Mechanoluminescence by Direct Laser Writing: A Seconds-Scale Fabrication Strategy

Mechanoluminescence by Direct Laser Writing: A Seconds-Scale Fabrication Strategy | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Mechanoluminescence by Direct Laser Writing: A Seconds-Scale Fabrication Strategy Wei Tao, Liang Shen, Yufeng Xue, Shicheng Yuan, Tingxuan Chen, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8106340/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Inorganic mechanoluminescent materials, which transduce mechanical force into light, are promising for power-free sensing, structural health monitoring and human-machine interfaces. However, their fabrication typically requires energy-intensive, protracted bulk synthesis methods such as solid-state sintering. Here, we report a rapid and in-situ laser writing strategy for fabricating mechanoluminophores, validated across multiple material systems. Using the classic ZnS/CaZnOS:Mn 2+ system, we demonstrate that the laser-induced luminophores retain the crystal structure and emission of their sintered counterparts, while exhibiting porous microstructures, shortened fluorescence lifetime (424.9 vs. 727.3 µs) and superior mechanoluminescent linearity with stress (fitting slope of 1.11 vs. 0.54). We leverage the top-down programmability of this approach to fabricate patterned mechanoluminescent sensors and demonstrate a deep learning-driven collision management system. Our work provides a general toolbox that accelerates the trial-and-error cycle of novel mechanoluminophores and enables high-precision luminescent patterning and on-demand sensor integration. Physical sciences/Materials science/Materials for optics Physical sciences/Optics and photonics/Optical techniques Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction In the era of embodied intelligence and the Internet of Everything, 1,2 smart systems from wearable devices to humanoid robots increasingly require real-time sensing of mechanical stimuli such as impact and friction to execute autonomous "sense-decide-act" cycles. 3 – 5 Mechanoluminescent (ML) materials have emerged as a transformative sensing technology, offering a unique solution by directly converting mechanical energy into visible light. 6–8 This photon-based signaling mechanism eliminates the need for external power, complex wiring, or intricate circuitry, enabling intuitive and visual monitoring of mechanical events. Since their initial development, 9 ML materials have diversified from single crystals to diverse systems encompassing organic emitters, 10 rare-earth-doped composites, 11 and transition metal-activated phosphors. 12 Their luminescent bands now span from ultraviolet to infrared, broaden their utility in bio and in vivo sensing. 13,14 Complementing this spectral versatility, their ultra-fast response down to microseconds and high spatial resolution far surpass those of conventional electrical sensors, 15,16 making them invaluable for dynamic impact detection and harsh-environment monitoring. 17,18 Despite these attractive features, the synthesis and integration of high-performance ML sensors present a major bottleneck. The prevailing method of high-temperature solid-state sintering typically requires prolonged annealing at temperatures exceeding 1000°C for serval hours, followed by grinding and blending with polymer matrices. 19 This process is energy-intensive and time-consuming, and often leads to grain coarsening or phase separation during compositing, degrading luminescence performance. Recent efforts to address these concerns have explored innovations in material synthesis and compositing processes. Alternative synthesis routes such as hydrothermal synthesis, 20 spray pyrolysis, 21 microwave-assisted synthesis, 22 and molten salt shielding method, 23 have shown promise in shortening synthesis time or lowering reaction temperatures. In parallel, integration strategies have evolved from simple bulk blending or sandwich configurations to more advanced techniques like 3D printing and electrospinning. 24–26 By modulating carrier migration behavior at the crystal–polymer interface, such approaches enable improved control over ML response modes, sensitivity ranges, and detection accuracy. 27 Nevertheless, limitations in energy efficiency, process complexity, or microstructure controllability persist, making a transformative manufacturing approach imperative. Here, we present a direct laser writing strategy for fabricating ML materials. Although laser-based processing offers a compelling pathway for additive and subtractive manufacturing, 28,29 its potential for the direct synthesis of ML materials remains largely unexplored. Laser thermal processing (LTP) could, in principle, reduce processing times from hours to seconds. The central challenge lies in achieving the highly localized, extreme temperatures required for ML crystal formation within a brief irradiation window. In this work, we systematically validate the feasibility of LTP across multiple material systems, including ZnS/CaZnOS:Mn, SrZnOS:Mn, SrP 2 O 7 :Mn, and Ca 9 Al(PO 4 ) 9 :Mn. By combining numerical simulation with experimental analysis, we characterize the morphology, crystal structure, photoluminescence, and mechanoluminescence properties of the laser-induced luminophores and compare them with their conventionally sintered counterparts. Capitalizing on the in-situ patterning capability of LTP, we further develop an intelligent vehicle collision management system that integrates deep learning-based recognition of ML patterns to enable automated emergency response, thereby closing a full "sense–decide–act" cycle. We believe this laser writing strategy will pave the way for future applications, such as structural health monitoring of laser-welded components and the rapid screening of novel ML phosphors. Results Fabrication of Laser-induced Luminophores Conventional high temperature solid-state sintering (HTSS) typically requires prolonged heat preservation at temperatures exceeding 1000°C, relying on thermally activated atomic diffusion, reaction, nucleation, and crystal growth of precursor powders to yield dense mechanoluminescent (ML) crystals. In contrast, this work pioneers employing laser thermal processing (LTP) to prepare ML crystals, as schematically illustrated in Fig. 1 a. The principle of LTP leverages localized high temperatures generated by laser irradiation to achieve in-situ densification and crystallization of precursor powders. Distinct from the sintering methods, the “top-down” LTP approach enables energy-efficient fabrication with high flexibility in pattern design via programmable laser scanning. The process initiates with mixing raw powders, which are then blade-coated into a uniform precursor film. Upon laser scanning, the film rapidly transforms into colored crystals within seconds. The resulting laser-induced luminophores (LILs) exhibit strong interparticle adhesion due to nucleation and growth, allowing them to be readily lifted from the surrounding loose precursor matrix using tweezers (Supplementary Fig. 1). It is worth noting that laser-induced thermal stress can cause partial powder detachment, resulting in irregular surface dispersion; however, this can be easily removed by gentle air blowing using an ear aspirator bulb. The incorporation of an optimized amount of B₂O₃ particles into the precursor effectively suppresses powder detachment during LTP (Supplementary Fig. 2). To elucidate the mechanism of LIL formation, we perform finite element method (FEM) simulations to analyze the influence of laser power and scanning speed on surface temperature and thermal stress distribution, using CaZnOS as a model system (simulation details are provided in Supplementary Note 1). As depicted in Fig. 1 b, under stationary laser irradiation at 10 W, a maximum surface temperature of 1840 K is achieved locally, well above the phase transition temperature of CaZnOS. 30 This confirms that the laser energy is sufficient to crystallize the precursor into ML-active CaZnOS. Notably, the temperature drops rapidly with distance from the laser spot (reaching only ~ 470 K at 1 cm away), underscoring the highly localized heating that enables both efficient crystallization and high-precision patterning. We further simulated axial surface temperature profiles under varying laser powers (Fig. 1 c). The results indicate a significant increase in local maximum temperature with increasing power, reaching 3700 K at 50 W. Under dynamic scanning conditions (i.e., laser writing), the peak temperature is lower than that under stationary irradiation; however, a distinct “temperature tail” is observed, where the temperature decays gradually post-irradiation. This tail mimics the heat preservation stage in HTSS and is essential for complete precursor crystallization. Additionally, we analyzed laser-induced thermal stress and its effect on LIL morphology. As illustrated in Fig. 1 d, at 10 W and 0.8 cm/s, the maximum thermal stress reaches 2.42 MPa, inducing a surface deformation of up to 525 µm. Further simulations show that higher laser power or slower scanning speed markedly increases the maximum thermal stress. These findings demonstrate that localized temperature and stress in LTP not only drive LIL formation but also critically influence surface morphology, as discussed in later sections. The FEM simulations confirm that LTP allows precise control over local temperature through adjustment of laser power and scanning speed, thereby accommodating the thermal requirements of various ML materials. To validate this versatility, we evaluate the ML performance of serval Mn-doped ML systems fabricated under identical LTP parameters, such as Ca 9 Al(PO 4 ) 9 :Mn, SrP 2 O 7 :Mn, SrZnOS:Mn, and ZnS/CaZnOS:Mn. For ML testing, a polyethylene terephthalate (PET) film is applied to the sample surface after fabrication, and ML emission is triggered by rubbing the PET with a glass rod. As shown in Fig. 1 e, the laser-scanned region of Ca 9 Al(PO 4 ) 9 :Mn appears light grey and exhibits deep red persistent photoluminescence (Pers PL) under 254 nm UV excitation, with a decay lifetime of ~ 0.5 s, but no detectable ML emission is observed. In the case of SrP 2 O 7 :Mn (Fig. 1 f), brighter red Pers PL with a longer lifetime of ~ 1.4s is observed under the same UV excitation; after PL completely fades, white ML emission is induced mechanically. In Fig. 1 g, SrZnOS:Mn exhibits a loose surface morphology after LTP, resulting in extremely weak ML emission. By contrast, ZnS/CaZnOS:Mn (ZCM) exhibits bright yellow mechanoluminescence (Fig. 1 h). Notably, neither SrZnOS:Mn nor ZCM shows significant Pers PL. The entire fabrication and ML characterization process of ZCM@LIL is captured in Supplementary Movie 1. These results confirm the versatility of LTP for producing functional mechanoluminophores and underscore its potential to serve as an alternative to conventional HTSS. Future efforts will focus on optimizing laser parameters and material compositions to further improve ML performance. Structural and Optical Characterization of the Mechanoluminephores Given the superior mechanoluminescence brightness of ZnS/CaZnOS:Mn among the LTP-fabricated systems, we systematically investigate the structural and optical properties of ZnS:Mn, CaZnOS:Mn, and their hybrid ZCM system. During LTP, both CaZnOS:Mn and ZCM develops three-dimensional porous architectures resembling “ginger-like” morphologies (Supplementary Fig. 3 and Fig. 2 a). We attribute this unique structure primarily to the decomposition of CaCO 3 , a common precursor in both systems, which releases CO 2 gas at 625°C. 31 Combined with localized laser-induced thermal stress, this gas evolution facilitates precursor crystallization and generates a fluffy, interconnected porous network. Although MnCO 3 decomposes at 300–350°C, 32 its low doping content is fails to induce detectable porosity in the ZnS:Mn system. As depicted in Fig. 2 a, ZCM@LIL appears yellow under daylight and emits orange photoluminescence under 256 nm UV excitation, with surface shadows highlighting its porous architecture. High-magnification SEM imaging reveals