBL) or GLAD.

Beyond the material domain, the structural dimensionality of chiral metasurfaces also plays a critical role in determining their optical performance. Three-dimensional metasurfaces offer enhanced degrees of freedom for symmetry manipulation and electromagnetic field control, which can enable more intricate and tailored chiral responses. However, their practical implementation is often constrained by increased fabrication difficulty and limited compatibility with conventional planar processes. By contrast, two-dimensional planar metasurfaces provide simpler fabrication routes, excellent compatibility with photonic integration platforms, and versatile polarization control through geometric phase engineering. Still, the inherent in-plane symmetry and restricted geometric diversity may limit the achievable strength and variety of chiral optical functionalities.

Together, these interrelated material and structural factors underscore the importance of deliberate platform selection in designing high-performance chiral metasurfaces for targeted applications in sensing, imaging, and light–matter interaction engineering.

4. Achiral Metasurfaces

Achiral metasurfaces are typically composed of nanostructural units exhibiting mirror symmetry. Although these structures themselves lack chirality, achiral metasurfaces can still exhibit a wide range of optical functionalities under specific polarization conditions. In particular, they show great potential for applications in polarization multiplexing, phase imaging, and polarization-insensitive device development. A detailed discussion is provided below.

4.1. Polarization Multiplexing

Polarization multiplexing is a multi-channel encoding and decoding technique based on the orthogonal polarization states of light waves. It enables a substantial increase in information capacity and transmission efficiency within limited spatial resources. In achiral metasurfaces, by independently manipulating the optical field distribution, phase gradient, or amplitude profile associated with each orthogonal polarization state, multiple isolated channels can be established on a single optical platform, thereby facilitating the realization of high-density information multiplexing systems.

In 2015, Arbabi et al. [88] designed a polarization-sensitive metasurface composed of high-refractive-index elliptical dielectric nanopillar arrays, as shown in Figure 12a. This metasurface enables simultaneous control over polarization and phase at the subwavelength scale, allowing the integration of multifunctional devices such as polarization-switchable holograms and vector beam generators. In the same year, Wen et al. [89] proposed a polarization-multiplexed metasurface holography technique. By precisely tuning the arrangement and parameters of the nanostructures, they achieved high-fidelity holographic reconstruction over a broad spectral range, as shown in Figure 12b. The structure encoded two different image contents under LCP and RCP, enabling polarization-domain-specific control and image switching. Additionally, Huang et al. [90] introduced a multidimensional information encryption and multiplexing platform that employed polarization state, spatial position, and incident angle as multiple keys for holographic reconstruction, as shown in Figure 12c. This approach supported a variety of multiplexing strategies while effectively suppressing spectral crosstalk among multiple images, thereby greatly improving the efficiency of optical information processing. In 2018, Zhao et al. [91] demonstrated a multi-polarization multiplexing platform based on rectangular silicon nanofins, as shown in Figure 12d. By exploiting the birefringence and geometric rotation degrees of freedom of the structure, the platform achieved flexible control over polarization-dependent manipulation. In combination with a modified Gerchberg–Saxton algorithm, this approach enabled independent multiplexing across up to 12 combinations of input and output polarization states. These capabilities provide a technical foundation for multi-channel optical communication.

In recent years, polarization-multiplexed metasurfaces have also demonstrated unique advantages in the generation of complex structured light fields. In 2021, Li et al. [92] proposed a Bessel beam generator based on geometric phase control, as shown in Figure 12e. By exploiting the independent responses of an achiral metasurface to orthogonal polarization states, the system enabled the flexible generation of Bessel beams corresponding to two distinct polarization channels. Based on this foundation, in 2024, Zhou et al. [93] developed an integrated polarization-multiplexed structured illumination system for super-resolution imaging, as shown in Figure 12f. The metasurface was constructed based on a single-layer silicon nanopillar array. Through the coordinated design of geometric and propagation phases, the system achieved spatial multiplexing of three phase-shifted structured light channels without requiring mechanical movement or dynamic manipulation, as shown. Its spatial resolution surpasses the diffraction limit of traditional systems by approximately 1.8 times, laying a solid technical foundation for the application of polarization multiplexing in the field of high-performance imaging.

4.2. Phase Imaging

Achiral metasurfaces can achieve efficient separation of orthogonal polarization states at the subwavelength scale by constructing periodic phase-gradient nanostructure arrays. This approach offers an integrated, scan-free solution for phase imaging technologies. In such devices, the geometric phase effect enables polarization-dependent directional separation of light in the Fourier space, thereby facilitating spatial mapping of phase gradients and interference-based modulation.

In 2019, Zhou et al. [94] proposed a phase-gradient imaging optical system based on an achiral metasurface, as shown in Figure 13a. In this design, the metasurface units are arranged along the x-axis with a periodic phase-gradient profile. When linearly polarized light is incident on the metasurface, the geometric phase effect decomposes the optical field into LCP and RCP components, which experience symmetric lateral displacements in the Fourier domain ($\Delta =\pm \lambda f/\Lambda$). This enables polarization-based beam splitting and controllable wavefront shearing of the object light. Due to the spatial separation of the two polarization channels, phase gradient information of the object wavefront can be extracted during fringe reconstruction in the interference process. When the shear amount Δ is much smaller than the image feature size, the weak shear condition is met, and the light intensity output by the system is proportional to the phase gradient of the object along the x-axis. Although this lateral shear interferometry method allows for one-dimensional phase retrieval, it has certain limitations in 2D imaging.

