© J. He et al., Published by EDP Sciences, 2026

1 Introduction

Micro-nano gratings are core optical components that modulate light waves through periodic micro/nanostructures, and they play pivotal roles in key fields such as spectroscopy, diffraction optics, and high-resolution imaging [1]. Conventional rigid-substrate gratings (e.g., glass, silicon, or metals) [2–4] offer excellent stability and optical performance, but their inherent physical rigidity limits applications in flexible electronics [5–7], wearable devices [8–10], and biomedical sensing [11], etc. This demands flexible gratings integrating mechanical adaptability (bending, stretching) and functional tunability (dynamic optical modulation). Flexible gratings serve as an interdisciplinary bridge connecting optics, materials science, and mechanics [12,13], combining the light-modulating ability with flexible materials (e.g., polydimethylsiloxane PDMS [14], liquid crystal composites [15], graphene nanocomposites [16,17]). They achieve two-level flexibility: (1) passive conformity to complex surfaces; (2) active tuning of period or refractive index [18] via external stimuli (e.g., light, electricity, heat, or mechanical force) [19–21]. Despite the promising application prospects of flexible micro-nano gratings in stretchable displays and implantable sensors [22–24], their industrial translation still faces unresolved bottlenecks—especially in the core application fields of wearable devices and implantable optical components.

Traditional microfabrication techniques, such as electron-beam lithography and laser direct writing, offer sub-10 nm precision but is costly, time-consuming, and incompatible with flexible substrates [25]. In contrast, surface mechanical instability-based patterning [26] leverages stress-driven self-organization (pre-straining + plasma treatment, thermal shrinkage of film-substrate bilayers) to form periodic structures (period 100 nm–10 µm) over large areas, reducing cost by 2–3 orders of magnitude vs. lithography—key for industrial translation. Foundational work by Bowden et al. [27] (1998) showed that compressive stress-induced buckling of metal films on PDMS generates regular micro/nano-patterns without expensive lithography, laying groundwork for flexible periodic structures. Chung et al. later highlighted surface instability's value in understanding soft material mechanics and enabling patterning, assembly, and property measurement at micro/nano-scales [28]. Recent researches have deeply explored the mechanical mechanisms of wrinkling instability in film/substrate systems and developed various methods to control pattern parameters such as period, amplitude, and orientation [29]. However, surface mechanical instability-based patterning does inot replace traditional lithography but complements it, addressing the limitations of lithography in flexible substrate compatibility and large-area, low-cost fabrication, thus expanding the application boundaries of micro-nano gratings.

In five years, research accelerated [30,31]. Fundamental work refined wrinkling models [32–34], and applied research enabled stretchable photonic devices [35,36], surface-enhanced Raman scattering (SERS) [37–41], and wearable sensors [22,23]. For wearable electronics, the large-scale manufacturing cost remains a critical barrier. Although surface mechanical instability-based patterning reduces cost by 2–3 orders of magnitude compared to lithography, the lack of long-range order in wrinkle patterns still requires additional post-processing (e.g., precise alignment) for device integration, which offsets part of the cost advantage. For implantable optical devices, biocompatibility is another unresolved issue—currently widely used PDMS substrates, while flexible, may induce chronic inflammatory responses when in long-term contact with human tissues, and the metal films (e.g., gold, silver) used in SERS or photonic devices further raise concerns about heavy metal ion release. These bottlenecks highlight the urgency of bridging the gap between laboratory-scale fabrication and industrial application, which is also a core focus of this review. We summarized flexible surface periodic micro/nanostructures (focus on gratings). Section 2 focuses on fabrication methods and their trade-offs in precision, scalability, and cost, laying the technical foundation for subsequent research; Section 3 explores structural control strategies to address the key challenge of precise regulation in flexible micro-nano structures; Section 4 highlights photonics, biosensing, and smart surface applications; Section 5 analyzes challenges and future directions. It aims to guide researchers and inspire flexible micro/nano-optics innovations.

2 Advanced fabrication strategies for periodic micro/nanostructures on flexible substrates

Periodic structures on flexible substrates usually originate from wavelike undulations arising from stress-induced instabilities at stiff/soft material interfaces. Researchers have developed a variety of approaches to induce and replicate such periodic micro/nanostructures on flexible surfaces, including oxygen plasma surface oxidation, inherent stress release (pre-strain methods), template molding replication, self-assembly, etc. This section will introduce several typical fabrication methods in these categories.

