Introduction

Mechanoluminescence is the direct emission of light in response to mechanical stimuli, such as impact, stretching, bending, or scratching, without requiring external excitation sources1,2,3,4. Unlike fluorescence, phosphorescence, and chemiluminescence, ML directly converts mechanical energy into light, offering advantages such as rapid response, low power consumption, and simplicity5,6,7,8,9,10. These features make ML materials suitable for applications in anti-counterfeiting, dynamic lighting, stress monitoring, electronic skin, and human-machine interfaces11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31. However, bottlenecks of low luminescent efficiency, insufficient photon utilization, and limited visibility under natural or low-stress conditions remain among the main challenges for ML materials. These limitations significantly hinder performance in high-precision displays, dynamic patterning, and practical deployment, where both brightness and signal reliability are essential.

To address these limitations, researchers have been exploring optimizations through chemical composition32,33, crystal structure34,35, and surface morphology36,37, alongside the introduction of piezoelectric and electrostatic effects to enhance photon emission38,39,40,41,42,43,44. Although piezoelectric and electrostatic effects can improve carrier generation and recombination, their efficacy diminishes under weak or non-uniform mechanical stresses. In addition, intrinsic bottlenecks in stress transfer efficiency persist, particularly in flexible matrix systems such as PDMS, where non-uniform stress distribution limits the effective activation of luminescent centers45,46. Photon losses at material interfaces and rapid luminescence decay exacerbate these challenges, limiting ML performance under working conditions. Thus, substantial enhancements in both mechanical stress transfer mechanisms and photon extraction efficiency are urgently required to optimize the performance of ML materials.

Here, we introduce a hierarchical multiscale structuring strategy at the ML material-PDMS interface to address key limitations in mechanoluminescent materials and achieve improved performance. By engineering microscale convex and concave patterns combined with nanoscale features, the hierarchical structures optimize stress distribution, enhance light-extraction efficiency, and improve light–matter coupling, resulting in a luminescent intensity of up to 366% of that of the bare ML film. This synergistic design localizes mechanical stress and simultaneously facilitates photon escape by enhancing multiscale scattering and reducing total internal reflection, thereby enabling spatially programmable mechanoluminescence. The programmed three-dimensional multiscale geometries allow precise stress localization and guide dynamic-contrast ML patterns emitting at 521 nm. Pre-encoded wafer-scale ML patterns, such as an image of a giant panda with 281,250 pixels at a resolution of 637 PPI, demonstrate the capability for stress-based readout mechanisms and pre-programmed dynamic luminescence. This approach provides quantitatively enhanced ML performance and supports applications in advanced visual stress sensors, dynamic displays, and force-responsive media, while ensuring minimal energy consumption and enabling diverse applications in intelligent sensing and multifunctional electronics.

Results

Programmable multiscale structures enhance ML

The hierarchical multiscale structures developed in this study combine macro-, micro-, and nanoscale design principles to modulate stress distribution and amplify light-matter interaction. On the macro scale, enabled by the hierarchical structures, millimeter-scale two-dimensional mechanoluminescent patterns can be embedded in an elastic thin-film without changing the mechanoluminescent material composition (Fig. 1a). The mechanism of ML pattern display is a combination of in-film stress concentration and enhanced light extraction. At the microscale, distinct arrays of convex and concave elements are patterned to redistribute mechanical stress within the thin-film (Fig. 1b). The nanoscale circular honeycomb arrays (2 μm in diameter, 320 nm in height, and 230 nm spacing) contribute to improved optical scattering and higher photon escape efficiency (Fig. 1c).