the hierarchical nature of this porosity. At 20 W laser power (Fig. 2 b, c), ZCM@LIL exhibits well-defined needle-like pores with uniform connectivity. At 30 W (Fig. 2 d), localized melting leads to irregular pore boundaries, and at 40 W (Fig. 2 e), extensive melting results in smoother surfaces and markedly reduced porosity. This trend is governed by the coupled effects of maximum temperature and thermal stress, where higher power promotes particle fusion and pore collapse. In Fig. 2 f, EDS elemental mapping confirms homogeneous distribution of Ca, Zn, O, S, and Mn in ZCM@LIL prepared at 40 W; Mn appears less uniform in highly undulated regions due to its low concentration. High-resolution small-area EDS further verifies successful synthesis and elemental distribution (Supplementary Fig. 4). To quantify porosity, we analyze the area of individual/non-connected pores in ZCM@LILs prepared at 20–50 W (Fig. 2 g). The average pore area decreases from ~ 0.05 mm² at 20 W to ~ 0.03 mm² at 50 W, underscoring the ability to tune microstructure via laser power for optimized ML performance. To verify the consistency in phase composition, we preform the XRD characterization on ZnS:Mn, CaZnOS:Mn, and ZCM systems prepared by LTP and HTSS (Supplementary Fig. 5). For ZnS:Mn, diffraction peaks of LILs fabricated across 10–40 W closely match hexagonal wurtzite ZnS (PDF#36-1450), with no cubic sphalerite phase (PDF#05-0566) detected. The transition from sphalerite to wurtzite occurs near ~ 1020°C; 33 the observed wurtzite dominance confirms that LTP locally exceeds this temperature, despite its short processing time, enabling formation of the thermodynamically stable phase preferred for ML applications. Minor residual precursor peaks (e.g., MnCO 3 ) are occasionally detected due to sampling of LILs from the precursor matrix. For CaZnOS:Mn, LIL diffraction patterns align well with hexagonal CaZnOS (PDF#01-076-3819) across the same power range. Trace CaO peaks appear at 32.20°, 37.36°, and 53.86°, consistent with partial decomposition above 1370 K, 34 along with slight CaCO₃ residues (PDF#00-005-0586). In ZCM, LIL XRD profiles match those of HTSS samples, confirming coexistence of wurtzite ZnS and hexagonal CaZnOS. Although LIL diffraction peaks are less intense than HTSS counterparts (likely due to shorter crystal growth during LTP’s rapid heating/cooling cycle), their identical peak positions verify that LTP successfully synthesizes target crystal phases for all three systems, laying the foundation for subsequent optical analysis. Furthermore, we conduct PL measurements on the three systems prepared by LTP and HTSS (Supplementary Fig. 6). With the emission wavelength fixed at 585 nm, the photoluminescence excitation (PLE) peak positions of ZnS:Mn differ minimally between LTP (346.3 nm) and HTSS (347.8 nm). In contrast, CaZnOS:Mn and ZCM exhibit ~ 10 nm blueshifts: CaZnOS:Mn peaks at 340.2 nm (LTP) and 351.5 nm (HTSS), and ZCM at 342.2 nm (LTP) and 352.1 nm (HTSS). These differences likely stem from process-dependent variations in crystal structure, doping environment, and defect distribution, altering excitation paths and energy-level transitions. In addition, under 342 nm excitation, ZnS:Mn emission peaks remain similar (588.5 nm for LTP vs. 591.0 nm for HTSS), whereas CaZnOS:Mn and ZCM exhibit noticeable emission shifts, consistent with PLE trends. To explore luminescence dynamics, we plot PL decay curves in logarithmic coordinates and fit them with a bi-exponential model ( ). Calculated fluorescence lifetimes reveal that all LIL samples have shorter lifetimes than HTSS counterparts. The ZCM system shows the most pronounced difference, with lifetimes decreasing from 727.3 ns (HTSS) to 424.9 ns (LIL), i.e., a 41.6% reduction. Shorter lifetimes typically indicate enhanced non-radiative transitions or faster energy transfer, suggesting that LILs may offer improved response speed and dynamic performance for mechano-optical sensing and display applications. Based on the SEM, XRD, and PL analyses, the LTP process is confirmed to successfully synthesize crystalline phases in ZnS:Mn, CaZnOS:Mn, and ZCM systems. To evaluate their mechanoluminescent performance, the three LIL systems are subjected to grinding and blending treatments, followed by systematic ML characterization under handwriting friction, compression loading, and impact excitation. As shown in Fig. 3 a, the ground LIL powders are sandwiched between PET films. Rubbing the PET surface with a glass rod elicits bright ML emission visible to the naked eye. Under identical applied force, ML emission spectra of the three systems are measured by a spectrometer, revealing that their emission peaks are all centered at approximately 590 nm. Furthermore, CIE chromaticity coordinate analysis of the ML emission spectra from the three systems showed negligible chromaticity differences (Fig. 3 b), with ZnS:Mn@LIL (0.560, 0.442), ZCM@LIL (0.566, 0.436), and CaZnOS:Mn@LIL (0.551, 0.444), indicating a high degree of consistency in ML color emission. In line with previous work, 35 the ZCM@LIL system exhibit slightly higher luminescence intensity than the CaZnOS:Mn@LIL system and is significantly brighter than ZnS:Mn, with an intensity approximately 1.6 times higher than the latter. The inset in Fig. 3 a displays time-lapse images of the ZCM@LIL system during handwriting, captured using a smartphone. Based on these findings, the ZCM system is selected as the primary research object for subsequent ML performance studies. The ML mechanism is illustrated in the inset of Fig. 3 b: when the ML crystals are subjected to mechanical stimulation, lattice distortion occurs, and the accumulation of piezoelectric charges induces band bending, thereby promoting carrier release and migration. The identical emission peak positions across the three systems confirm that their ML originates from the transition of 3 d 5 electrons in Mn 2+ from the 4 T 1 to the 6 A 1 energy level. 36–38 To assess the bulk luminescence behavior of LILs, ZCM@LIL powders are blended with epoxy resin (ER) to form ML-ER blends. Under uniaxial compression testing (Fig. 3 c), the ML intensity depends strongly on the compression rate. At a lower compression rate (1 mm/s), the ML intensity of the samples is relatively weak, and a pre-compression of approximately 3 mm is required to trigger stable luminescence. However, at a compression rate of 3 mm/s, the samples exhibit a ML response with enhanced intensity without pre-compression, further demonstrating the force-light response capability of the LIL system. The embedded figure shows the ML image at the peak intensity. When the compression rate increases to 5 mm/s, the ML intensity reaches its maximum, 2.49 times that at 1 mm/s. However, as the compression force further increases, the luminescence intensity rapidly decays, indicating that the ML luminescence intensity and response range are dependent on the rate of mechanical stimulation. The potential for stress propagation detection is evaluated using cylindrical ML-ER composites compressed at 1.67 mm/s (Fig. 3 d). Peak ML intensity occurs at 8% strain, and the stripes fade after loading ceases. After unloading, surface deformation marks align with the ML stripe distribution, confirming that the ZCM@LIL–ER composite visually maps stress propagation and strain distribution. Supplementary Movie 2 and 3 visually capture the dynamic evolution of ML emission in cylindrical and cubic composites, respectively, demonstrating the capability of the ZCM@LIL system to visualize localized stress distribution. Finally, to demonstrate the practical utility of LTP for functional device design, we fabricate patterned mechanoluminescent composites and evaluate their spatial selectivity and impact response. As schematically illustrated in Fig. 3 e, a flexible ML-ER lamina is prepared by drop-casting epoxy resin onto a laser-patterned LIL film, followed by curing and peeling from the substrate. High-speed imaging during iron ball impact reveals spatially selective mechanoluminescence exclusively from the laser-written “crosshair” region (Fig. 3 f), confirming the patterning precision of LTP and its potential for spatially resolved stress sensing. We further compare the impact-induced ML performance of non-patterned ZCM@LIL-ER laminas with conventional HTSS-derived counterparts by varying the drop height of the iron ball. As summarized in Fig. 3 g, the ZCM@LIL system exhibits lower overall ML intensity across all tested heights, which we attribute to its inherent porous architecture and lower particle packing density relative to the densely sintered HTSS samples. Insets in Fig. 3 g show representative emission images at impact heights ranging from 5 to 40 cm. Notably, linear fitting of ML intensity versus impact force reveals that the ZCM@LIL composite exhibits a significantly steeper slope compared to ZCM@HTSS (1.11 vs. 0.54), indicating superior force–light sensitivity. This enhanced linearity suggests that the porous LIL structure undergoes structural collapse and facilitates efficient stress transfer under high impact, amplifying luminescence output. These results underscore the potential of ZCM@LIL-ER laminas as visual impact sensors with high discriminative capacity across a broad dynamic range, providing a foundation for applications in stress mapping and impact monitoring. Deep Learning-Driven Intelligent System via LIL Sensors To demonstrate a "sense-decide-act" cycle using LIL sensors, we develop an intelligent vehicle collision management system that integrates programmable ZCM@LIL sensors, convolutional neural network (CNN)-based optical analysis, microcontroller signal processing, and WiFi-enabled actuator control. This closed-loop framework accurately identifies impact locations through mechanoluminescent pattern recognition and facilitates rapid automated countermeasures to mitigate collision consequences. As illustrated in Fig. 4 a, ZCM@LIL sensors are patterned on PET substrates using trajectory-programmable LTP. Four alphabetic characters, namely “W”, “S”, “A”, and “D”, are designed to represent front, rear, left, and right impact positions, respectively. After laser-induced crystallization, ER is applied to the LIL surface and cured, forming a protective layer that improves mechanical robustness while maintaining high optical transparency for clear ML signal transmission. A comprehensive dataset is constructed by recording ML images under varied impact conditions, including different ball masses, drop heights, and imaging setups. Representative ML images for each character are shown in Fig. 4 a, with full datasets and detailed network architecture provided in Supplementary Note 2. To enhance model robustness, ML images undergo feature enhancement and data augmentation (Fig. 4 b). Attention maps visualize the model's focus regions during classification, revealing that raw ML images exhibit weak or ambiguous features. Image sharpening using a Laplacian operator improves contour clarity, and binarization further enhances character distinctness. To address the limited sample size (15 ML images per letter class, 60 in total), data augmentation techniques including salt-and-pepper noise and random erasing expand the dataset to 3600 images, significantly improving generalization. A four-stage CNN architecture (Fig. 4 c) processes 224×224 pixel inputs through convolutional layers with 3×3 kernels, BatchNorm2d, LeakyReLU (α = 0.1), and 2×2 max-pooling. This progressively reduces spatial dimensions from 224 to 14 while increasing channels from 32 to 256, enabling hierarchical feature extraction from local edges to global structures. Dropout (rates: 0.5, 0.5, 0.4) and BatchNorm1d layers before fully connected modules help suppress non-critical activations, effectively implementing an implicit spatial attention mechanism. Feature maps are resized via AdaptiveAvgPool2d (7×7) and mapped through three linear layers (output sizes: 512, 256, and 4), yielding classification probabilities for “A”, “D”, “S”, and “W”. The confusion matrix (Fig. 4 d) shows 100% accuracy for “D”, “S”, and “W”, and 95.7% for “A” (4.3% misclassified as “W”), with an overall accuracy