2.1 Strain-release induced wrinkling

The pre-strain method is one of the most classic and commonly used strategies to fabricate periodic structures on flexible surfaces. The basic principle is to first apply a certain mechanical strain (tensile or thermal expansion) to a flexible substrate (such as an elastomer), deposit or form a thin stiff film on the substrate while it is in the strained state, and then release the strain. Due to the differential contraction between the substrate and the thin film, compressive stress develops within the system. This stress is subsequently relieved through a buckling instability, resulting in the formation of a wave-like, periodic wrinkled structure. For example, Bowden et al. (1998) heated a PDMS substrate to thermally expand it, deposited a metal film on the expanded PDMS, and upon cooling, thermal mismatch stress between the metal and elastomer induced periodic instability and wrinkling once the film reached a critical compressive stress, large-area ordered ripple patterns were achieved [27] (Fig. 1a). In 2006, Khang's team reported a new method to fabricate stretchable single-crystal silicon electronics on elastic substrates, which allowed inherently brittle single-crystal silicon to withstand large stretching and compression (Fig. 1b). The researchers used a silicon-on-insulator (SOI) wafer and created silicon ribbon structures 20–320 nm thick via lithography and etching, then transferred them onto a pre-stretched PDMS elastic substrate. After releasing the pre-strain, the silicon ribbons together with the PDMS surface formed a highly periodic sinusoidal wave structure [42]. In 2024, Byun demonstrated a simple and efficient method to fabricate periodically ordered wrinkles by combining Controlled Evaporative Self-Assembly (CESA) with mechanically driven surface wrinkling [43] (Fig. 1c). In this approach, a “wedge-plane” gradient geometry was used: through solvent evaporation, gradient-thickness, gradient-width poly (methyl methacrylate PMMA) stripe patterns were created on a silicon substrate. These gradient stripes were then transferred onto a pre-stretched PDMS substrate, and upon releasing the strain, periodic wrinkles spontaneously formed due to the pre-strain release. Furthermore, by a second transfer and orthogonal stacking of the stripes, the wrinkles could be constrained to form within defined regions, resulting in hierarchical grid-like wrinkle patterns. This pre-strain method is not limited to PDMS substrates; it has also been applied to other flexible substrates such as polyimide (PI), enabling the fabrication of periodic structures on diverse flexible platforms. These demonstrate the flexibility and application potential of the method in micro/nanopatterning.

Fig. 1 (a) A fabrication method for periodic wrinkled metal films on PDMS [27]. (b) A method to fabricate stretchable single-crystal silicon electronics on elastic substrates [42]. (c) Hierarchical grid-like wrinkle patterns were fabricated from periodic wrinkle structures [43]. (d)Optical micrographs of the film after transfer to PDMS and application of strain to induce buckling [44]. (e) Schematic of the spontaneous isotropic wrinkling in oxidized PDMS films in air versus in inert atmosphere [45].

2.2 Oxygen plasma irradiation-induced surface hardening wrinkling

Using plasma oxidation to generate an ultrathin stiff skin on a flexible polymer surface is another common approach to wrinkle formation. The core of this technique is to deliberately create a “stiff shell–soft core” mechanical system and then trigger its instability by releasing pre-strain. The process can be divided into three key steps: 1. Pre-stretching: The PDMS substrate is stretched uniaxially by a certain strain( from 0% to 20%) to store elastic energy. Uniaxial stretching ensures that the wrinkles formed later will have a uniform orientation (perpendicular to the stretching direction). 2. Plasma surface modification: While maintaining the PDMS in the stretched state, expose it to a low-temperature plasma (commonly O₂, Ar, or air plasma). The cleaning power were fixed at 50 W .and the treatment time was varied from 0.5 to 10 min [46]. In this process, energetic reactive species bombard the PDMS surface, breaking Si–CH₃ bonds and causing oxidation and crosslinking that produces a thin, stiff silica-like glass layer (SiO_x) on the surface. This results in a classic bilayer system of a rigid thin film on a soft substrate. 3. Strain release and wrinkle self-assembly: After slowly releasing the pre-stretch, the soft PDMS substrate attempts to return to its original length and thus contracts, while the stiff SiO_x layer on the surface cannot contract accordingly. This mismatch induces a large uniaxial compressive stress in the surface layer. When the compressive stress exceeds a critical threshold, the flat surface becomes unstable. To minimize total energy, the system relieves the compression via buckling deformation, spontaneously forming periodic wavy wrinkles. The resulting wrinkle wavelength and amplitude depend on the thickness of the surface layer and the mechanical properties of both the surface layer and substrate, and they can be described by [40,44,47–49]:

$$\lambda =2\pi h{\left[\frac{\left(1-{{\nu }_s}^2\right)E_f}{3\left(1-{{\nu }_f}^2\right)E_s}\right]}^{\frac{1}{3}}$$(1)

$$A=h_f\sqrt{\frac{{}^{\epsilon }\mathrm{pr}e}{{\epsilon }_c}-1}$$(2)

where E_f and E_s denote the Young's modulus of the top stiff layer and the bottom substrate, respectively; ν_f and ν_s represent their Poisson's ratios; and h_f is the thickness of the stiff thin film. The critical strain ε_c can be theoretically expressed as [50]:

$${\epsilon }_c=\frac{1}{4}{\left[\frac{3\left(1-{{\nu }_f}^2\right)E_s}{\left(1-{{\nu }_s}^2\right)E_f}\right]}^{\frac{2}{3}}.$$(3)

Compared to the mechanical pre-strain method, the plasma oxidation method eliminates the need for pre-depositing a stiff film on the substrate and involves fewer fabrication steps, making it more widely adopted in practical applications. In 2004, Stafford and colleagues introduced an innovative measurement technique based on strain-induced elastic buckling instability (SIEBIMM) for fast and accurate determination of the elastic modulus of polymer thin films [44] (Fig. 1d). This method involves applying a pre-strain to a soft elastic substrate (e.g., PDMS) coated with a thin, stiff film. Subsequent release of the pre-strain generates compressive stress, inducing a buckling instability that produces a periodic wrinkled pattern on the surface.