Fig. 1: Hierarchically encoded pattern of a giant panda using multiscale structures.
Fig. 1: Hierarchically encoded pattern of a giant panda using multiscale structures.The alternative text for this image may have been generated using AI.
Full size image

a Schematic and characterization of the 3D pattern of a giant panda with hierarchical multiscale structures. b Design of hierarchical structures integrating multiscale features, with a 3D surface topography of microstructures. c SEM and 3D optical topography of the nanostructure; height (µm) color-coded. d ML intensity comparison of mechanoluminescent films: planar, microstructured, and hierarchical multiscale-structured films. e Luminescent image of a giant panda generated by the multiscale film under mechanical excitation; ML intensity in 103 arb.u. f Fabrication of hierarchical multiscale-structured mechanoluminescent films: template preparation using SU-8 photolithography and UV nanoimprint lithography, followed by soft lithography to produce mechanoluminescent films. Scale bars, 2 µm (c).

This hierarchical design achieves two key objectives: (1) about a threefold enhancement in luminescence intensity compared to planar films, owing to stress concentration at structural interfaces and improved photon extraction efficiency (Fig. 1d); and (2) the programmability of ML light-emitting patterns, enabling dynamic stress visualization through a responsive three-dimensional (3D) surface relief pattern under mechanical stimuli (Fig. 1e).

To achieve such precision, a three-step fabrication pipeline was developed, combining SU-8 photolithography, nanoimprint lithography (NIL), and soft lithography (Fig. 1f and Supplementary Figs. 5–8). This method ensures reproducibility across scales and offers tunability in key structural parameters, such as height, diameter, and spacing.

Local stress mapping via luminescent encoding

The ML film consists of PDMS and ZnS:Cu particles, with a particle diameter of ≈ 17 μm (Supplementary Fig. 12). When the ML film is subjected to compression or tension (Fig. 2a), the relative motion between the ML particles and PDMS induces contact-separation and lateral sliding (Fig. 2b). This motion generates an electric field that excites electrons in the ZnS:Cu particles. The excited electrons transition to a higher energy state and emit light as they return to their ground state (Fig. 2c). The PDMS matrix surrounding the ZnS:Cu particles plays a crucial role in facilitating mechanical deformation and enabling relative motion between particles, which enhances stress localization and further promotes luminescence.

Fig. 2: Factors influencing the luminescent intensity of mechanoluminescent films.
Fig. 2: Factors influencing the luminescent intensity of mechanoluminescent films.The alternative text for this image may have been generated using AI.
Full size image

a Schematic of ML film deformation under tensile and compressive stress. b Force directions acting on polymer chains and mechanoluminescent particles under tension and compression. c Schematic of pore gap formation and the working process of ZnS:Cu/polymer ML composites under mechanical deformation, where q⁺ and q⁻ represent positive and negative triboelectric charges generated at the ZnS:Cu–polymer interfaces. d Effect of the ML-to-PDMS mixing ratio (ZnS:Cu to PDMS, w/w) on luminescent intensity. e Impact of the PDMS-to-curing agent ratio on film luminescent intensity. f 3D plots showing the relationship between maximum strain and Young’s modulus in films with varying PDMS-to-ML material and PDMS-to-curing agent ratios. g Stress birefringence images of PDMS films with convex and concave macroscopic structures under mechanical stress. h Luminescent patterns of mechanoluminescent films with convex and concave macroscopic structures under mechanical excitation. Scale bars, 2 mm (g,h). Source data are provided as a Source Data file.

The optimization of ML intensity is closely tied to the mechanical properties of the PDMS-ZnS:Cu composite, specifically its Young’s modulus. To systematically evaluate this, the ZnS:Cu content and PDMS-to-curing agent ratio were varied to balance luminescence performance and mechanical durability. Experimental results showed that increasing the ZnS:Cu content enhances luminescence intensity; with an optimal mixing ratio of 6:4 (ZnS:Cu: PDMS), the particle dispersion and film uniformity are maximized (Fig. 2d). Beyond this threshold, excessive viscosity introduces defects, resulting in uneven film thickness and luminescence non-uniformity. Similarly, the PDMS-to-curing agent ratio optimization indicated that, with a ratio of 20:1, relatively high luminescent intensity and strain tolerance can be realized (Fig. 2e). Under these optimized conditions, the ML film achieved a maximum strain of 146.38% while maintaining a relatively high Young’s modulus (Fig. 2f).