of 98.93%. The full system architecture is illustrated in Fig. 4 e. ZCM@LIL-ER laminas labeled “W”, “S”, “A”, and “D” are installed inside the front, rear, left, and right sections of a vehicle model. A darkroom environment enhances ML signal detection. Upon collision, an embedded camera captures real-time ML emission from the impacted region. As shown in Fig. 4 a, ML images acquired under varying impact conditions simulate real-world collisions with differing locations and force intensities. A microcontroller unit processes the ML images, converts them to binary format, and classifies the impact location using the embedded CNN. Results are transmitted via WiFi to the central control system, triggering actuators such as alarms, brake assistance, steering correction, or airbag alerts. To verify the functionality of the integrated system, we present a simplified demonstration using a model car's motor/steering control as an example (Supplementary Movie 4). This system can achieve four emergency avoidance operations based on the luminescent images. for example, advancing forward in response to a rear impact (“W”), or moving forward while turning left in response to a right-side impact (“A”). As shown in Video S4, the integrated system exhibits an average response time of ~ 150 ms from ML luminescence image reading to motor/steering control, demonstrating real-time capability. This LIL–deep learning integrated system offers a novel strategy for intelligent collision management, highlighting high sensitivity, rapid response, and applicability in smart sensing and human–machine interaction. Discussion In summary, this work establishes a versatile laser writing strategy for the rapid in-situ fabrication of mechanoluminophores, exemplified by the synthesis of SrP 2 O 7 :Mn, SrZnOS:Mn, and ZnS/CaZnOS:Mn. By replacing energy-intensive and prolonged sintering with CO 2 laser writing strategy, we achieve crystallization within seconds while yielding porous microstructures favourable for stress–light coupling. Structural and optical analyses confirm that the laser-induced luminophores maintain crystalline phases analogous to sintered benchmarks, yet exhibit a 41.6% reduction in fluorescence lifetime, enabling accelerated luminescence dynamics. More importantly, the LIL-ER composites display intense orange mechanoluminescence with significantly enhanced linear response to stress, offering improved sensing accuracy. Capitalizing on the top-down patterning capability of this approach, we further construct programmable ML sensors and integrate them into a deep learning-augmented collision management platform. A customized CNN architecture trained on an augmented dataset achieves 98.93% classification accuracy for impact localization. The full-system demonstration of incorporating the ML sensing lamina, microcontroller, and wireless communication enables real-time collision mapping and triggers appropriate emergency responses. Our study not only introduces a general and efficient toolbox for developing novel mechanoluminophores and high-resolution luminescent patterning, but also illustrates a seamless materials-algorithm co-design framework toward intelligent autonomous systems. Methods Chemicals and materials ZnS, CaCO 3, MnCO 3 , SrCO 3 , H 6 NO 4 P, Al 2 O 3 , epoxide resin, aromatic amine-type curing agent, anhydrous ethanol, are obtained from Shanghai Macklin Biochemical Co., Ltd. All materials are used as received without further purification. For Mn-doped samples, including ZnS:Mn, CaZnOS:Mn, SrZnOS:Mn, SrP 2 O 7 :Mn and Ca 9 Al(PO 4 ) 9 :Mn, the Mn doping level was maintained at 1 mol%. For ZnS/CaZnOS:Mn system, ZnS, CaCO 3 , and MnCO 3 are proportioned at 3:2:2%. 35 After proportioning according to the corresponding molar ratios, the raw materials are transferred to an agate mortar, an appropriate amount of absolute ethanol is added, and the mixture is ground for thirty minutes until the raw materials are uniformly mixed. Subsequently, it is placed in an oven and dried at 60°C for 4 hours for later use. Laser thermal processing The dried precursor mixture is uniformly coated onto a glass slide using a blade coater. A CO 2 laser system (Gd Han’s Yueming Laser Group Co., Ltd.) with a fixed focal length of 8 mm is employed for irradiation. Laser-induced luminophores are fabricated in ambient atmosphere by scanning the laser beam over the precursor film according to pre-defined patterns, with laser power and scanning speed adjusted as needed. High-temperature solid-state sintering The dried precursor mixtures were placed in an alumina crucible and transferred to a tube furnace (Hefei Kejing Materials Technology Co., Ltd.). Under a continuous argon flow, the temperature was raised to 1050°C at a rate of 10°C/min, held for 4 h, and then cooled naturally to room temperature to obtain the final ML crystals. Fabrication of ML samples preparation For PET-encapsulated samples (PET-ML-PET), ground ML crystals are being uniformly dispersed between two square PET sheets (0.5 mm thick) to form a sandwich structure. Mechanoluminescence (ML) is being excited by pressing or sliding a glass rod over the PET surface; the resulting emission is being recorded using a smartphone camera, and spectra are being acquired using a fiber-optic spectrometer. For ML-ER blends, epoxy resin and amine curing agent are being mixed at a mass ratio of 2:1. The resulting solution is being combined with ground ML crystals at a 1:2 mass ratio (resin:crystals), cast into a custom silicone mold, and cured at 80°C for 6 h before demolding. For ML-ER laminas, LILs are first being prepared on glass or PET substrates via LTP. A 3D-printed square frame mold (20 mm × 20 mm × 2 mm) is being fixed onto the sample surface, filled with the ER solution, and cured at room temperature for 24 h to form a protective, transparent encapsulation layer. Characterizations The XRD patterns are measured using a Bruker D8ADVANCE Multifunctional X-ray Diffractometer. SEM images of ML samples are obtained by a ThermoFisher VeriosG4UC. The photoluminescence fluorescence spectra are acquired via EDINBURGH-FLS 1000. An Ocean Optics QE65pro fiber optic spectrometer is used to collect the ML signals of the samples encapsulated by PET. Luminescence images were captured using a smartphone and analyzed in MATLAB to extract intensity values. Declarations Data Availability All data are available in the main text or the supplementary information. Acknowledgements This work is supported by General Program of the Fundamental Research Funds for the Central Universities (2023ZYGXZR052), Guangzhou Applied Basic Research Program Project (2024A04J3601), The 10th Youth Talent Cultivation Project of China Association for Science and Technology (YESS20240469). The authors acknowledge the support from the Guangdong Provincial Key Laboratory of Intelligent Disaster Prevention and Emergency Technologies for Urban Lifeline Engineering, and Dongguan University of Technology Analytical and Testing Center. Ethics declarations The authors declare no competing interests. Supplementary Information Supplementary Information References Li J et al (2025) AI-embodied multi-modal flexible electronic robots with programmable sensing, actuating and self-learning. Nat Commun 16:8818 Obayashi N, Abdulali A, Iida F, Hughes J (2025) Embodied intelligence paradigm for human-robot communication. Sci Rob 10:eads8528 Kim H et al (2024) Inherently integrated microfiber-based flexible proprioceptive sensor for feedback-controlled soft actuators. npj Flex Electron 8:15 Chen C et al (2025) Bio-inspired multimodal soft actuator with environmental self-adaptation. Nat Commun 16:7630 Chen S et al (2025) A sensor-actuator–coupled gustatory interface chemically connecting virtual and real environments for remote tasting. Sci Adv 11:eadr4797 Fardian-Melamed N et al (2025) Infrared nanosensors of piconewton to micronewton forces. Nature 637:70–75 Casar JR et al (2025) Upconverting microgauges reveal intraluminal force dynamics in vivo. Nature 637:76–83 Chang S et al (2025) Hair-Like Mechanoluminescent Structures with Ultralow Activation Threshold for Dynamic Force Sensing. Adv Mater, 2507634 Xu C, Watanabe T, Akiyama M, Zheng X (1999) Artificial skin to sense mechanical stress by visible light emission. Appl Phys Lett 74:1236–1238 Yin Y et al (2024) Force-triggered hypso-and bathochromic bidirectional fluorescence switching beyond 120 nm and its anticounterfeiting applications. Sci Adv 10:eadk5444 Zhang S et al (2024) Tunable Full-Color Mechanoluminescence in Rare Earth‐Doped Transparent Amorphous Glass. Adv Funct Mater 34:2404439 Wang H et al (2025) Oscillatory mechanoluminescence of Mn2+-doped SrZnOS in dynamic response to rapid compression. Nat Commun 16:548 Cheng K et al (2025) A Universal Strategy for Multicolor Mechanoluminescence via Radiative Energy Transfer Based on Ultraviolet Mechanoluminescent Material Ca9Al (PO4) 7: Ce3+. Laser Photonics Rev, 2401524 Liu S et al (2025) Bright Chromium-Sensitized Lanthanide NIR‐II Mechanoluminescence in a Piezoelectric Oxide. Adv Mater, e06957 Chen C et al (2023) Revealing the Intrinsic Decay of Mechanoluminescence for Achieving Ultrafast-Response Stress Sensing. Adv Funct Mater 33:2304917 Wang X et al (2017) Full Dynamic-Range Pressure Sensor Matrix Based on Optical and Electrical Dual-Mode Sensing. Adv Mater 29:1605817 Zhou B et al (2023) Mechanoluminescent-triboelectric bimodal sensors for self-powered sensing and intelligent control. Nano-Micro Lett 15:72 Hou B et al (2022) An interactive mouthguard based on mechanoluminescence-powered optical fibre sensors for bite-controlled device operation. Nat Electron 5:682–693 Tao W et al (2025) Flexible mechano-optical sensors from mechanoluminescence to mechanoplasmonics: designs, applications, and prospects. J Mater Chem C 13:2058–2090 Wang Z, Tai Y, Nam J, Yin Y (2023) Calcination-induced transformation of ZnS: Mn2 + nanorods to microparticles for enhanced mechanoluminescence. Chem Mater 35:6845–6852 Pugina RS et al (2021) Nd3+: YAG microspheres powders prepared by spray pyrolysis: Synthesis, characterization and random laser application. Mater Chem Phys 269:124764 Ugbo FC et al (2023) Optimizing the Mechanoluminescent Properties of CaZnOS: Tb via Microwave-Assisted Synthesis: A Comparative Study with Conventional Thermal Methods. Materials 16:3511 Zhu M et al (2023) Novel Red Mechanoluminescence in Mn-Doped ZnGa2S4 Crystal Phosphors via a Molten Salt Shielding Method. Laser Photonics Rev 17:2300517 Hajra S et al (2023) Flexible composite material for self-powered applications via triboelectricity and mechanoluminescence: PDMS/ZnS: Cu composites. Nano Energy 114:108668 Zheng T et al (2023) Mechanoluminescence and photoluminescence heterojunction for superior multimode sensing platform of friction, force, pressure, and temperature in fibers and 3D-printed polymers. Adv Mater 35:2304140 Wang Z, Tai Y, Ye Z, Nam J, Yin Y (2024) Integration of ZnS: Mn2⁺ Microparticles into Electrospun PVDF-Based Nanofibers for Enhanced Mechanoluminescence. Adv Funct Mater 34:2410358 Jeong HI et al (2025) Super elastic and negative triboelectric polymer matrix for high performance mechanoluminescent platforms. Nat Commun 16:854 Chen Z et al (2025) 25 nm-Feature, 104-aspect-ratio, 10 mm2-area single-pulsed laser nanolithography. Nat Commun 16:7434 Han J et al (2025) Non-contact ultrasound to assist laser additive manufacturing. Nat Commun 16:7613 Zhang J-C et al (2013) An intense elastico-mechanoluminescence material CaZnOS: Mn2 + for sensing and imaging multiple mechanical stresses. Opt Express 21:12976–12986 Zhuang D, Chen Z, Sun B (2025) Thermal Decomposition of Calcium Carbonate at Multiple Heating Rates in Different Atmospheres Using the Techniques of TG, DTG, and DSC. Crystals 15:108 Kissinger H, McMurdie H, Simpson B (1956) Thermal decomposition of manganous and ferrous carbonates. J Am Ceram Soc 39:168–172 Moore D, Wang ZL (2006) Growth of anisotropic one-dimensional ZnS nanostructures. J Mater Chem 16:3898–3905 Gulyaeva R, Selivanov E, Vershinin A, Chumarev V (2006) Thermal expansion of CaZnSO. Inorg Mater 42:897–900 Peng D et al (2020) A ZnS/CaZnOS heterojunction