However, wrinkled films prepared by this plasma oxidation method often suffer from random crack networks in the brittle surface layer, which compromise the integrity of the ordered wrinkle pattern [51,52]. Recent studies have proposed solutions to overcome this issue. Ahmad et al. (2024) found that increasing the thermal conductivity of the substrate during plasma oxidation can effectively suppress the intrinsic cracking in plasma-induced wrinkles, achieving crack-free continuous wavy patterns [53]. Furthermore, it was discovered that it is not thermal stress but rather ambient humidity that is the primary driver of wrinkling in the oxidized PDMS surface layer: water vapor absorption by the oxide skin causes swelling stress that leads to wrinkling. Ahmad et al. (2025) demonstrated this moisture-induced wrinkling mechanism by controlling the humidity environment after plasma treatment [45] (Fig. 1e).

2.3 Hierarchical wrinkling and reconfigurable structures

By applying multi-step loading and releasing of stress, one can create hierarchical (multi-level) wrinkle structures on the same surface. A typical approach is to first generate a primary set of wrinkles, then superimpose another set of wrinkles in a different direction or with a different wavelength. For example, performing two sequential perpendicular uniaxial stretch-and-release cycles on a substrate can produce a “wrinkles-on-wrinkles” double-wave structure. On a two-dimensional surface, sequentially controlling strains in different directions can yield more complex topographies such as grid-like networks or intersecting stripe patterns. Duan et al. (2017) reported a simple method to fabricate orthogonal grating structures on both sides of a PDMS substrate [54] (Figs. 2a and 2b). In their method, a PDMS film was first pre-stretched uniaxially, then treated with oxygen plasma on its surface to form a stiff SiO₂ layer; upon releasing the strain, periodic wrinkles formed spontaneously on that surface. The sample was then rotated 90°, and the same process was repeated on the opposite side of the PDMS, ultimately obtaining a double-sided grating with orthogonal orientations. In 2021, Knapp et al. controlled the formation of wrinkle “line defects”—a key structure for guiding light propagation in optical devices, and proposed a new method to precisely control line defects in PDMS surface wrinkles via patterned plasma treatment [55] (Fig. 2c). They used low-pressure plasma (H₂ or N₂) on selectively masked regions of a pre-stretched PDMS surface, employing masks to create combinations of treatment durations and gases. After strain release, this induced adjacent wrinkle regions with markedly different wavelengths (up to a 7-fold contrast), such that an abrupt wavelength change occurred at the boundaries, effectively pinning the wrinkle line defects at those borders. In 2021, Tan's group quantitatively revealed for the first time the critical influence of a tunable phase grating's gradient refractive index distribution on diffraction behavior [56] (Figs. 2d and 2e). Furthermore, by analyzing static light-scattering diffraction patterns, they inversely reconstructed the surface profiles and extended this method to the fabrication of two-dimensional multiaxial gratings.

Fig. 2 (a) Fabrication process of PDMS double optical grating. (b) One-dimensional diffraction spots were generated by the single-direction grating. Two-dimensional spot array was generated by the double cross optical grating [54]. (c) Schematic illustration of the masked wrinkling process, and detailed characterization of λ, amplitude, and branching degree [55]. (d) The SLS image of a 2D pattern with θ = 20° and 50°. (e) Schematic of 2D wrinkling pattern formation [56]. (f) Schematic illustration of the fabrication process of flexible PDMS films with surface microdome array and the corresponding pressure sensors [57]. (g) Self-organized preparation process of the PS microsphere nanoarray strain sensor [58].

2.4 Colloidal crystal self-assembly

By leveraging the self-organization of colloidal particles (e.g., SiO₂ or polystyrene microspheres) in a solution, large-area ordered two-dimensional (2D) or three-dimensional (3D) “colloidal crystal” structures can be constructed. For example, monodisperse microspheres can assemble at a liquid/air interface into a hexagonally close-packed monolayer and then be deposited onto a substrate to form a colloidal crystal film. The lattice constant of this colloidal monolayer is determined by the sphere diameter, typically ranging from sub-micron down to hundreds of nanometers. Such an array of microspheres essentially behaves as a photonic crystal, which can produce visible structural colors or serve as a grating (via lattice diffraction). Furthermore, by using the microsphere array as a template to deposit material and then removing the microspheres, one can obtain inverse structures such as etched micropit arrays or inverted pyramid textures, which are often used for antireflection or light-trapping purposes. Recently, colloidal self-assembly has been combined with transfer techniques to fabricate micro-dome structured PDMS flexible pressure sensors. Zhang et al. (2017) obtained a PDMS film with a convex microdome array through a two-step PDMS transfer process. Then a gold nanolayer was deposited on the microdome structures by sputtering, and two such functionalized PDMS films were assembled face-to-face (microdomes facing each other) to form a resistive pressure sensor [57] (Fig. 2f). In 2024, Wang et al. reported a simple fabrication of a PS microsphere nanoarray optical strain sensor based on a hemispherical-indentation-assisted self-assembly technique [58] (Fig. 2g). They first formed a large-area monolayer of 1000 nm diameter PS spheres on a hydrophilically treated silicon wafer via an air-liquid interface assembly, then transferred it onto a pre-fixed flexible PDMS substrate. After drying, this produced a periodic nanostructure with hexagonal symmetry on the PDMS. The sensor monitors strain in real time via changes in its reflection spectrum: as the nanostructure's spacing changes during stretching, a characteristic resonance peak shifts to shorter wavelength (blue shifts).