To further elucidate the interaction between structural design and ML intensity, finite-element analysis was performed to model the stress distribution in columnar and groove geometries (Supplementary Fig. 15). The results reveal that surface topography plays a pivotal role in modulating stress and luminescence. The elevated convex structures concentrate stress at their apex, resulting in localized luminescence enhancement, while recessed groove regions distribute stress along edges, producing lower luminescence intensity (Fig. 2g, h). As a practical demonstration, the letters “SJTU” (abbreviation of Shanghai Jiao Tong University) were patterned onto the ML film using arrays of pillars (“S” and “J”) and grooves (“T” and “U”) (Supplementary Fig. 16). Stress-induced birefringence mapping clearly indicated brighter regions corresponding to high residual stress areas, which correlated with enhanced luminescence (Fig. 2g, h). Quantitative analysis (Supplementary Fig. 17) confirmed that higher stress regions consistently exhibited greater luminescent intensity, providing direct experimental validation of the stress-luminescence relationship.

These findings demonstrate that macroscale surface structuring—through convex (pillar) and concave (groove) geometries—provides an essential framework for manipulating local stress distribution and establishing the basis for subsequent multiscale enhancements. At the macroscale, raised pillar and recessed groove architectures redistribute mechanical stress across the film, guiding and amplifying stress localization at finer scales. Raised pillars are particularly effective at concentrating stress, resulting in regions of enhanced luminescence, while recessed grooves lead to broader stress distribution and thus lower luminescence contrast. This initial macroscale modulation of mechanical stress establishes a foundation for precise control of mechanoluminescent output. By enabling both localized stress concentration and broader mechanical responsiveness, macrostructuring plays a critical role in optimizing overall mechanoluminescent performance.

Microstructure-driven mechanoluminescence control

Figure 3 explores the influence of microstructure parameters—shape, feature size, and the perimeter-to-area ratio (P/A, µm⁻¹)—on the ML intensity of mechanoluminescent films. These parameters are critical for controlling the luminescent performance of ML films. Microstructure arrays with feature sizes of 40, 60, 80, and 100 µm (Fig. 3a and Supplementary Figs. 18–20) were fabricated using high-aspect-ratio SU-8 photolithography and soft lithography, with a uniform structural height of 90 µm (Fig. 3b). Spectral analysis at the emission peak (521 nm) revealed significant variations in ML intensity, demonstrating that microstructure design enables controllable ML contrast.

Fig. 3: Influence of microstructure arrays on ML intensity in mechanoluminescent films.
Fig. 3: Influence of microstructure arrays on ML intensity in mechanoluminescent films.The alternative text for this image may have been generated using AI.
Full size image

a SEM images of microstructure arrays (feature size 80 µm) with perimeter-to-area ratios (P/A, µm⁻¹) of 22, 12, and 5. b 3D optical microscope image of 80 µm microstructure arrays. c Dependence of ML intensity at the emission peak (521 nm) on P/A for films with concave and convex arrays (40 and 100 µm). d ML intensity at the emission peak (521 nm) vs. perimeter-to-area ratio P/A for concave arrays with feature sizes of 40, 60, 80, and 100 µm. e Same as (d) for convex arrays. Scale bars, 200 µm (a). Source data are provided as a Source Data file.

Films with convex pillar arrays exhibited higher ML intensity than planar films. For convex pillars, the 40 µm array (P/A = 25 µm⁻¹) achieved the maximum ML intensity of 1462 (arb.u.), representing 130% of the planar-film value. The ML intensity of convex pillar arrays increased with decreasing feature size and increasing P/A. In contrast, concave hole arrays showed the opposite trend: ML intensity increased with increasing feature size and decreasing P/A, reaching a maximum of 1366 (arb.u.) for the 100 µm array (P/A = 4 µm⁻¹). Spectral measurements were performed for feature sizes of 40, 60, 80, and 100 µm (Fig. 3c and Supplementary Fig. 21), and the relationships between ML intensity, feature size, and P/A were established for both convex and concave structures (Fig. 3d, e).