for efficient mechanical-to‐optical energy conversion by conduction band offset. Adv Mater 32:1907747 Mukhina MV et al (2021) Single-particle studies reveal a nanoscale mechanism for elastic, bright, and repeatable ZnS: Mn mechanoluminescence in a low-pressure regime. ACS Nano 15:4115–4133 Wang Z et al (2024) Stress-induced multi‐stimulus‐responsive mechanoluminescence in Mn2 + doped double perovskite compound. Adv Opt Mater 12:2301796 Wang Y et al (2025) Smart Semiconductors: An Overview of AZnOS (A = Ca, Sr, Ba) Based Phosphors, Multi-Stimuli Responsive Luminescence, and Potential Application Prospects. Adv Sci Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMovie1.mp4 Supplementary Movie 1: Fabrication and ML performance of laser-induced luminophores SupplementaryMovie2.mp4 Supplementary Movie 2: ML performance of ZCM@LIL-ER cylindrical blends SupplementaryMovie3.mp4 Supplementary Movie 3: ML performance of ZCM@LIL-ER cubic blends SupplementaryMovie4.mp4 Supplementary Movie 4: Demonstration of the deep learning-driven control system SupplementaryInformation.docx Supplementary Information Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8106340","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":570338272,"identity":"d82a9728-4de1-42d2-89ba-54518cbfa37b","order_by":0,"name":"Wei Tao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAz0lEQVRIiWNgGAWjYNACAxvGBjCDjQjFPBAtaUAtzCRpYThMghZ79sPHJD4UnJedH5F/gOFD2WEG/tkNBGzhSUuTnGFw23jjjWQGxhnnDjNI3DlAyGE5Zrd5DG4nbpyRzMDM23aYwUAigYAW/jdmt/8YnINo+UuUFgmgLQwGBxLnSwC1MBKl5caz9J89BsnGG3geGxzsOZfOI3GDgBb2/uTDBj/+2MnOb098+OBHmbUc/wwCWuDA4AADwwEGeEQRA+QbiFc7CkbBKBgFIwwAADrFQlYtGgzzAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-4877-4737","institution":"Dongguan University of Technology","correspondingAuthor":true,"prefix":"","firstName":"Wei","middleName":"","lastName":"Tao","suffix":""},{"id":570338273,"identity":"ea93f5ac-0324-4330-aa3a-41a153a54ae8","order_by":1,"name":"Liang Shen","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Liang","middleName":"","lastName":"Shen","suffix":""},{"id":570338274,"identity":"d4ef0a36-fd75-4cdd-b7fd-fe981384772b","order_by":2,"name":"Yufeng Xue","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Yufeng","middleName":"","lastName":"Xue","suffix":""},{"id":570338275,"identity":"03224579-6897-43cb-baa9-87fb7f85d92f","order_by":3,"name":"Shicheng Yuan","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Shicheng","middleName":"","lastName":"Yuan","suffix":""},{"id":570338276,"identity":"3e8de37c-bff1-4362-9819-58a7b5f70e92","order_by":4,"name":"Tingxuan Chen","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Tingxuan","middleName":"","lastName":"Chen","suffix":""},{"id":570338277,"identity":"6bc71437-f905-4797-a02f-e8b534dda4aa","order_by":5,"name":"Qinhua Hu","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Qinhua","middleName":"","lastName":"Hu","suffix":""},{"id":570338278,"identity":"71125444-81a1-44a7-b0d7-15e9d461edab","order_by":6,"name":"Wenjie Wu","email":"","orcid":"","institution":"South China University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Wenjie","middleName":"","lastName":"Wu","suffix":""},{"id":570338279,"identity":"fe4edb70-2461-46bf-8680-6e0b4e782add","order_by":7,"name":"Xiaokang Chen","email":"","orcid":"","institution":"South China University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiaokang","middleName":"","lastName":"Chen","suffix":""},{"id":570338280,"identity":"4a29b5d8-aee8-4969-89bd-2dfc12e02350","order_by":8,"name":"Enjie Wang","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Enjie","middleName":"","lastName":"Wang","suffix":""},{"id":570338281,"identity":"ca85194b-82a8-4964-af5f-6a98c9a86341","order_by":9,"name":"Ye Liu","email":"","orcid":"","institution":"Dongguan University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Ye","middleName":"","lastName":"Liu","suffix":""},{"id":570338282,"identity":"07159c21-2d1e-46c0-bd5b-3bf262f452a6","order_by":10,"name":"Thomas Maurer","email":"","orcid":"","institution":"École Centrale de Lille","correspondingAuthor":false,"prefix":"","firstName":"Thomas","middleName":"","lastName":"Maurer","suffix":""},{"id":570338283,"identity":"e0ca1c93-31f4-4ee0-ab4c-ce00ccb76f58","order_by":11,"name":"Dengfeng Peng","email":"","orcid":"https://orcid.org/0000-0002-9714-2317","institution":"Shenzhen University","correspondingAuthor":false,"prefix":"","firstName":"Dengfeng","middleName":"","lastName":"Peng","suffix":""}],"badges":[],"createdAt":"2025-11-13 13:50:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8106340/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8106340/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99750363,"identity":"e415d3a2-1ed8-4869-ab43-0662c2e3bdc7","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2185655,"visible":true,"origin":"","legend":"","description":"","filename":"Maintext.docx","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/fca8746bbe2c47d4174429d2.docx"},{"id":99750360,"identity":"192e71f1-846b-4b66-9f0a-b23e17eef603","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12192,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS2591816.json","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/aaffdb21c96a5697434fbb4c.json"},{"id":99750357,"identity":"65cae627-6b40-4c5c-a281-a6957fde2b1e","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2516378,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/75ef1c35bc86433cdc7a1c56.docx"},{"id":99750350,"identity":"e295e120-2c39-499e-8b25-6032c97925fa","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":88453,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25918160enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/8c5c2e87efa8ca0d7daab210.xml"},{"id":99750346,"identity":"040ece19-4f2f-48a1-a242-632a3adcbbf5","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":486720,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/b677da1640bea11b362d2f5a.jpeg"},{"id":99750345,"identity":"11fbe7c3-a1e2-42c9-9dc1-b90b4acdc323","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":783073,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/86d876080df57d22e5a87ef5.jpeg"},{"id":99750339,"identity":"4ff8ba92-3d57-4aa2-a395-2df3c6d8a9e6","added_by":"auto","created_at":"2026-01-08 03:39:14","extension":"wmf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1660,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.wmf","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/ce2d00b05be0b36331b7f454.wmf"},{"id":99750355,"identity":"cf5efab9-181f-47b3-867a-df90410fc19c","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":333945,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/dae4dc0e2a8c582896509ba4.jpeg"},{"id":99750359,"identity":"d608432c-e239-40e0-bb2d-6612f3ba104a","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":506222,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/71d9e78467f16e62b5aee10f.jpeg"},{"id":99750343,"identity":"4251e6be-8768-4a40-a7b5-748c60afd8f5","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"wmf","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1660,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.wmf","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/cca2515eafab628a9384b41e.wmf"},{"id":99797883,"identity":"6af4719a-6e9f-4a71-bd74-55a7a7bfea51","added_by":"auto","created_at":"2026-01-08 13:46:49","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":167106,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/3ca324e76b6eede5b470c552.png"},{"id":99750352,"identity":"07ba7fc5-3b25-4422-a71c-a38b093e79d4","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":319056,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/b3e92b83174ef71b7b48b4d6.png"},{"id":99750353,"identity":"6f303505-a183-422e-a32c-23f3f76c9a35","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1093,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/df95a8a0579b22d0c17d6a8d.png"},{"id":99750351,"identity":"99ed5dc4-d8c0-41c1-aff8-1300bcf56bbf","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":95065,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/f080746bb8f0deb5d34bcc20.png"},{"id":99750348,"identity":"e6d7eaf5-9de3-41d8-8379-9b3ed26ae6c4","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":187022,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/10418ca43f94290b8e7bb59b.png"},{"id":99750364,"identity":"c88fa571-a920-481c-ac14-ec25e4a5a6f0","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1093,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/709bc29345a84522496409ef.png"},{"id":99750338,"identity":"8182f6b2-225c-4290-b64c-cee38efc1e74","added_by":"auto","created_at":"2026-01-08 03:39:13","extension":"xml","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":84358,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25918160structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/dee3048f48dd562eb6774523.xml"},{"id":99750356,"identity":"076300e9-ac46-47d4-bce7-cb0e2e7d7880","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"html","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":94829,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/aa5e69109b1356afdb8595ec.html"},{"id":99750344,"identity":"31e1731d-efbf-4999-b646-9b2e96e0416f","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":330906,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFabrication, theoretical simulation and mechanoluminescence characterization of laser-induced luminophores.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e Schematic diagram of the LIL fabrication workflow, including raw materials mixing, blade coating of percussor films, to laser thermal processing. \u003cstrong\u003eb-e\u003c/strong\u003e Finite element method simulations showing the distribution of temperature and thermal stress during LTP. \u003cstrong\u003ef\u003c/strong\u003e Simulated thermal stress under representative laser parameters. \u003cstrong\u003eg-i\u003c/strong\u003e Persistent photoluminescence and mechanoluminescence properties of various laser-induced luminophores: Ca\u003csub\u003e9\u003c/sub\u003eAl(PO\u003csub\u003e4\u003c/sub\u003e)\u003csub\u003e9\u003c/sub\u003e: Mn, and SrP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e7\u003c/sub\u003e: Mn, SrZnOS: Mn, and ZnS/CaZnOS:Mn.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/47994f18ec8d919f477ee914.jpg"},{"id":99750341,"identity":"25daa677-9d63-4339-b0bf-ec159af3dd53","added_by":"auto","created_at":"2026-01-08 03:39:14","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":526826,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMorphological and compositional characterization of ZnS/CaZnOS:Mn @LIL.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea \u003c/strong\u003eOptical images showing ginger-like porous structures under daylight and UV light. \u003cstrong\u003eb-e \u003c/strong\u003eEM images of hierarchical porous structures at different laser powers. \u003cstrong\u003ef\u003c/strong\u003eEDS elemental maps of Ca, Zn, O, S, and Mn for LIL prepared at P = 40W. \u003cstrong\u003eg \u003c/strong\u003eQuantitative analysis of individual pore area as a function of laser power.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/26071cc11d4ce09dce11cf9e.jpg"},{"id":99797617,"identity":"d152d337-3788-41b9-a59c-b0ae13e45a86","added_by":"auto","created_at":"2026-01-08 13:46:10","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":254693,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMechanoluminescent Performance of ZnS/CaZnOS:Mn@LIL composites.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea-b\u003c/strong\u003e ML response, CIE chromaticity, and mechanism of PET-encapsulated ZCM@LIL powders under glass rod friction. \u003cstrong\u003ec\u003c/strong\u003e ML intensity of ZCM@LIL-ER blends under uniaxial compression at different rates. \u003cstrong\u003ed\u003c/strong\u003e Stress propagation visualization in a cylindrical ML-ER blend under\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/aa92452eb71e8d445e2b793a.jpg"},{"id":99750361,"identity":"6aadd7f3-127c-409a-92ed-764257d10d57","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":364708,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDeep learning-integrated collision management system for ZCM@LIL lamina.