Plasmonic nanocomposite helices for weather-adaptive LiDAR function Authors: JuHyeong Lee, Gyurin Kim, Doeun Kim, Hyun Min Kim, Jang-Hwan Han, Hyeon-Ho Jeong Published: 2026 ABSTRACT AbstractReliable light detection and ranging (LiDAR) is often challenged by multiscale water droplets on protective covers, as macroscopic raindrops cause refraction and diffraction, while microscopic fog condensation scatters light, distorting returned signals. Here, we present a bioinspired strategy for multiscale droplet clearance using plasmonic nanocomposite helices that couple passive photothermal antifogging with pressure-stable hydrophobic water repellence. Inspired by penguin feather barbules, copper nanoparticles are embedded within three-dimensional silica nanohelices via glancing angle co-deposition. The copper nanoparticles provide visible-light plasmonic heating while preserving >85% transmittance at 905 nm, and the helical architecture imparts hierarchical roughness, yielding a 143° contact angle. Under 1 sun illumination, the surface shows a 9.3°C temperature rise, clearing condensation within 6 seconds. During natural rainfall (3-5 mm/h), LiDAR transmission through the plasmonic nanohelices remains at 100%, whereas bare glass drops to 70% after 5 minutes. These results demonstrate weather-adaptive LiDAR operation for robust autonomous sensing.Subject terms: Metamaterials, Nanoparticles Introduction Just as raindrops blur human vision, fog and rain can obscure light detection and ranging (LiDAR) sensors in autonomous vehicles, distorting their ability to perceive and navigate safely1–3. LiDAR is a cornerstone of vision-based sensing systems for autonomous driving, yet moisture accumulation on sensor covers severely degrades optical signals (Fig. 1). The effect depends strongly on droplet size, i.e., microscopic fog droplets scatter incident beams and blur returns4, while macroscopic raindrops refract/diffract and attenuate light5, both leading to degraded imaging and compromised localization accuracy. Because perception underpins the entire autonomous decision-making pipeline, even small signal distortions can propagate through planning and control, ultimately increasing the risk of hazardous malfunction. These challenges highlight the urgent need for a robust strategy that can adaptively clear moisture across multiple droplet scales. Challenges to LiDAR operation under adverse weather conditions. a Schematic illustration of safety risks caused by water droplet accumulation on the LiDAR window or protective cover. b Schematic illustration of multi-scale moisture formation on car window (top). Penetration depth of near-infrared (NIR) light through water, showing absorption-limited propagation (bottom)70. Additional NIR signal distortions due to different water droplet sizes, (i) refraction and diffraction from macroscopic raindrops and (ii) scattering from microscopic fog droplets. c LiDAR vision images through bare glass (left) and plasmonic nanohelices (right), obtained under simulated weather conditions: clean (top), rainy (middle), and foggy (bottom) environments. d Corresponding pixel intensity profiles normalized to the clean condition, showing progressive signal degradation under moisture exposure for bare glass, whereas the proposed plasmonic nanocomposite helices surface maintains stable signal transmission under the same conditions. Photothermal coatings provide a power-free route by converting sunlight into localized heat6–8. Carbon-based absorbers such as graphene9–11 and carbon nanotubes12–14, as well as MXenes15–17 and polymeric films18–21, demonstrate strong light-to-heat conversion, often producing surface temperature rises of 10–30°C under 1 sun illumination. However, their absorption is typically broadband (> 500 nm wavelength range), which reduces transparency at the 905 nm LiDAR band, where transmission frequently drops below ~50%22–25. This spectral non-selectivity limits their compatibility with LiDAR sensors. By contrast, plasmonic heating enables wavelength-selective absorption26,27. Metallic nanostructures can convert visible sunlight into heat while preserving near-infrared (NIR) transparency. For example, titania thin films with percolated Au layers have demonstrated plasmonic antifogging26, but their performance is largely confined to fog-scale condensates (≤ 20 µm) and insufficient against larger adhered droplets28. Subsequent approaches that integrate plasmonic films with micro/nano-structured surfaces have achieved water contact angles above 150°, roll-off angles below 10°, and defogging within tens of seconds under ~1 sun29–32. Despite these advances, two critical limitations