Clear trends emerged from the results: for convex pillar arrays, smaller feature sizes and higher P/A ratios produced stronger luminescence intensities. In contrast, for concave hole arrays, larger features and lower P/A ratios resulted in higher intensities. These findings demonstrate that the luminescent properties of ML films can be systematically controlled by tuning the shape, size, and P/A ratio of microstructure arrays, enabling the fabrication of films with customizable brightness levels and spatial resolution. This provides a versatile toolbox for programming ML thin films as visual force sensors and force-responsive display units.

Hierarchical structures for stress-responsive luminescence

The incorporation of nanostructures onto the films forms hierarchical multiscale structures, further increasing photon escape efficiency. As demonstrated in Fig. 4, this figure elucidates the mechanism by which these multiscale structures enhance luminescent intensity by modifying the interaction between mechanical deformation and light emission. This design leads to an improvement in the overall performance of mechanoluminescent films.

Fig. 4: Mechanisms enhancing luminescence in mechanoluminescent films via hierarchical multiscale structures.
Fig. 4: Mechanisms enhancing luminescence in mechanoluminescent films via hierarchical multiscale structures.The alternative text for this image may have been generated using AI.
Full size image

a Schematic illustration of stress directions on micropillar arrays during detachment from the SU-8 template. b Finite-element simulation of stress distribution in a single convex micropillar under tensile loading. c Results during 11 s of cyclic operation: variation of film ML intensity at the emission peak (521 nm) under cyclic loading, displacements of the film in the X- and Z-directions, device velocity along the X-direction, and a schematic of the testing system. d Stress distribution in PDMS films with convex micropillar arrays under static conditions visualized by stress birefringence imaging. e Stress distribution in PDMS films with convex micropillar arrays under tensile loading visualized by stress birefringence imaging. f Comparison of films with hierarchical multiscale and microscale structures: luminescent images (left), SEM images (middle), and SEM of nanoscale surface patterns (right). g ML intensity comparison for films with hierarchical multiscale structures versus microscale structures. Scale bars, 80 µm (d, e); 1 mm (f, left); 100 µm (f, middle); 1 µm (f, right). Source data are provided as a Source Data file.

During the fabrication process — including coating, heating, curing, and peeling — intrinsic stresses develop within the material as a result of molecular rearrangement and local density variations (Fig. 4a). The geometry of the microstructures plays a critical role in determining the distribution and magnitude of both residual and externally applied stresses. In films with pillar-like microstructures, external mechanical loading induces pronounced stress concentration at the bases and apexes of the pillars due to geometric confinement, resulting in highly localized elastic deformation (Fig. 4b). This enhances mechanical activation of ZnS:Cu particles and increases interfacial contact–separation and sliding, facilitating triboelectric charge generation and strong, spatially localized mechanoluminescent emission. In contrast, concave (hole-type) microstructures distribute mechanical stress more broadly toward their edges, forming a ring-shaped stress pattern. This geometry leads to a more diffuse and lower-intensity luminescent response compared to the sharply localized, high-intensity emission of pillar-based structures (Supplementary Fig. 23). Overall, the geometry of the microstructures fundamentally determines the local stress landscape and, consequently, the efficiency and spatial distribution of mechanoluminescent output.

Under cyclic stretching, internal mechanical stress dynamics further influence luminescence behavior. As shown in Fig. 4c, the luminescence intensity reaches its peak during the first stretch cycle, owing to the rapid release of pre-existing stresses introduced during fabrication. In subsequent cycles, the luminescence gradually decreases and stabilizes at a steady-state bimodal pattern, indicating stress redistribution and mechanical equilibrium. This cyclic luminescence behavior occurs independently of the testing system equipment displacement or stretching speed (Supplementary Fig. 3), confirming that internal stress dynamics dominate the observed behavior. To visualize and quantify these stress patterns, stress birefringence imaging was applied (Fig. 4d and Supplementary Fig. 24). Before stretching, stress naturally accumulates at the bases of convex pillars and the edges of concave holes, reflecting the inherent stress concentration induced by residual stress accumulation during microfabrication. Upon mechanical stretching, stress redistributes directionally along the loading axis, forming highly localized stress concentrations. These regions appear as bright directional stripes in birefringence images, aligning closely with luminescent gradients observed under mechanical excitation.