3 Multi-dimensional precise control mechanisms for periodic structures

3.1 Geometric regulation

3.1.1 Period control

In 2008, Tsougeni's team reported a simple method to produce oriented, ordered nanostructures on PDMS films and stamps via oxygen plasma treatment [59] (Fig. 3a). The study found that the periodicity (λ) of the formed nanostructures strongly depended on the size of the initial micropattern—the smaller the initial pattern dimensions, the smaller the induced nanowrinkle period—demonstrating the potential of using initial pattern scale to tune nanoscale morphology. In 2017, Zhang et al. reported a method to increase the grating line density of PDMS [60] (Fig. 3b). First, a micron-scale grating structure was replicated onto PDMS via soft lithography (e.g., microcontact printing) to create a negative mold. The PDMS was then stretched to compress the grating period along its direction, and finally, the pattern was transferred to a PMMA substrate through nanoimprint lithography. This process could be iterated to progressively reduce the grating period from the micron scale down to 820 nm. Ma et al. (2021) proposed a method of stretching a PDMS mold before imprinting and curing, which increased the line density of an original 1200 lines/mm grating by about 26%, thereby obtaining a structure with a smaller period [61] (Fig. 3c). In 2021, Li's team reported a low-cost, efficient method for fabricating variable line-space (VLS) gratings based on a PDMS self-assembly technique [26] (Fig. 3d). In this method, a PDMS substrate with a wedge-shaped thickness gradient was prepared, and by combining lateral or longitudinal stretching, oxygen plasma surface treatment, and strain release, a grating structure with continuously varying periodicity was spontaneously formed on the PDMS surface. Because the substrate thickness varied continuously (the wedge profile), the stress experienced at different positions was different upon release, thereby achieving a gradient distribution of grating periods across the substrate.

Fig. 3 (a) AFM 2D images of a PDMS stamp after exposure to O2 plasma for 10 min: structures formed on the top of the stamp [59]. (b) Fabrication process flow [60]. (c) The fabrication process of original grating, PDMS grating and re-replication [61]. (d) Preparation of PDMS VLS grating [26]. (e) Reversible wrinkling/dewrinkling process of temperature-responsive surface wrinkles [45].

3.1.2 Amplitude control

The amplitude of wrinkles (peak-to-valley height) affects their surface properties, such as contact angle hysteresis and optical scattering intensity. Wrinkle amplitude is primarily determined by the magnitude of the applied strain or stress: the greater the strain released, the more the thin film “buckles out”, and the higher the wrinkle peaks. Classical models predict that the wrinkle amplitude increases as the applied strain beyond the critical value increases, with amplitude growing continuously from zero as the instability threshold is reached. Experimental observations align with this trend. For example, in a humidity-induced plasma-oxidation wrinkling system, increasing the ambient humidity causes greater swelling of the oxide layer and thus significantly increases the wrinkle height, while the wavelength remains essentially unchanged [45] (Fig. 3e). For practical applications, there is often interest in dynamically and reversibly tuning the wrinkle amplitude to achieve switchable functionalities—such as switching between a diffusely reflecting state and a transparent window [28]. In recent years, fast dual-stimuli-responsive wrinkles (for example, responsive to both temperature and solvent) have been developed as efforts toward this goal [62]. Wrinkle amplitude is an important tunable parameter for creating tunable optical devices and switchable interfaces. Relevant control can be achieved through three main approaches: strain modulation (adjusting the pre-release strain magnitude), swelling stimuli (e.g., humidity or solvent-induced swelling of the surface layer), and responsive material design (using temperature- or pH-sensitive polymers).

3.1.3 Pattern directionality control

Unconstrained wrinkles typically appear as isotropic random wave patterns or island-like herringbone motifs (especially under equibiaxial compressive stress). However, many applications require highly ordered, unidirectionally aligned patterns—uniaxial pre-strain is the simplest method: if a substrate is stretched in one direction and then released, the wrinkles tend to form as parallel ripples perpendicular to the stretching direction. Such uniaxially aligned wrinkles usually have a very uniform orientation and can function as diffraction gratings. For more complex 2D patterns like intersecting networks or orthogonal grids, a two-step orthogonal stretch-and-release sequence can be employed: a first uniaxial strain produces parallel wrinkles, and after rotating the substrate, a second stretch and release overlays a new set of wrinkles in another direction, forming a checkerboard or herringbone-like biaxial pattern [56]. Mu et al. (2025) used this strategy to obtain one-dimensional and two-dimensional controllable flexible grating structures on PDMS, where the 2D crossed wrinkles served as anti-reflection microstructures for optical devices [46] (Fig. 4a). Clearly, through stress field engineering (whether uniaxial, biaxial, or custom-designed distributions) and the use of pre-patterned templates, one can effectively control the orientation of wrinkles—transforming them from random to ordered patterns—to achieve the desired aligned structures.

Fig. 4 (a) Wrinkle morphologies and light diffraction patterns induced by biaxial loading [46]. (b) Thermal response of the ${{\displaystyle \hat{n}}}_{\parallel }$ and ${{\displaystyle \hat{n}}}_{\perp }$ μLCE (the liquid crystal elastomers patterned with micro-scale surface relief gratings) via optical diffraction measurements [63]. (c) Temperature response of PNIPAM-AA-Al hydrogel film [64].