To solve the above problems, in 2020, this team [95] designed a radially symmetric geometric phase metasurface to achieve 2D spatial differentiation and isotropic phase gradient imaging, as shown in Figure 13b. It can be regarded as the result of radial rotation and superposition of infinite one-dimensional phase gradient geometric phase metasurfaces. Combined with the numerical integration algorithm, the system can recover the isotropic phase distribution. In addition to shear interferometry, achiral metasurfaces can also enable phase imaging based on the Transport-of-Intensity Equation (TIE) by controlling beam propagation through polarization manipulation, as illustrated. In 2021, Engay et al. [96] proposed a single-shot phase imaging method using an all-dielectric achiral metasurface, as shown in Figure 13c. By tailoring the structural design, the metasurface spatially separates the incident TE and TM polarization components and introduces a specific defocus phase delay to the TM component. This design enables the formation of both in-focus and defocused intensity images simultaneously within a 4F optical system. These two types of images can be captured in a single camera exposure, and the phase distribution on the sample surface is subsequently reconstructed by numerically solving the TIE using a finite difference method. This approach eliminates the need for multi-frame exposures required in traditional TIE-based techniques, significantly improving imaging efficiency and system integration.

4.3. Polarization-Insensitive

Based on the phase modulation mechanism of achiral metasurfaces, polarization-insensitive optical devices typically use nanostructural units with high rotational symmetry (e.g., ${\mathrm{C}}_4$ or higher) as their basic structural components. These units exhibit equivalent optical responses under different polarization states, effectively suppressing polarization dependence and enhancing the stability and applicability of devices in complex polarization environments.

Polarization-insensitive metasurfaces are typically realized by tuning the geometrical parameters of subwavelength nanopillars, such as diameter, height, and duty cycle, thereby constructing a precise equivalent refractive index distribution in space and enabling continuous modulation of the optical phase across the surface. In 2016, Khorasaninejad et al. [97] developed a polarization-insensitive metalens based on high-index titanium dioxide (${\mathrm{TiO}}_2$) nanopillars, as shown in Figure 14a. By varying the diameter of the nanopillars, the transmission phase was modulated to achieve equivalent focusing performance for differently polarized linearly incident light at a wavelength of 532 nm. Experimental results confirmed that the focal spot shape, size, and focusing efficiency showed negligible variation across polarization states, validating the excellent polarization robustness of the design. In 2021, Sun et al. [98] proposed a computational wavefront-encoded, polarization-independent achromatic metalens, as shown in Figure 14b. By optimizing the arrangement of simple-shaped nanopillars (circular or square), they achieved focal depth overlap across a broad spectral range (1300–1700 nm), thereby realizing broadband achromatic focusing that is insensitive to polarization.

5. Conclusions

In metasurfaces, structural symmetry and asymmetry play a pivotal role in regulating light–matter interactions. By carefully designing the symmetry of the unit cells, it is possible to achieve polarization insensitivity, band degeneracy, and efficient wavefront control. In contrast, the introduction of symmetry breaking can stimulate a rich variety of optical phenomena, including chiral responses, anisotropic scattering, and asymmetric transmission. Through precise structural control, metasurfaces have demonstrated broad application prospects across frontier fields such as optical communications, biosensing, quantum optics, imaging systems, and information processing.

This review systematically summarizes recent advances and applications related to the intrinsic relationship between structural symmetry and the optical functionalities of metasurfaces, from the representative perspectives of achiral and chiral systems. Achiral metasurfaces, by synergistically modulating the propagation and geometric phases, enable efficient control over polarization states, phase distributions, and wavefront profiles, exhibiting significant advantages in applications such as polarization multiplexing, polarization-insensitive devices, planar lenses, and optical field shaping. In contrast, chiral metasurfaces enhance optical chirality through intrinsic geometric configurations or extrinsic mode excitation, enabling strong selective control over circularly polarized light and offering unique solutions in areas such as chiral sensing, optical imaging, and dynamic chirality manipulation.

Driven by advances in fabrication technologies and theoretical modeling, symmetry control in metasurfaces is expected to evolve from static geometric optimization toward dynamic programmability, multi-physics coupling, and intelligent sensing. On the one hand, the integration of emerging platforms such as flexible materials, liquid crystals, and phase-change materials is anticipated to enable the dynamic reconfiguration of symmetric structures and real-time optical responses, thereby advancing the development of adaptive optics, real-time imaging, and reconfigurable optical computing. On the other hand, with continuous advances in fabrication technologies and theoretical modeling, symmetry control in metasurfaces is evolving from static geometric optimization toward dynamic programmability and intelligent design. Recently, Hou et al. [99] employed a target-conditioned generative network (TCGN) to achieve inverse design of all-silicon chiral metasurfaces in the terahertz regime, obtaining a CD value exceeding 36%. In addition, deep learning models such as convolutional neural networks (CNNs) [100] and variational autoencoders (VAEs) [101] have been widely adopted to learn the complex nonlinear mappings between metasurface geometries and their optical responses, enabling efficient forward prediction and inverse structural retrieval. These AI-driven approaches are increasingly becoming essential tools for the next generation of metasurface design.