At the nanoscale, hierarchical structures primarily enhance photon escape by increasing multiscale scattering and reducing total internal reflection at the film-air interface. Nanopatterned features such as nanopores and nanopillars enlarge the effective surface area, promoting multiple scattering events that randomize photon trajectories and facilitate light out-coupling (Fig. 4f). Finite-difference time-domain (FDTD) simulations (Supplementary Figs. 25–27) demonstrate that these nanopatterns increase the proportion of photons escaping from the film, leading to a substantial improvement in overall luminescent output (Fig. 4g). The observed enhancement is thus mainly attributed to improved scattering and modulation of local refractive index boundaries.

The hierarchical multiscale-structured ML films achieve substantial luminescence enhancement by integrating microscale stress concentration with nanoscale photon management. Microscale convex and concave patterns efficiently localize mechanical stress, leading to stronger activation of ZnS:Cu luminescent centers and introducing pronounced angle-dependent emission (Supplementary Fig. 28), with structured films maintaining high brightness over a wider range of viewing angles than planar controls. Nanoscale features, such as nanopillars and nanopores, act as scattering centers that disrupt total internal reflection and facilitate photon escape, as confirmed by both FDTD simulations and experimental data. This synergistic structural design enables precise spatial control of mechanical activation and maximizes light extraction, resulting in up to about a threefold increase in emission intensity and providing a clear framework for the rational engineering of high-performance mechanoluminescent devices.

Programmable visual stress mapping and information storage

Multiscale-structured two-dimensional pattern arrays offer a dynamic and real-time approach for visualizing in-film stress distribution. By introducing hierarchical checkerboard architectures, the stress and strain distributions can be effectively visualized with high spatial resolution. These designs utilize the stress concentration effects of microstructures to amplify luminescence under localized stress while enhancing light-material interactions through increased surface complexity. This combination optimizes strain visualization, where the precise adjustment of size, shape, spacing, and hierarchy of microstructures ensures enhanced stress mapping performance (Supplementary Fig. 32).

The checkerboard structure integrates macroscopic square blocks with embedded microscale features, forming a hierarchical design that operates across multiple scales. At the macroscopic level, regularly arranged raised and recessed square blocks introduce foundational stress modulation. At the microscale, features such as microholes and micropillars enhance light-surface interactions by increasing the effective surface area and generating localized optical effects. When subjected to external forces, the hierarchical microstructures exhibit varying brightness levels corresponding to local strain, thereby enabling real-time stress visualization (Fig. 5a). This design significantly improves stress detection at both macro and micro levels, producing detailed stress maps through luminescence variations as the film deforms (Fig. 5d, e).

Fig. 5: Programmable visual information retrieval and storage.
Fig. 5: Programmable visual information retrieval and storage.The alternative text for this image may have been generated using AI.
Full size image

a Real-time visualization of stress distribution in mechanoluminescent films using checkerboard hierarchical structures. b Planar film. c Conceptual schematic of strain distribution in checkerboard-structured films. d Macrostress monitoring: ML intensity maps under applied line and surface stresses. e Microstress detection: ML intensity maps under applied point and line stresses. f 3D visualization of ML intensity (102 arb.u.) for macrostress monitoring. g 3D visualization of ML intensity (102 arb.u.) for microstress monitoring. h SEM images of programmable “giant panda” arrays with feature sizes of 40, 80, and 100 µm. i Graded brightness of the “giant panda” pattern under mechanoluminescence for feature sizes of 40, 80, and 100 µm. j ML intensity values from aligned frames for the graded patterns shown in (i). k 3D surface map of ML intensity for the “giant panda” pattern. l ML brightness comparison between films with and without multiscale patterns. Scale bars, 2 mm (a,b,d,e,i,l); 500 µm (h). Colors: in ML images, color indicates ML intensity at 521 nm (brighter indicating higher intensity); in (f and g), the 3D surfaces directly show stress magnitude, with darker shades indicating higher stress.