3.2 Dynamic modulation

3.2.1 Thermoresponsive modulation

Flexible periodic structures based on thermally induced deformable materials can undergo reversible morphological evolution in response to temperature changes, demonstrating significant potential for smart optical devices, sensors, and adaptive systems. Moorhouse et al. developed a method for fabricating submicron diffractive optical elements based on the inherent anisotropic deswelling of liquid crystal elastomers (LCEs) [63] (Fig. 4b). The period of this LCE grating could contract to 707 nm, exhibiting dual tuning capabilities via both thermal and mechanical responses. Wen et al. utilized a CD-R disc template replication technique to fabricate one-dimensional periodic grating structures on the surface of poly(N-isopropylacrylamide) (PNIPAM) hydrogel [64] (Fig. 4c). Regarding thermoresponse, leveraging the phase transition property of PNIPAM around 35°C, the grating period decreased with increasing temperature, leading to a blue-shifted in the diffraction wavelength of visible light—specifically, the structural color changed from orange to green. When the temperature exceeded the critical transition point, the hydrophilic-hydrophobic transition induced aqueous phase separation, enhanced light scattering, masked the structural color, and turned the film into a milky, opaque state. Ma et al. developed a highly sensitive micro-strain sensing technique based on wrinkled PDMS/Au gratings [65]. By adhering the PDMS/Au gratings to different substrates (Cu, Si), their coefficients of thermal expansion (CTE) could be measured.

3.2.2 Solvent response

The interaction between solvent molecules and PDMS primarily induces changes in optical properties through physical diffusion causing material swelling or refractive index variation. For instance, the aforementioned study by Wen et al. also noted that when ethanol (refractive index 1.36) filled the grating grooves, the grating structure became optically invisible and the film turned transparent due to the close match between the refractive index of ethanol and that of the hydrogel (1.48); the structural color reappeared after solvent evaporation [64]. Similarly, when water acts as the solvent [66], PDMS-based films achieve water-induced chromism via Mie scattering effects, exhibiting strong scattering in the dry state and transparency in the wet state, with a differential transmittance change reaching 44.93% (Fig. 5a). In 2024, Hernik's team recorded a sinusoidal surface relief on a photopolymer using holographic technique and then replicated it into PDMS to fabricate gratings with depths ranging from 120 to 530 nm [22]. Deeper gratings showed higher sensitivity to VOCs; the diffraction efficiency change reached 0.44% upon toluene exposure, with a sensitivity of 0.017 µW/ppm and a detection limit as low as 186 ppm. The response of photonic crystal structures manifests as a photonic bandgap (PBG) shift (Fig. 5b). Solvent penetration leads to an increase in lattice spacing, causing a redshift in the reflection spectrum. For example, the bandgap shift of PDMS-based photonic crystals in aromatic hydrocarbon vapors can exceed 100 nm [67]. The advantages of these sensors include low cost, ease of integration, and real-time response. However, challenges remain in selectivity and long-term stability (e.g., structural degradation caused by aromatic hydrocarbons). Future efforts could enhance selectivity through functionalized coatings.

Fig. 5 (a) Structural color change of the best-optimized hydrochromic adhesive film [66]. (b) A mechanism of detecting hydrocarbons with a sensor based on a 3D PhC [67]. (c) Schematics of SMP grating under different stretching conditions [68]. (d) The evolution of structural colors at a viewing angle of 27° during uniaxial tension along direction A and recovery [69].

3.2.3 Memory effect

As a type of smart optical device, flexible gratings have recently achieved reversible tuning through the introduction of shape memory polymers (SMPs), exhibiting remarkable dynamic performance, particularly under heating/cooling stimuli. Shape memory polymers possess the ability to be deformed at high temperatures, fixed in a temporary shape upon cooling, and recover their original shape upon reheating, providing a basis for grating period control. In 2022, Sun et al. fabricated nanogratings on an SMP surface using a dual-beam interference method, where the initial period could be precisely controlled by the interference angle [68]. When stretched parallel or perpendicular to the grating direction, the grating period could be continuously tuned, and recovery triggered by heating to 80°C restored the original period, overcoming fabrication precision limits (Fig. 5c). In the same year, Zhao et al. constructed one-dimensional nanogratings on PLLA films using a template method and induced changes in the microscopic structural period via macroscopic stretching [69]. Heating and stretching caused the structural color to shift from red to blue-green; after cooling fixed the temporary shape, reheating completely restored the initial red color (Fig. 5d). The memory effect in flexible gratings enables reversible and precise control of grating parameters through the temperature-controlled deformation of SMPs, offering new avenues for dynamic optics, sensing, and anti-counterfeiting. Future research may focus on integrating multi-stimuli responsiveness and optimizing large-scale fabrication processes.

4 Multi-functional applications of flexible subwavelength gratings

4.1 Micro-optics and tunable photonic devices

Periodic structures are inherently diffraction gratings and can be used to control the propagation and reflection of light and to produce color. Periodic wrinkles on flexible substrates, by virtue of being stretchable, impart tunability to optical devices. For example, a regular wrinkled PDMS surface can serve as a stretchable grating; by stretching it to change its period, one can tune the diffraction angle or optical phase. Ma et al. (2013) employed a metal film/PDMS wrinkle structure as a tunable grating for optical measurement of small strains, demonstrating the sensitive tunability of wrinkles as optical elements [48] (Fig. 6a). In the area of structural colors, flexible wrinkles provide a new route to achieve color variability—unlike traditional lithography-based structural color fabrication, which is limited to rigid substrates, flexible wrinkles enable strain-tunable color changes on deformable platforms. Tan et al. (2022) reported that by introducing gradient and multiaxial wrinkles on PDMS, they could obtain angle-dependent structural color patterns, and under mechanical stretching the colors changed reversibly, achieving a strain-responsive color-changing device [19]. Such mechano-tunable coloration surfaces hold promise for stress sensing and dynamic display applications. Meanwhile, flexible wrinkled gratings can conform to irregular surfaces as bendable diffractive elements for imaging, spectroscopy, and other optical analyses [46]. In optical applications, flexible periodic wrinkles offer deformability and tunability that are difficult to achieve with traditional rigid gratings, enabling novel photonic devices such as rollable displays, wearable optical sensors, and adaptive optical components.