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e ML senor design and luminescence patterns under varied impact conditions and shooting equipment. \u003cstrong\u003eb\u003c/strong\u003e Data augmentation strategies and attention map analysis during deep learning training. \u003cstrong\u003ec-d\u003c/strong\u003e CNN architecture and confusion matrix for ML pattern classification. \u003cstrong\u003ee\u003c/strong\u003e Workflow of the vehicle collision emergency response system from ML signal capture to actuator control.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/7d525350451db0d42df46210.jpg"},{"id":104399382,"identity":"da0d514d-deac-4a73-922e-1892c3657eeb","added_by":"auto","created_at":"2026-03-11 12:05:50","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2077333,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/9517fa67-4d20-414d-9b31-32de26926e17.pdf"},{"id":99750347,"identity":"c68adb50-bb01-44ff-84f6-dee0f9363ae4","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"mp4","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":87457432,"visible":true,"origin":"","legend":"Supplementary Movie 1: Fabrication and ML performance of laser-induced luminophores","description":"","filename":"SupplementaryMovie1.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/b30352ab81f961c6c97575e0.mp4"},{"id":99750349,"identity":"e0f8a939-b4d7-4cdb-86c6-7d6cd44809fa","added_by":"auto","created_at":"2026-01-08 03:39:15","extension":"mp4","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2190800,"visible":true,"origin":"","legend":"Supplementary Movie 2: ML performance of ZCM@LIL-ER cylindrical blends","description":"","filename":"SupplementaryMovie2.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/6ba41cdb60d8ee38cfe78fc9.mp4"},{"id":99750362,"identity":"1caf23ff-88a8-4852-a8bf-16d44a72437c","added_by":"auto","created_at":"2026-01-08 03:39:16","extension":"mp4","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2037801,"visible":true,"origin":"","legend":"Supplementary Movie 3: ML performance of ZCM@LIL-ER cubic blends","description":"","filename":"SupplementaryMovie3.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/6a92bd687e88d1b26c17dff6.mp4"},{"id":99750342,"identity":"a76406e5-d072-47fa-b514-003f94b9a5ea","added_by":"auto","created_at":"2026-01-08 03:39:14","extension":"mp4","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":12283595,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Movie 4: Demonstration of the deep learning-driven control system\u003c/p\u003e","description":"","filename":"SupplementaryMovie4.mp4","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/4c0482ed8a0500fa8106fb86.mp4"},{"id":99750337,"identity":"701fc775-dcf7-4681-be04-a0bffc947bb5","added_by":"auto","created_at":"2026-01-08 03:39:13","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":2516378,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Information\u003c/p\u003e","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8106340/v1/26aadcdf0467a4caa7550684.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Mechanoluminescence by Direct Laser Writing: A Seconds-Scale Fabrication Strategy","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn the era of embodied intelligence and the Internet of Everything, \u003csup\u003e1,2\u003c/sup\u003e smart systems from wearable devices to humanoid robots increasingly require real-time sensing of mechanical stimuli such as impact and friction to execute autonomous \"sense-decide-act\" cycles.\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e Mechanoluminescent (ML) materials have emerged as a transformative sensing technology, offering a unique solution by directly converting mechanical energy into visible light. \u003csup\u003e6\u0026ndash;8\u003c/sup\u003e This photon-based signaling mechanism eliminates the need for external power, complex wiring, or intricate circuitry, enabling intuitive and visual monitoring of mechanical events. Since their initial development, \u003csup\u003e9\u003c/sup\u003e ML materials have diversified from single crystals to diverse systems encompassing organic emitters, \u003csup\u003e10\u003c/sup\u003e rare-earth-doped composites, \u003csup\u003e11\u003c/sup\u003e and transition metal-activated phosphors. \u003csup\u003e12\u003c/sup\u003e Their luminescent bands now span from ultraviolet to infrared, broaden their utility in bio and in vivo sensing. \u003csup\u003e13,14\u003c/sup\u003e Complementing this spectral versatility, their ultra-fast response down to microseconds and high spatial resolution far surpass those of conventional electrical sensors, \u003csup\u003e15,16\u003c/sup\u003e making them invaluable for dynamic impact detection and harsh-environment monitoring. \u003csup\u003e17,18\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eDespite these attractive features, the synthesis and integration of high-performance ML sensors present a major bottleneck. The prevailing method of high-temperature solid-state sintering typically requires prolonged annealing at temperatures exceeding 1000\u0026deg;C for serval hours, followed by grinding and blending with polymer matrices. \u003csup\u003e19\u003c/sup\u003e This process is energy-intensive and time-consuming, and often leads to grain coarsening or phase separation during compositing, degrading luminescence performance. Recent efforts to address these concerns have explored innovations in material synthesis and compositing processes. Alternative synthesis routes such as hydrothermal synthesis,\u003csup\u003e20\u003c/sup\u003e spray pyrolysis,\u003csup\u003e21\u003c/sup\u003e microwave-assisted synthesis,\u003csup\u003e22\u003c/sup\u003e and molten salt shielding method,\u003csup\u003e23\u003c/sup\u003e have shown promise in shortening synthesis time or lowering reaction temperatures. In parallel, integration strategies have evolved from simple bulk blending or sandwich configurations to more advanced techniques like 3D printing and electrospinning. \u003csup\u003e24\u0026ndash;26\u003c/sup\u003e By modulating carrier migration behavior at the crystal\u0026ndash;polymer interface, such approaches enable improved control over ML response modes, sensitivity ranges, and detection accuracy. \u003csup\u003e27\u003c/sup\u003e Nevertheless, limitations in energy efficiency, process complexity, or microstructure controllability persist, making a transformative manufacturing approach imperative.\u003c/p\u003e \u003cp\u003eHere, we present a direct laser writing strategy for fabricating ML materials. Although laser-based processing offers a compelling pathway for additive and subtractive manufacturing, \u003csup\u003e28,29\u003c/sup\u003e its potential for the direct synthesis of ML materials remains largely unexplored. Laser thermal processing (LTP) could, in principle, reduce processing times from hours to seconds. The central challenge lies in achieving the highly localized, extreme temperatures required for ML crystal formation within a brief irradiation window. In this work, we systematically validate the feasibility of LTP across multiple material systems, including ZnS/CaZnOS:Mn, SrZnOS:Mn, SrP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e7\u003c/sub\u003e:Mn, and Ca\u003csub\u003e9\u003c/sub\u003eAl(PO\u003csub\u003e4\u003c/sub\u003e)\u003csub\u003e9\u003c/sub\u003e:Mn. By combining numerical simulation with experimental analysis, we characterize the morphology, crystal structure, photoluminescence, and mechanoluminescence properties of the laser-induced luminophores and compare them with their conventionally sintered counterparts. Capitalizing on the in-situ patterning capability of LTP, we further develop an intelligent vehicle collision management system that integrates deep learning-based recognition of ML patterns to enable automated emergency response, thereby closing a full \"sense\u0026ndash;decide\u0026ndash;act\" cycle. We believe this laser writing strategy will pave the way for future applications, such as structural health monitoring of laser-welded components and the rapid screening of novel ML phosphors.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eFabrication of Laser-induced Luminophores\u003c/p\u003e\n\u003cp\u003eConventional high temperature solid-state sintering (HTSS) typically requires prolonged heat preservation at temperatures exceeding 1000\u0026deg;C, relying on thermally activated atomic diffusion, reaction, nucleation, and crystal growth of precursor powders to yield dense mechanoluminescent (ML) crystals. In contrast, this work pioneers employing laser thermal processing (LTP) to prepare ML crystals, as schematically illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ea. The principle of LTP leverages localized high temperatures generated by laser irradiation to achieve in-situ densification and crystallization of precursor powders. Distinct from the sintering methods, the \u0026ldquo;top-down\u0026rdquo; LTP approach enables energy-efficient fabrication with high flexibility in pattern design via programmable laser scanning. The process initiates with mixing raw powders, which are then blade-coated into a uniform precursor film. Upon laser scanning, the film rapidly transforms into colored crystals within seconds. The resulting laser-induced luminophores (LILs) exhibit strong interparticle adhesion due to nucleation and growth, allowing them to be readily lifted from the surrounding loose precursor matrix using tweezers (Supplementary Fig.\u0026nbsp;1). It is worth noting that laser-induced thermal stress can cause partial powder detachment, resulting in irregular surface dispersion; however, this can be easily removed by gentle air blowing using an ear aspirator bulb. The incorporation of an optimized amount of B₂O₃ particles into the precursor effectively suppresses powder detachment during LTP (Supplementary Fig.\u0026nbsp;2).\u003c/p\u003e\n\u003cp\u003eTo elucidate the mechanism of LIL formation, we perform finite element method (FEM) simulations to analyze the influence of laser power and scanning speed on surface temperature and thermal stress distribution, using CaZnOS as a model system (simulation details are provided in Supplementary Note 1). As depicted in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb, under stationary laser irradiation at 10 W, a maximum surface temperature of 1840 K is achieved locally, well above the phase transition temperature of CaZnOS.\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e This confirms that the laser energy is sufficient to crystallize the precursor into ML-active CaZnOS. Notably, the temperature drops rapidly with distance from the laser spot (reaching only\u0026thinsp;~\u0026thinsp;470 K at 1 cm away), underscoring the highly localized heating that enables both efficient crystallization and high-precision patterning. We further simulated axial surface temperature profiles under varying laser powers (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ec). The results indicate a significant increase in local maximum temperature with increasing power, reaching 3700 K at 50 W. Under dynamic scanning conditions (i.e., laser writing), the peak temperature is lower than that under stationary irradiation; however, a distinct \u0026ldquo;temperature tail\u0026rdquo; is observed, where the temperature decays gradually post-irradiation. This tail mimics the heat preservation stage in HTSS and is essential for complete precursor crystallization. Additionally, we analyzed laser-induced thermal stress and its effect on LIL morphology. As illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ed, at 10 W and 0.8 cm/s, the maximum thermal stress reaches 2.42 MPa, inducing a surface deformation of up to 525 \u0026micro;m. Further simulations show that higher laser power or slower scanning speed markedly increases the maximum thermal stress. These findings demonstrate that localized temperature and stress in LTP not only drive LIL formation but also critically influence surface morphology, as discussed in later sections.