The hierarchical design achieves refined microscale stress detection by creating high-contrast bright and dark regions at stress concentration points, allowing for precise stress localization (Supplementary Figs. 33, 34). Moreover, variations in microstructure height and luminescent intensity introduce a quasi-3D effect, enabling the visualization of stress gradients with a depth-like appearance (Fig. 5f, g). This capability differentiates point, line, and surface stress characteristics, supporting accurate static and dynamic stress monitoring.

By fine-tuning the microstructure geometry and arrangement, the hierarchical design enables the creation of gradient stress patterns with enhanced resolution and contrast. For instance, simulated depth perception is achieved by adjusting luminous intensity and stress gradients across different regions (Fig. 5h, i, and Supplementary Fig. 35). Notably, dynamic patterns—such as the giant panda pattern—are generated, showcasing spatially graded luminescence and intricate 3D effects (Fig. 5i–k). The stress gradients vary responsively under external forces, thereby producing a detailed and adaptive 3D stress map.

The hierarchical multiscale structures facilitate the creation of complex visual patterns, including checkerboard and giant panda arrays, within ML films. By modulating light intensity, these patterns achieve quasi-3D visualization with high precision. Importantly, the technology demonstrates pixelated mechanoluminescence with a minimal pitch of 40 μm on a single light-emitting film, encoding 281,250 pixels with a resolution of 637 PPI (Fig. 5h–l).

This design improves stress engineering and light extraction in ML films, supporting applications in real-time stress visualization, passive force sensors, and mechanical imaging at μm-scale pixel pitch. These findings indicate that hierarchical architectures enable pixelated mechanoluminescence for intelligent sensing and dynamic force mapping systems.

Discussion

This study presents a programmable mechanoluminescent film with hierarchical multiscale structures, enabling visualization of in-film stress distribution dynamics with enhanced sensitivity and resolution. The hierarchical ML films induce concentrated stress patterns using microstructures and improve photon escape efficiency using nanostructures. Through deliberate arrangement of these multiscale structures on an ML film, programmable, pixelated ML patterns can be achieved. Overall, this approach achieves about a threefold increase in ML intensity and enables quasi-3D visualization of stress gradients in real time. The technology supports applications in passive wireless stress monitoring, wearable electronics, and structural health monitoring, offering real-time, energy-efficient feedback for dynamic environments. This work provides a framework for visualizing film stress using ML materials, and demonstrates the use of ML films as programmable two-dimensional information storage media with a simple force-readout mechanism.

Methods

Materials

SU-8 negative photoresist and its dedicated developer purchased from KAYAKU. Trimethylchlorosilane (TMCS): ≥ 99.99%, purchased from Sigma-Aldrich. Zinc sulfide (ZnS): ≥ 99.99% purity, purchased from Sigma-Aldrich. Cu-doped ZnS mechanoluminescent phosphor (ZnS:Cu): average particle size 20 µm, obtained from Shanghai Scientific & Technical Co., Ltd., purity ≥ 99.9%. Polydimethylsiloxane (PDMS) elastomer kit (Sylgard 184): base and curing agent (both ≥ 99%), Dow Corning. Ethanol and isopropanol (for cleaning substrates): HPLC grade, ≥ 99.8%, Merck. Other reagents (e.g., acetone, hexane): analytical grade (≥ 99%), Sinopharm Chemical Reagent Co., Ltd.