Fig. 6 (a) Optical setup for micro-strain measurement [48]. (b) Images of light spot position under different forces: F = 5N, F = 7N, F = 10N [70]. (c) Flexible capacitance-based pressure sensor [72]. (d) Structure and photograph of the organophotovoltaic (OPV) [13]. (e) The fabrication process of gold (Au)-grating using oblique angle deposition (OAD) with e-beam evaporation, and cross-section profiling using focused ion beam (FIB) in a SEM [73]. (f) Schematic diagram for the fabrication of the 3D flexible SERS substrate [74].

4.2 Strain and pressure sensors

Periodic wrinkle structures, due to their unique stress–strain response, have been widely used in flexible sensing and stretchable electronic devices. Strain sensing is one of the most direct applications: when the substrate is stretched, the wrinkles flatten out; when released, the wrinkles revert. By monitoring this morphology change, one can determine the amount of strain. Jin et al. (2020) proposed a mechanical measurement method based on a flexible PDMS grating: they attached a PDMS wrinkled grating onto the structure to be measured, and determined small strains or forces by optically reading the change in the wrinkle spacing [70] (Fig. 6b). Compared to a conventional metal foil strain gauge, the wrinkled grating strain sensor is highly compliant and can conform to curved surfaces, and its optical readout avoids interference from electrical noise. Similarly, wrinkle structures can be used for pressure sensing: when pressure is applied to an elastic dielectric with surface wrinkles, the wrinkles are pressed flat, changing the thickness of the dielectric layer and thus its capacitance [71]. This principle can be employed to create a high-sensitivity flexible capacitive pressure sensor [72] (Fig. 6c).

4.3 Biomedical monitoring

In biomedical sensing, wrinkle structures—owing to their wavy morphology resembling extracellular matrix—can be used to fabricate flexible biosensing patches for highly sensitive detection of vital signs such as pulse and respiration [13] (Fig. 6d). At the same time, the wrinkle pattern helps mitigate mechanical mismatch when the sensor patch stretches with the skin, improving wearer comfort and signal stability. In a separate study, a flexible and stretchable photonic crystal sensor based on a TiO₂/PDMS structure was demonstrated for biosensing and tactile sensing applications [75]. This sensor exhibits high sensitivity and represents a potentially cost-effective solution for these applications. In essence, the value of flexible wrinkles in sensor applications lies in the coupling of structural deformability with device function: the wrinkles provide a strain-accommodating buffer and a mechanism for signal transduction, allowing devices to operate stably under large deformations while still producing readable signals. A 2025 study fabricated metal/organic hybrid electrodes on PDMS substrates, providing new achievements for wearable electronic products [76]. This capability is crucial for the development of the next generation of wearable electronics.

4.4 SERS substrate

Periodic micro/nanostructures can significantly enhance the localized field effect for Raman scattering. When metal nanoparticles or nanostructures are assembled on periodic wrinkle surfaces, the density of “hot spots” between adjacent metallic structures increases, leading to a remarkable enhancement of the Raman signal from adsorbed molecules. Flexible wrinkles, owing to their large area, bendability, and ease of metal deposition, have become an ideal platform for fabricating high-performance SERS substrates [77]. For instance, Chen et al. (2019) self-assembled a uniform layer of gold nanoparticles on a PDMS wrinkle substrate to construct a flexible SERS platform for trace-molecule detection [40]. The accumulation of nanoparticles within the wrinkle grooves induces strong localized surface plasmon coupling, resulting in a substantial increase in the Raman enhancement factor (typically ranging from 10⁴ to 10⁶). Such substrates not only exhibit high sensitivity but can also conform to irregular surfaces for in situ detection. In another study, Ghosh et al. deposited titanium (5 nm) and gold (50 nm) layers via electron beam evaporation at different angles (0° for surface plasmon polariton sensors and 65° for localized surface plasmon resonance sensors) on a periodic wrinkled PDMS grating structure, achieving either continuous or discrete metal gratings [73] (Fig. 6e). Its anisotropic grating structure further supports multi-mode plasmon excitation, improving signal selectivity.

In 2024, Li et al. developed a flexible SERS substrate using a layer-by-layer transfer technique, assembling gold nanospheres on a PDMS-supported polystyrene microsphere array, which achieved high reproducibility and tunable SERS enhancement [74] (Fig. 6f). The same year, Mi's group introduced a simple low-cost approach using polystyrene microspheres as a sacrificial template. Combining plasma etching for precise interparticle gap control and thermal demolding, they fabricated a large-area bowl-shaped gold nanoparticle array on PDMS as a flexible SERS substrate [78].

Flexible periodic wrinkles offer a new design strategy for SERS substrates: by using their tunable periodic structures to organize metal nano-units, high enhancement factors can be achieved while retaining mechanical flexibility. This meets the demand for low detection limits and in-field portable detection in chemical and biological sensing.