\u003c/p\u003e\n\u003cp\u003eThe FEM simulations confirm that LTP allows precise control over local temperature through adjustment of laser power and scanning speed, thereby accommodating the thermal requirements of various ML materials. To validate this versatility, we evaluate the ML performance of serval Mn-doped ML systems fabricated under identical LTP parameters, such as Ca\u003csub\u003e9\u003c/sub\u003eAl(PO\u003csub\u003e4\u003c/sub\u003e)\u003csub\u003e9\u003c/sub\u003e:Mn, SrP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e7\u003c/sub\u003e:Mn, SrZnOS:Mn, and ZnS/CaZnOS:Mn. For ML testing, a polyethylene terephthalate (PET) film is applied to the sample surface after fabrication, and ML emission is triggered by rubbing the PET with a glass rod. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ee, the laser-scanned region of Ca\u003csub\u003e9\u003c/sub\u003eAl(PO\u003csub\u003e4\u003c/sub\u003e)\u003csub\u003e9\u003c/sub\u003e:Mn appears light grey and exhibits deep red persistent photoluminescence (Pers PL) under 254 nm UV excitation, with a decay lifetime of ~\u0026thinsp;0.5 s, but no detectable ML emission is observed. In the case of SrP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e7\u003c/sub\u003e:Mn (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ef), brighter red Pers PL with a longer lifetime of ~\u0026thinsp;1.4s is observed under the same UV excitation; after PL completely fades, white ML emission is induced mechanically. In Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eg, SrZnOS:Mn exhibits a loose surface morphology after LTP, resulting in extremely weak ML emission. By contrast, ZnS/CaZnOS:Mn (ZCM) exhibits bright yellow mechanoluminescence (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eh). Notably, neither SrZnOS:Mn nor ZCM shows significant Pers PL. The entire fabrication and ML characterization process of ZCM@LIL is captured in Supplementary Movie 1. These results confirm the versatility of LTP for producing functional mechanoluminophores and underscore its potential to serve as an alternative to conventional HTSS. Future efforts will focus on optimizing laser parameters and material compositions to further improve ML performance.\u003c/p\u003e\n\u003cp\u003eStructural and Optical Characterization of the Mechanoluminephores\u003c/p\u003e\n\u003cp\u003eGiven the superior mechanoluminescence brightness of ZnS/CaZnOS:Mn among the LTP-fabricated systems, we systematically investigate the structural and optical properties of ZnS:Mn, CaZnOS:Mn, and their hybrid ZCM system. During LTP, both CaZnOS:Mn and ZCM develops three-dimensional porous architectures resembling \u0026ldquo;ginger-like\u0026rdquo; morphologies (Supplementary Fig. 3 and Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea). We attribute this unique structure primarily to the decomposition of CaCO\u003csub\u003e3\u003c/sub\u003e, a common precursor in both systems, which releases CO\u003csub\u003e2\u003c/sub\u003e gas at 625\u0026deg;C.\u003csup\u003e31\u003c/sup\u003e Combined with localized laser-induced thermal stress, this gas evolution facilitates precursor crystallization and generates a fluffy, interconnected porous network. Although MnCO\u003csub\u003e3\u003c/sub\u003e decomposes at 300\u0026ndash;350\u0026deg;C,\u003csup\u003e32\u003c/sup\u003e its low doping content is fails to induce detectable porosity in the ZnS:Mn system. As depicted in Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea, ZCM@LIL appears yellow under daylight and emits orange photoluminescence under 256 nm UV excitation, with surface shadows highlighting its porous architecture. High-magnification SEM imaging reveals the hierarchical nature of this porosity. At 20 W laser power (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb, c), ZCM@LIL exhibits well-defined needle-like pores with uniform connectivity. At 30 W (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ed), localized melting leads to irregular pore boundaries, and at 40 W (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ee), extensive melting results in smoother surfaces and markedly reduced porosity. This trend is governed by the coupled effects of maximum temperature and thermal stress, where higher power promotes particle fusion and pore collapse. In Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ef, EDS elemental mapping confirms homogeneous distribution of Ca, Zn, O, S, and Mn in ZCM@LIL prepared at 40 W; Mn appears less uniform in highly undulated regions due to its low concentration. High-resolution small-area EDS further verifies successful synthesis and elemental distribution (Supplementary Fig. 4). To quantify porosity, we analyze the area of individual/non-connected pores in ZCM@LILs prepared at 20\u0026ndash;50 W (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eg). The average pore area decreases from ~\u0026thinsp;0.05 mm\u0026sup2; at 20 W to ~\u0026thinsp;0.03 mm\u0026sup2; at 50 W, underscoring the ability to tune microstructure via laser power for optimized ML performance.\u003c/p\u003e\n\u003cp\u003eTo verify the consistency in phase composition, we preform the XRD characterization on ZnS:Mn, CaZnOS:Mn, and ZCM systems prepared by LTP and HTSS (Supplementary Fig. 5). For ZnS:Mn, diffraction peaks of LILs fabricated across 10\u0026ndash;40 W closely match hexagonal wurtzite ZnS (PDF#36-1450), with no cubic sphalerite phase (PDF#05-0566) detected. The transition from sphalerite to wurtzite occurs near ~\u0026thinsp;1020\u0026deg;C; \u003csup\u003e33\u003c/sup\u003e the observed wurtzite dominance confirms that LTP locally exceeds this temperature, despite its short processing time, enabling formation of the thermodynamically stable phase preferred for ML applications. Minor residual precursor peaks (e.g., MnCO\u003csub\u003e3\u003c/sub\u003e) are occasionally detected due to sampling of LILs from the precursor matrix. For CaZnOS:Mn, LIL diffraction patterns align well with hexagonal CaZnOS (PDF#01-076-3819) across the same power range. Trace CaO peaks appear at 32.20\u0026deg;, 37.36\u0026deg;, and 53.86\u0026deg;, consistent with partial decomposition above 1370 K, \u003csup\u003e34\u003c/sup\u003e along with slight CaCO₃ residues (PDF#00-005-0586). In ZCM, LIL XRD profiles match those of HTSS samples, confirming coexistence of wurtzite ZnS and hexagonal CaZnOS. Although LIL diffraction peaks are less intense than HTSS counterparts (likely due to shorter crystal growth during LTP\u0026rsquo;s rapid heating/cooling cycle), their identical peak positions verify that LTP successfully synthesizes target crystal phases for all three systems, laying the foundation for subsequent optical analysis.\u003c/p\u003e\n\u003cp\u003eFurthermore, we conduct PL measurements on the three systems prepared by LTP and HTSS (Supplementary Fig. 6). With the emission wavelength fixed at 585 nm, the photoluminescence excitation (PLE) peak positions of ZnS:Mn differ minimally between LTP (346.3 nm) and HTSS (347.8 nm). In contrast, CaZnOS:Mn and ZCM exhibit\u0026thinsp;~\u0026thinsp;10 nm blueshifts: CaZnOS:Mn peaks at 340.2 nm (LTP) and 351.5 nm (HTSS), and ZCM at 342.2 nm (LTP) and 352.1 nm (HTSS). These differences likely stem from process-dependent variations in crystal structure, doping environment, and defect distribution, altering excitation paths and energy-level transitions. In addition, under 342 nm excitation, ZnS:Mn emission peaks remain similar (588.5 nm for LTP vs. 591.0 nm for HTSS), whereas CaZnOS:Mn and ZCM exhibit noticeable emission shifts, consistent with PLE trends. To explore luminescence dynamics, we plot PL decay curves in logarithmic coordinates and fit them with a bi-exponential model (\u003cimg style=\"width: 199px;\" src=\"data:image/png;base64,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\" alt=\"\" /\u003e). Calculated fluorescence lifetimes reveal that all LIL samples have shorter lifetimes than HTSS counterparts. The ZCM system shows the most pronounced difference, with lifetimes decreasing from 727.3 ns (HTSS) to 424.9 ns (LIL), i.e., a 41.6% reduction. Shorter lifetimes typically indicate enhanced non-radiative transitions or faster energy transfer, suggesting that LILs may offer improved response speed and dynamic performance for mechano-optical sensing and display applications.\u003c/p\u003e\n\u003cp\u003eBased on the SEM, XRD, and PL analyses, the LTP process is confirmed to successfully synthesize crystalline phases in ZnS:Mn, CaZnOS:Mn, and ZCM systems. To evaluate their mechanoluminescent performance, the three LIL systems are subjected to grinding and blending treatments, followed by systematic ML characterization under handwriting friction, compression loading, and impact excitation. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea, the ground LIL powders are sandwiched between PET films. Rubbing the PET surface with a glass rod elicits bright ML emission visible to the naked eye. Under identical applied force, ML emission spectra of the three systems are measured by a spectrometer, revealing that their emission peaks are all centered at approximately 590 nm. Furthermore, CIE chromaticity coordinate analysis of the ML emission spectra from the three systems showed negligible chromaticity differences (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb), with ZnS:Mn@LIL (0.560, 0.442), ZCM@LIL (0.566, 0.436), and CaZnOS:Mn@LIL (0.551, 0.444), indicating a high degree of consistency in ML color emission. In line with previous work,\u003csup\u003e35\u003c/sup\u003e the ZCM@LIL system exhibit slightly higher luminescence intensity than the CaZnOS:Mn@LIL system and is significantly brighter than ZnS:Mn, with an intensity approximately 1.6 times higher than the latter. The inset in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea displays time-lapse images of the ZCM@LIL system during handwriting, captured using a smartphone. Based on these findings, the ZCM system is selected as the primary research object for subsequent ML performance studies. The ML mechanism is illustrated in the inset of Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb: when the ML crystals are subjected to mechanical stimulation, lattice distortion occurs, and the accumulation of piezoelectric charges induces band bending, thereby promoting carrier release and migration. The identical emission peak positions across the three systems confirm that their ML originates from the transition of \u003csup\u003e3\u003c/sup\u003ed\u003csup\u003e5\u003c/sup\u003e electrons in Mn\u003csup\u003e2+\u003c/sup\u003e from the \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003eT\u003csub\u003e1\u003c/sub\u003e to the \u003csup\u003e\u003cspan class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003eA\u003csub\u003e1\u003c/sub\u003e energy level. \u003csup\u003e36\u0026ndash;38\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo assess the bulk luminescence behavior of LILs, ZCM@LIL powders are blended with epoxy resin (ER) to form ML-ER blends. Under uniaxial compression testing (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ec), the ML intensity depends strongly on the compression rate. At a lower compression rate (1 mm/s), the ML intensity of the samples is relatively weak, and a pre-compression of approximately 3 mm is required to trigger stable luminescence. However, at a compression rate of 3 mm/s, the samples exhibit a ML response with enhanced intensity without pre-compression, further demonstrating the force-light response capability of the LIL system. The embedded figure shows the ML image at the peak intensity. When the compression rate increases to 5 mm/s, the ML intensity reaches its maximum, 2.49 times that at 1 mm/s. However, as the compression force further increases, the luminescence intensity rapidly decays, indicating that the ML luminescence intensity and response range are dependent on the rate of mechanical stimulation. The potential for stress propagation detection is evaluated using cylindrical ML-ER composites compressed at 1.67 mm/s (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ed). Peak ML intensity occurs at 8% strain, and the stripes fade after loading ceases. After unloading, surface deformation marks align with the ML stripe distribution, confirming that the ZCM@LIL\u0026ndash;ER composite visually maps stress propagation and strain distribution. Supplementary Movie 2 and 3 visually capture the dynamic evolution of ML emission in cylindrical and cubic composites, respectively, demonstrating the capability of the ZCM@LIL system to visualize localized stress distribution.