Microstructured template fabrication

The microstructure template was fabricated using high aspect-ratio SU-8 photolithography through the following steps: First, 30 mL of SU-8 3050 negative photoresist was carefully dispensed onto the clean wafer surface. The wafer underwent manual spin-coating at a low speed to ensure uniform coverage, followed by high-speed spin-coating at 1200 rpm for 30 s to achieve a thin, consistent photoresist layer. After coating, the wafer was soft-baked on a temperature-controlled hotplate, first at 65 °C for 5 min, followed by 95 °C for 20 min. This gradual baking process helped drive off the solvent and form a stable, tack-free photoresist layer, completing the initial soft-baking step required for SU-8 processing. Once the soft baking was complete, photolithography was carried out using an SUSS MA6 mask aligner (UV), applying a dose of 210 mJ cm−2 to expose the photoresist to UV light through a photomask with the desired pattern. The wafer then underwent post-exposure bake at 95 °C for 5 min to enhance the chemical crosslinking of the exposed areas. Following post-exposure baking, the wafer was developed using SU-8 developer to remove the unexposed areas of the photoresist, revealing the desired microstructure pattern. Finally, the wafer was hard-baked at 150 °C for 10 min to ensure the complete curing of the photoresist, resulting in a stable photoresist microstructure template with μm-scale lateral features suitable for further processing.

Fabrication of hierarchical multiscale-structured templates

The nanostructures were added to the microstructured templates using nanoimprint lithography through the following steps: First, an adhesion promoter (AP) layer was spin-coated onto the microstructured template surface. Subsequently, a layer of nanoimprint resist was applied to the wafer through spin-coating. The wafer, coated with nanoimprint resist, was placed into a nanoimprint tool, where the nanostructured imprint template was carefully aligned with the photoresist-patterned wafer. Finally, UV nanoimprint lithography was employed to cure the UV nanoimprint resist, completing the transfer of the nanoimprint mold structure onto the template.

Fabrication of hierarchical multiscale-structured mechanoluminescent films

The mechanoluminescent films were fabricated using soft lithography as follows: The hierarchical multiscale structures template was first treated with trimethylchlorosilane vapor in a sealed Petri dish for 5 min to ensure surface anti-adhesion. A mixture of ZnS:Cu powder and PDMS was then prepared by blending PDMS with its curing agent at a 20:1 (w/w) ratio, followed by the addition of ZnS:Cu powder at a PDMS:ZnS:Cu (phosphor) ratio of 6:4 (w/w). The resulting mixture was thoroughly degassed under vacuum and poured onto the treated template. After spreading the material evenly, it was coated using a spin coater and cured in an oven at 55 °C for 7 h. Once cured, the film was carefully peeled off and trimmed to yield the hierarchical micro-nanostructured mechanoluminescent film.

Finite-element analysis

Finite-element analysis (FEA) was employed to investigate the stress distribution characteristics of microstructure arrays, single structures, and planar regions in mechanoluminescent films during deformation. The cross-sectional structure used in the simulations was consistent with the actual thickness of the fabricated films. Material properties such as the elastic modulus of the ML elastomer film used in the FEA were based on experimentally measured values. Similarly, the strain and deformation rate of the film were set according to experimentally observed data from the testing apparatus. To simulate the stress distribution during film stretching, boundary loads were applied to the front edge of the film, and transient conditions were used to model the deformation process.

Characterization

The morphology of the samples was characterized using a field-emission scanning electron microscope (JSM-7800F, JEOL). The three-dimensional morphology of the samples was analyzed using a 3D optical microscope (VHX-7000, Keyence). Stress-induced birefringence was measured using a polarized light microscope (Olympus BX53). The stress–strain curves were measured using a dynamic mechanical analyzer (DMA Q850, TA Instruments). In this study, a strain–optical testing system was developed to control the movement and deformation of the samples for testing optical performance and imaging-based strain mapping. All spectral data were acquired with the fiber-optic spectrometer (NOVA2S, 200–1100 nm) operated in cooled mode. Optical images were captured using the MTR3CMOS series dual-stage semiconductor deep-cooled CMOS camera (MTR3CMOS08300KPA), which was used for imaging the strain distribution.