4.5 Superhydrophobic/self-cleaning surfaces

In recent years, significant progress has been made in the development of flexible superhydrophobic surfaces, particularly demonstrating great potential in biomimetic structural design and functional applications [79,80]. Inspired by the microstructures of natural surfaces such as lotus leaves and water strider legs, Hou et al. achieved a remarkable enhancement in superhydrophobic performance by constructing hierarchical micro/nano structures [81] (Fig. 7a). On the other hand, the development of fluorine-free flexible superhydrophobic surfaces offers new avenues for practical applications. Lu et al. formed a maze-like wrinkled structure by combining a zinc film with a PDMS substrate, achieving a high contact angle of 168.5° and an ultralow sliding angle close to 0°, along with excellent mechanical durability [82] (Fig. 7b).

The central challenge for flexible superhydrophobic surfaces lies in balancing mechanical durability, environmental adaptability, and functional stability. The combination of bioinspired multi-scale structures and flexible substrates provides an effective solution to this challenge. Future research may further explore dynamic responsive structures and smart wettability regulation, expanding their application boundaries in flexible electronics, aerospace, and biomedical fields.

Fig. 7 (a) Droplet impacts on the sub-mm, micro, and sub-mm/micro surfaces [81]. (b) The droplet profiles on the wrinkled Zn/PDMS surfaces with different Zn film thicknesses [82]. (c) TENG connected to a full-wave bridge rectifier to get rectified output and the illustration of the device and circuit on a breadboard illuminating an LED [83]. (d) Power density of PDMS/MXene TENG (6W) and PDMS TENG (0N) as a function of load resistance [84].

4.6 Triboelectric nanogenerator

Periodically ordered structures formed via self-assembly or template-induced patterning can effectively enhance the interfacial charge transfer efficiency of triboelectric layers, while functional fillers such as MXene further improve electrical conductivity and polarization response. Kumar et al. fabricated a nanograting-patterned PDMS layer using a DVD-R disc as a template to construct a flexible metal–dielectric triboelectric nanogenerator (TENG) [83] (Fig. 7c). Their results demonstrated that the surface nanogratings significantly increased the effective contact area and charge density, leading to a 97% enhancement in open-circuit voltage compared with flat PDMS. In contrast, Aiswarya et al. introduced micro-wrinkled structures into PDMS/MXene composite films using a sandpaper template [84] (Fig. 7d). The incorporation of highly conductive MXene enhanced the film's electronegativity, resulting in a substantial performance improvement, with an open-circuit voltage of 790 V, a short-circuit current of 140 µA, and a power density of 17.1 W · m⁻²—a value that is approximately 3–5 times higher than that of flat PDMS/MXene composite film-based TENGs.

5 Summary and outlook

Despite the significant progress in research on periodic micro/nanostructures on flexible material surfaces, numerous challenges remain in scaling up for practical applications and achieving higher-level functionalities.

First, in terms of precise controllability, although various methods exist to tune wrinkle period and orientation, it remains challenging to achieve the same level of fine positional and shape control at the nanoscale as that offered by lithography. Currently, most wrinkle patterns lack the long-range order of conventional lithographic patterns. Current research has successfully implemented guided self-assembly to achieve more ordered templates, which may serve as a viable solution for future long-range ordered structures [85]. In the future, more controllable self-assembly strategies need to be developed—for example, deliberately inducing stress at specific locations in order to obtain programmable wrinkle patterns.

Second, regarding materials and durability—a key constraint for wearable and implantable applications, current research is still mainly focused on silicone elastomers like PDMS. While PDMS is convenient to use, its mechanical properties (such as creep and aging) may affect the long-term stability of wrinkle structures in wearable devices (e.g., repeated stretching in daily use can lead to wrinkle amplitude attenuation). More critically, PDMS lacks inherent biocompatibility: in implantable optical devices (e.g., intraocular pressure monitoring gratings), its surface may trigger protein adsorption and cell adhesion, leading to device failure or tissue inflammation. Although some studies have attempted to change the substrate material or chemically functionalize the PDMS surface to introduce active functional groups [86]. In practical applications, one must also consider environmental factors (illumination, humidity, temperature) on wrinkle morphology to prevent performance degradation due to aging or external disturbances. Therefore, future material selection and structural design should emphasize stability and durability—for example, developing fatigue-resistant polymer substrates or self-healing wrinkle materials.

Third, in device integration and large-scale manufacturing—a bottleneck for wearable device commercialization, incorporating micro/nano wrinkle structures into real devices (such as sensor arrays or optical components) requires addressing compatibility with existing electronic process flows. For example, wearable sensor arrays need to integrate flexible gratings with printed electrodes and signal processing modules, but the current “stretch-release wrinkling” process is difficult to align with roll-to-roll printing (a mainstream large-scale manufacturing technology for flexible electronics), resulting in low production efficiency and high unit cost. Additionally, the lack of long-range order in wrinkle patterns increases the difficulty of batch calibration (e.g., consistent diffraction efficiency across a 100 mm × 100 mm wearable display panel), further hindering industrial adoption. Due to the uncontrollability of wrinkles' self-assembly, it can be attempted to reverse them into templates to replicate stable periodic structures, thereby achieving standardized production. Recently, a study has successfully achieved the repeated printing of nanoscale stripes by using self-assembled PDMS gratings as a template [87]. Future research should focus on developing manufacturing techniques that are scalable—for instance, combining plasma-induced wrinkling (for grating fabrication) with printed electronics (e.g., inkjet printing of electrodes) to enable batch fabrication of flexible, optical, and electronic integrated devices.