\u003c/p\u003e\n\u003cp\u003eFinally, to demonstrate the practical utility of LTP for functional device design, we fabricate patterned mechanoluminescent composites and evaluate their spatial selectivity and impact response. As schematically illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ee, a flexible ML-ER lamina is prepared by drop-casting epoxy resin onto a laser-patterned LIL film, followed by curing and peeling from the substrate. High-speed imaging during iron ball impact reveals spatially selective mechanoluminescence exclusively from the laser-written \u0026ldquo;crosshair\u0026rdquo; region (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ef), confirming the patterning precision of LTP and its potential for spatially resolved stress sensing. We further compare the impact-induced ML performance of non-patterned ZCM@LIL-ER laminas with conventional HTSS-derived counterparts by varying the drop height of the iron ball. As summarized in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eg, the ZCM@LIL system exhibits lower overall ML intensity across all tested heights, which we attribute to its inherent porous architecture and lower particle packing density relative to the densely sintered HTSS samples. Insets in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eg show representative emission images at impact heights ranging from 5 to 40 cm. Notably, linear fitting of ML intensity versus impact force reveals that the ZCM@LIL composite exhibits a significantly steeper slope compared to ZCM@HTSS (1.11 vs. 0.54), indicating superior force\u0026ndash;light sensitivity. This enhanced linearity suggests that the porous LIL structure undergoes structural collapse and facilitates efficient stress transfer under high impact, amplifying luminescence output. These results underscore the potential of ZCM@LIL-ER laminas as visual impact sensors with high discriminative capacity across a broad dynamic range, providing a foundation for applications in stress mapping and impact monitoring.\u003c/p\u003e\n\u003cp\u003eDeep Learning-Driven Intelligent System via LIL Sensors\u003c/p\u003e\n\u003cp\u003eTo demonstrate a \"sense-decide-act\" cycle using LIL sensors, we develop an intelligent vehicle collision management system that integrates programmable ZCM@LIL sensors, convolutional neural network (CNN)-based optical analysis, microcontroller signal processing, and WiFi-enabled actuator control. This closed-loop framework accurately identifies impact locations through mechanoluminescent pattern recognition and facilitates rapid automated countermeasures to mitigate collision consequences. As illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea, ZCM@LIL sensors are patterned on PET substrates using trajectory-programmable LTP. Four alphabetic characters, namely \u0026ldquo;W\u0026rdquo;, \u0026ldquo;S\u0026rdquo;, \u0026ldquo;A\u0026rdquo;, and \u0026ldquo;D\u0026rdquo;, are designed to represent front, rear, left, and right impact positions, respectively. After laser-induced crystallization, ER is applied to the LIL surface and cured, forming a protective layer that improves mechanical robustness while maintaining high optical transparency for clear ML signal transmission. A comprehensive dataset is constructed by recording ML images under varied impact conditions, including different ball masses, drop heights, and imaging setups. Representative ML images for each character are shown in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea, with full datasets and detailed network architecture provided in Supplementary Note 2.\u003c/p\u003e\n\u003cp\u003eTo enhance model robustness, ML images undergo feature enhancement and data augmentation (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb). Attention maps visualize the model's focus regions during classification, revealing that raw ML images exhibit weak or ambiguous features. Image sharpening using a Laplacian operator improves contour clarity, and binarization further enhances character distinctness. To address the limited sample size (15 ML images per letter class, 60 in total), data augmentation techniques including salt-and-pepper noise and random erasing expand the dataset to 3600 images, significantly improving generalization. A four-stage CNN architecture (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ec) processes 224\u0026times;224 pixel inputs through convolutional layers with 3\u0026times;3 kernels, BatchNorm2d, LeakyReLU (\u0026alpha;\u0026thinsp;=\u0026thinsp;0.1), and 2\u0026times;2 max-pooling. This progressively reduces spatial dimensions from 224 to 14 while increasing channels from 32 to 256, enabling hierarchical feature extraction from local edges to global structures. Dropout (rates: 0.5, 0.5, 0.4) and BatchNorm1d layers before fully connected modules help suppress non-critical activations, effectively implementing an implicit spatial attention mechanism. Feature maps are resized via AdaptiveAvgPool2d (7\u0026times;7) and mapped through three linear layers (output sizes: 512, 256, and 4), yielding classification probabilities for \u0026ldquo;A\u0026rdquo;, \u0026ldquo;D\u0026rdquo;, \u0026ldquo;S\u0026rdquo;, and \u0026ldquo;W\u0026rdquo;. The confusion matrix (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ed) shows 100% accuracy for \u0026ldquo;D\u0026rdquo;, \u0026ldquo;S\u0026rdquo;, and \u0026ldquo;W\u0026rdquo;, and 95.7% for \u0026ldquo;A\u0026rdquo; (4.3% misclassified as \u0026ldquo;W\u0026rdquo;), with an overall accuracy of 98.93%.\u003c/p\u003e\n\u003cp\u003eThe full system architecture is illustrated in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ee. ZCM@LIL-ER laminas labeled \u0026ldquo;W\u0026rdquo;, \u0026ldquo;S\u0026rdquo;, \u0026ldquo;A\u0026rdquo;, and \u0026ldquo;D\u0026rdquo; are installed inside the front, rear, left, and right sections of a vehicle model. A darkroom environment enhances ML signal detection. Upon collision, an embedded camera captures real-time ML emission from the impacted region. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ea, ML images acquired under varying impact conditions simulate real-world collisions with differing locations and force intensities. A microcontroller unit processes the ML images, converts them to binary format, and classifies the impact location using the embedded CNN. Results are transmitted via WiFi to the central control system, triggering actuators such as alarms, brake assistance, steering correction, or airbag alerts. To verify the functionality of the integrated system, we present a simplified demonstration using a model car's motor/steering control as an example (Supplementary Movie 4). This system can achieve four emergency avoidance operations based on the luminescent images. for example, advancing forward in response to a rear impact (\u0026ldquo;W\u0026rdquo;), or moving forward while turning left in response to a right-side impact (\u0026ldquo;A\u0026rdquo;). As shown in Video S4, the integrated system exhibits an average response time of ~\u0026thinsp;150 ms from ML luminescence image reading to motor/steering control, demonstrating real-time capability. This LIL\u0026ndash;deep learning integrated system offers a novel strategy for intelligent collision management, highlighting high sensitivity, rapid response, and applicability in smart sensing and human\u0026ndash;machine interaction.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn summary, this work establishes a versatile laser writing strategy for the rapid in-situ fabrication of mechanoluminophores, exemplified by the synthesis of SrP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e7\u003c/sub\u003e:Mn, SrZnOS:Mn, and ZnS/CaZnOS:Mn. By replacing energy-intensive and prolonged sintering with CO\u003csub\u003e2\u003c/sub\u003e laser writing strategy, we achieve crystallization within seconds while yielding porous microstructures favourable for stress\u0026ndash;light coupling. Structural and optical analyses confirm that the laser-induced luminophores maintain crystalline phases analogous to sintered benchmarks, yet exhibit a 41.6% reduction in fluorescence lifetime, enabling accelerated luminescence dynamics. More importantly, the LIL-ER composites display intense orange mechanoluminescence with significantly enhanced linear response to stress, offering improved sensing accuracy. Capitalizing on the top-down patterning capability of this approach, we further construct programmable ML sensors and integrate them into a deep learning-augmented collision management platform. A customized CNN architecture trained on an augmented dataset achieves 98.93% classification accuracy for impact localization. The full-system demonstration of incorporating the ML sensing lamina, microcontroller, and wireless communication enables real-time collision mapping and triggers appropriate emergency responses. Our study not only introduces a general and efficient toolbox for developing novel mechanoluminophores and high-resolution luminescent patterning, but also illustrates a seamless materials-algorithm co-design framework toward intelligent autonomous systems.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eChemicals and materials\u003c/p\u003e \u003cp\u003eZnS, CaCO\u003csub\u003e3,\u003c/sub\u003e MnCO\u003csub\u003e3\u003c/sub\u003e, SrCO\u003csub\u003e3\u003c/sub\u003e, H\u003csub\u003e6\u003c/sub\u003eNO\u003csub\u003e4\u003c/sub\u003eP, Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e, epoxide resin, aromatic amine-type curing agent, anhydrous ethanol, are obtained from Shanghai Macklin Biochemical Co., Ltd. All materials are used as received without further purification. For Mn-doped samples, including ZnS:Mn, CaZnOS:Mn, SrZnOS:Mn, SrP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e7\u003c/sub\u003e:Mn and Ca\u003csub\u003e9\u003c/sub\u003eAl(PO\u003csub\u003e4\u003c/sub\u003e)\u003csub\u003e9\u003c/sub\u003e:Mn, the Mn doping level was maintained at 1 mol%. For ZnS/CaZnOS:Mn system, ZnS, CaCO\u003csub\u003e3\u003c/sub\u003e, and MnCO\u003csub\u003e3\u003c/sub\u003e are proportioned at 3:2:2%.\u003csup\u003e35\u003c/sup\u003e After proportioning according to the corresponding molar ratios, the raw materials are transferred to an agate mortar, an appropriate amount of absolute ethanol is added, and the mixture is ground for thirty minutes until the raw materials are uniformly mixed. Subsequently, it is placed in an oven and dried at 60\u0026deg;C for 4 hours for later use.\u003c/p\u003e \u003cp\u003eLaser thermal processing\u003c/p\u003e \u003cp\u003eThe dried precursor mixture is uniformly coated onto a glass slide using a blade coater. A CO\u003csub\u003e2\u003c/sub\u003e laser system (Gd Han\u0026rsquo;s Yueming Laser Group Co., Ltd.) with a fixed focal length of 8 mm is employed for irradiation. Laser-induced luminophores are fabricated in ambient atmosphere by scanning the laser beam over the precursor film according to pre-defined patterns, with laser power and scanning speed adjusted as needed.\u003c/p\u003e \u003cp\u003eHigh-temperature solid-state sintering\u003c/p\u003e \u003cp\u003eThe dried precursor mixtures were placed in an alumina crucible and transferred to a tube furnace (Hefei Kejing Materials Technology Co., Ltd.). Under a continuous argon flow, the temperature was raised to 1050\u0026deg;C at a rate of 10\u0026deg;C/min, held for 4 h, and then cooled naturally to room temperature to obtain the final ML crystals.\u003c/p\u003e \u003cp\u003eFabrication of ML samples preparation\u003c/p\u003e \u003cp\u003eFor PET-encapsulated samples (PET-ML-PET), ground ML crystals are being uniformly dispersed between two square PET sheets (0.5 mm thick) to form a sandwich structure. Mechanoluminescence (ML) is being excited by pressing or sliding a glass rod over the PET surface; the resulting emission is being recorded using a smartphone camera, and spectra are being acquired using a fiber-optic spectrometer. For ML-ER blends, epoxy resin and amine curing agent are being mixed at a mass ratio of 2:1. The resulting solution is being combined with ground ML crystals at a 1:2 mass ratio (resin:crystals), cast into a custom silicone mold, and cured at 80\u0026deg;C for 6 h before demolding. For ML-ER laminas, LILs are first being prepared on glass or PET substrates via LTP. A 3D-printed square frame mold (20 mm \u0026times; 20 mm \u0026times; 2 mm) is being fixed onto the sample surface, filled with the ER solution, and cured at room temperature for 24 h to form a protective, transparent encapsulation layer.\u003c/p\u003e \u003cp\u003eCharacterizations\u003c/p\u003e \u003cp\u003eThe XRD patterns are measured using a Bruker D8ADVANCE Multifunctional X-ray Diffractometer. SEM images of ML samples are obtained by a ThermoFisher VeriosG4UC. The photoluminescence fluorescence spectra are acquired via EDINBURGH-FLS 1000. An Ocean Optics QE65pro fiber optic spectrometer is used to collect the ML signals of the samples encapsulated by PET. Luminescence images were captured using a smartphone and analyzed in MATLAB to extract intensity values.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data are available in the main text or the supplementary information.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by General Program of the Fundamental Research Funds for the Central Universities (2023ZYGXZR052), Guangzhou Applied Basic Research Program Project (2024A04J3601), The 10th Youth Talent Cultivation Project of China Association for Science and Technology (YESS20240469). The authors acknowledge the support from the Guangdong Provincial Key Laboratory of Intelligent Disaster Prevention and Emergency Technologies for Urban Lifeline Engineering, and Dongguan University of Technology Analytical and Testing Center.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary Information\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eLi J et al (2025) AI-embodied multi-modal flexible electronic robots with programmable sensing, actuating and self-learning. Nat Commun 16:8818\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eObayashi N, Abdulali A, Iida F, Hughes J (2025) Embodied intelligence paradigm for human-robot communication. Sci Rob 10:eads8528\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim H et al (2024) Inherently integrated microfiber-based flexible proprioceptive sensor for feedback-controlled soft actuators. npj Flex Electron 8:15\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen C et al (2025) Bio-inspired multimodal soft actuator with environmental self-adaptation. Nat Commun 16:7630\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen S et al (2025) A sensor-actuator\u0026ndash;coupled gustatory interface chemically connecting virtual and real environments for remote tasting. Sci Adv 11:eadr4797\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFardian-Melamed N et al (2025) Infrared nanosensors of piconewton to micronewton forces. Nature 637:70\u0026ndash;75\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCasar JR et al (2025) Upconverting microgauges reveal intraluminal force dynamics in vivo. Nature 637:76\u0026ndash;83\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChang S et al (2025) Hair-Like Mechanoluminescent Structures with Ultralow Activation Threshold for Dynamic Force Sensing. Adv Mater, 2507634\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu C, Watanabe T, Akiyama M, Zheng X (1999) Artificial skin to sense mechanical stress by visible light emission. Appl Phys Lett 74:1236\u0026ndash;1238\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYin Y et al (2024) Force-triggered hypso-and bathochromic bidirectional fluorescence switching beyond 120 nm and its anticounterfeiting applications. Sci Adv 10:eadk5444\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang S et al (2024) Tunable Full-Color Mechanoluminescence in Rare Earth‐Doped Transparent Amorphous Glass. Adv Funct Mater 34:2404439\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang H et al (2025) Oscillatory mechanoluminescence of Mn2+-doped SrZnOS in dynamic response to rapid compression. Nat Commun 16:548\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng K et al (2025) A Universal Strategy for Multicolor Mechanoluminescence via Radiative Energy Transfer Based on Ultraviolet Mechanoluminescent Material Ca9Al (PO4) 7: Ce3+. Laser Photonics Rev, 2401524\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu S et al (2025) Bright Chromium-Sensitized Lanthanide NIR‐II Mechanoluminescence in a Piezoelectric Oxide. Adv Mater, e06957\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen C et al (2023) Revealing the Intrinsic Decay of Mechanoluminescence for Achieving Ultrafast-Response Stress Sensing. Adv Funct Mater 33:2304917\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang X et al (2017) Full Dynamic-Range Pressure Sensor Matrix Based on Optical and Electrical Dual-Mode Sensing. Adv Mater 29:1605817\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou B et al (2023) Mechanoluminescent-triboelectric bimodal sensors for self-powered sensing and intelligent control. Nano-Micro Lett 15:72\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHou B et al (2022) An interactive mouthguard based on mechanoluminescence-powered optical fibre sensors for bite-controlled device operation. Nat Electron 5:682\u0026ndash;693\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao W et al (2025) Flexible mechano-optical sensors from mechanoluminescence to mechanoplasmonics: designs, applications, and prospects. J Mater Chem C 13:2058\u0026ndash;2090\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z, Tai Y, Nam J, Yin Y (2023) Calcination-induced transformation of ZnS: Mn2\u0026thinsp;+\u0026thinsp;nanorods to microparticles for enhanced mechanoluminescence. Chem Mater 35:6845\u0026ndash;6852\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePugina RS et al (2021) Nd3+: YAG microspheres powders prepared by spray pyrolysis: Synthesis, characterization and random laser application. Mater Chem Phys 269:124764\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUgbo FC et al (2023) Optimizing the Mechanoluminescent Properties of CaZnOS: Tb via Microwave-Assisted Synthesis: A Comparative Study with Conventional Thermal Methods. Materials 16:3511\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu M et al (2023) Novel Red Mechanoluminescence in Mn-Doped ZnGa2S4 Crystal Phosphors via a Molten Salt Shielding Method. Laser Photonics Rev 17:2300517\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHajra S et al (2023) Flexible composite material for self-powered applications via triboelectricity and mechanoluminescence: PDMS/ZnS: Cu composites. Nano Energy 114:108668\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZheng T et al (2023) Mechanoluminescence and photoluminescence heterojunction for superior multimode sensing platform of friction, force, pressure, and temperature in fibers and 3D-printed polymers. Adv Mater 35:2304140\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z, Tai Y, Ye Z, Nam J, Yin Y (2024) Integration of ZnS: Mn2⁺ Microparticles into Electrospun PVDF-Based Nanofibers for Enhanced Mechanoluminescence. Adv Funct Mater 34:2410358\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJeong HI et al (2025) Super elastic and negative triboelectric polymer matrix for high performance mechanoluminescent platforms. Nat Commun 16:854\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen Z et al (2025) 25 nm-Feature, 104-aspect-ratio, 10 mm2-area single-pulsed laser nanolithography. Nat Commun 16:7434\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan J et al (2025) Non-contact ultrasound to assist laser additive manufacturing. Nat Commun 16:7613\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J-C et al (2013) An intense elastico-mechanoluminescence material CaZnOS: Mn2\u0026thinsp;+\u0026thinsp;for sensing and imaging multiple mechanical stresses. Opt Express 21:12976\u0026ndash;12986\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhuang D, Chen Z, Sun B (2025) Thermal Decomposition of Calcium Carbonate at Multiple Heating Rates in Different Atmospheres Using the Techniques of TG, DTG, and DSC. Crystals 15:108\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKissinger H, McMurdie H, Simpson B (1956) Thermal decomposition of manganous and ferrous carbonates. J Am Ceram Soc 39:168\u0026ndash;172\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoore D, Wang ZL (2006) Growth of anisotropic one-dimensional ZnS nanostructures. J Mater Chem 16:3898\u0026ndash;3905\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGulyaeva R, Selivanov E, Vershinin A, Chumarev V (2006) Thermal expansion of CaZnSO. Inorg Mater 42:897\u0026ndash;900\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeng D et al (2020) A ZnS/CaZnOS heterojunction for efficient mechanical-to‐optical energy conversion by conduction band offset. Adv Mater 32:1907747\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMukhina MV et al (2021) Single-particle studies reveal a nanoscale mechanism for elastic, bright, and repeatable ZnS: Mn mechanoluminescence in a low-pressure regime. ACS Nano 15:4115\u0026ndash;4133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z et al (2024) Stress-induced multi‐stimulus‐responsive mechanoluminescence in Mn2\u0026thinsp;+\u0026thinsp;doped double perovskite compound. Adv Opt Mater 12:2301796\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y et al (2025) Smart Semiconductors: An Overview of AZnOS (A\u0026thinsp;=\u0026thinsp;Ca, Sr, Ba) Based Phosphors, Multi-Stimuli Responsive Luminescence, and Potential Application Prospects. Adv Sci\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8106340/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8106340/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eInorganic mechanoluminescent materials, which transduce mechanical force into light, are promising for power-free sensing, structural health monitoring and human-machine interfaces. However, their fabrication typically requires energy-intensive, protracted bulk synthesis methods such as solid-state sintering. Here, we report a rapid and in-situ laser writing strategy for fabricating mechanoluminophores, validated across multiple material systems. Using the classic ZnS/CaZnOS:Mn\u003csup\u003e2+\u003c/sup\u003e system, we demonstrate that the laser-induced luminophores retain the crystal structure and emission of their sintered counterparts, while exhibiting porous microstructures, shortened fluorescence lifetime (424.9 vs. 727.3 \u0026micro;s) and superior mechanoluminescent linearity with stress (fitting slope of 1.11 vs. 0.54). We leverage the top-down programmability of this approach to fabricate patterned mechanoluminescent sensors and demonstrate a deep learning-driven collision management system. Our work provides a general toolbox that accelerates the trial-and-error cycle of novel mechanoluminophores and enables high-precision luminescent patterning and on-demand sensor integration.\u003c/p\u003e","manuscriptTitle":"Mechanoluminescence by Direct Laser Writing: A Seconds-Scale Fabrication Strategy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-08 03:38:38","doi":"10.21203/rs.3.rs-8106340/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"04b59d99-dccb-44bb-9832-5d50319c0c04","owner":[],"postedDate":"January 8th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":60688705,"name":"Physical sciences/Materials science/Materials for optics"},{"id":60688706,"name":"Physical sciences/Optics and photonics/Optical techniques"}],"tags":[],"updatedAt":"2026-02-27T15:50:50+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-08 03:38:38","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8106340","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8106340","identity":"rs-8106340","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}