It is foreseeable that with improvements in fabrication techniques and interdisciplinary collaboration, more breakthrough results will emerge in this field, propelling the development of flexible devices and smart material technologies. To address the bottlenecks of wearable and implantable applications: (1) For biocompatibility, develop novel flexible substrates (e.g., biodegradable poly(lactic-co-glycolic acid) (PLGA)-based elastomers) and non-toxic metal alternatives (e.g., biocompatible titanium nitride nanostructures) to replace PDMS and traditional noble metals ; (2) For large-scale manufacturing, integrate surface mechanical instability-based patterning with roll-to-roll processes—for example, designing continuous pre-stretching and plasma treatment modules compatible with roll-to-roll lines to achieve batch fabrication of wrinkle gratings with uniform long-range order ; (3) For durability, explore self-healing mechanisms (e.g., dynamic covalent bonds in elastomers) to enable automatic recovery of wrinkle structures after repeated stretching, meeting the long-term use requirements of wearable devices.

Funding

This research was funded by the National Natural Science Foundation of China (12374398), Key Project of the National Natural Science Foundation of China (52532007), National Key Research and Development Program of China (2022YFB3806003), Beijing Natural Science Foundation (4222081), and Fundamental Research Funds for the Central Universities.

Conflicts of interest

The authors declare no conflict of interest.

Data availability statement

No new data were generated or analyzed in this study. As a review article, all discussed information and data are derived from previously published literature, which is cited throughout the text.

Author contribution statement

J.H conceived the idea and wrote the manuscript. G.D supervised the project and participated in revision. Y.Q provided grammatical guidance. Z.Z assisted in data curation. X.S participated in literature analysis. W.T contributed to data collection. All authors edited the article.

References

All Figures

Fig. 1 (a) A fabrication method for periodic wrinkled metal films on PDMS [27]. (b) A method to fabricate stretchable single-crystal silicon electronics on elastic substrates [42]. (c) Hierarchical grid-like wrinkle patterns were fabricated from periodic wrinkle structures [43]. (d)Optical micrographs of the film after transfer to PDMS and application of strain to induce buckling [44]. (e) Schematic of the spontaneous isotropic wrinkling in oxidized PDMS films in air versus in inert atmosphere [45].

In the text

Fig. 2 (a) Fabrication process of PDMS double optical grating. (b) One-dimensional diffraction spots were generated by the single-direction grating. Two-dimensional spot array was generated by the double cross optical grating [54]. (c) Schematic illustration of the masked wrinkling process, and detailed characterization of λ, amplitude, and branching degree [55]. (d) The SLS image of a 2D pattern with θ = 20° and 50°. (e) Schematic of 2D wrinkling pattern formation [56]. (f) Schematic illustration of the fabrication process of flexible PDMS films with surface microdome array and the corresponding pressure sensors [57]. (g) Self-organized preparation process of the PS microsphere nanoarray strain sensor [58].

In the text

	Fig. 3 (a) AFM 2D images of a PDMS stamp after exposure to O2 plasma for 10 min: structures formed on the top of the stamp [59]. (b) Fabrication process flow [60]. (c) The fabrication process of original grating, PDMS grating and re-replication [61]. (d) Preparation of PDMS VLS grating [26]. (e) Reversible wrinkling/dewrinkling process of temperature-responsive surface wrinkles [45].
In the text

	Fig. 4 (a) Wrinkle morphologies and light diffraction patterns induced by biaxial loading [46]. (b) Thermal response of the ${{\displaystyle \hat{n}}}_{\parallel }$ and ${{\displaystyle \hat{n}}}_{\perp }$ μLCE (the liquid crystal elastomers patterned with micro-scale surface relief gratings) via optical diffraction measurements [63]. (c) Temperature response of PNIPAM-AA-Al hydrogel film [64].
In the text

	Fig. 5 (a) Structural color change of the best-optimized hydrochromic adhesive film [66]. (b) A mechanism of detecting hydrocarbons with a sensor based on a 3D PhC [67]. (c) Schematics of SMP grating under different stretching conditions [68]. (d) The evolution of structural colors at a viewing angle of 27° during uniaxial tension along direction A and recovery [69].
In the text

Fig. 6 (a) Optical setup for micro-strain measurement [48]. (b) Images of light spot position under different forces: F = 5N, F = 7N, F = 10N [70]. (c) Flexible capacitance-based pressure sensor [72]. (d) Structure and photograph of the organophotovoltaic (OPV) [13]. (e) The fabrication process of gold (Au)-grating using oblique angle deposition (OAD) with e-beam evaporation, and cross-section profiling using focused ion beam (FIB) in a SEM [73]. (f) Schematic diagram for the fabrication of the 3D flexible SERS substrate [74].

In the text

Fig. 7 (a) Droplet impacts on the sub-mm, micro, and sub-mm/micro surfaces [81]. (b) The droplet profiles on the wrinkled Zn/PDMS surfaces with different Zn film thicknesses [82]. (c) TENG connected to a full-wave bridge rectifier to get rectified output and the illustration of the device and circuit on a breadboard illuminating an LED [83]. (d) Power density of PDMS/MXene TENG (6W) and PDMS TENG (0N) as a function of load resistance [84].

In the text