remain, (i) most designs are not spectrally optimized for LiDAR, risking attenuation or back-reflection at 905 nm, and (ii) hydrophobic and photothermal functions are mostly separated into multilayers29–31, increasing thickness, cost, and interfacial losses that hinder practical integration. Here, we present plasmonic nanocomposite helices that unify both functions within a single nanostructured coating. Inspired by penguin feather barbules, which combine melanosome absorption with hierarchical keratin architectures for water repellence, our design consists of Cu nanoparticles (< 10 nm) embedded within three-dimensional silica nanohelices, fabricated by glancing angle co-deposition. Embedding Cu in silica stabilizes the nanoparticles against oxidation, overcoming a longstanding barrier to their use as plasmonic absorbers. The resulting surfaces combine visible-light-driven photothermal antifogging with pressure-stable hydrophobicity, exhibiting a water contact angle of 143°, maintaining >80% transmittance at 905 nm, and clearing condensation within 6 s under solar illumination. In outdoor LiDAR field tests, they maintain 100% signal transmission under rainfall, whereas bare glass degrades to 70% after 5 min. These results establish plasmonic nanocomposite helices as a scalable and robust platform for weather-adaptive LiDAR optics, with broad potential in autonomous vehicles, robotics, and outdoor sensing applications. Results and discussion Concept of bioinspired plasmonic nanocomposite helices The concept of our plasmonic nanocomposite helices draws inspiration from the hierarchical structures of penguin feather barbules, which combine embedded melanosomes (acting as light absorbers) with a branched keratin architecture to achieve both water repellence and thermal regulation in extreme environments (Fig. 2a)33. In species such as Pygoscelis papua, body feathers contain micron to centimeter scale densely branched hair-like barbules. These exhibit static contact angles with water in the range of 141–148°, enabling droplets with diameters between 10 and 80 µm to detach and self-propel during coalescence. Although this does not satisfy the conventional criterion for superhydrophobicity, ecologically, such a structure can still prevent continuous water film formation and subsequent ice transition in cold environments by suppressing liquid penetration and minimizing droplet-surface contact time, thereby helping to maintain a pressure-stable wetting condition34. In addition, colored feathers, often located on the back, contain melanosomes that absorb ultraviolet radiation and dissipate it as heat, assisting thermoregulation by influencing heating and cooling rates in different environments. This biological synergy between droplet repellence and light-to-heat conversion forms the foundation of our device design. Bioinspired concept of plasmonic nanocomposite helices. a Hierarchical structure of penguin (Pygoscelis papua) feathers, showing branched barbules at the microscale (top) and melanosomes functioning as light-absorbing inclusions at the nanoscale (bottom). b Schematic translation of these dual functions into an artificial design. SiO2 micro-helices provide surface roughness for hydrophobicity, while embedded Cu nanoparticles act as plasmonic absorbers, generating visible-light-driven photothermal heating while maintaining NIR transparency. c Integrated mechanism of the bioinspired architecture, where the helical surface geometry repels macroscopic water droplets (pressure-stable hydrophobicity) and sunlight-induced plasmonic heating removes condensed microscopic droplets (antifogging). Translating this biological function into a synthetic platform, we introduce artificial structures consisting of three-dimensional silica nanohelices embedded with metallic nanoparticles (NPs), acting as artificial light absorbers (Fig. 2b). The helical geometry is selected as a practical three-dimensional anisotropic architecture enabled by the fabrication process, rather than as a uniquely required structure. These metal@oxide nanocomposite helices provide two distinct functions (Fig. 2c), (i) pressure-stable hydrophobicity arising from the hierarchical roughness of the helical glass array, and (ii) localized surface plasmon resonance (LSPR) of the embedded metal NPs for photothermal antifogging. The LSPR effect provides wavelength-selective absorption, where resonant excitation induces collective electron oscillations that relax non-radiatively within picoseconds, releasing energy as heat7. This photothermal heating accelerates the evaporation of condensed droplets and suppresses fog formation. Simultaneously, analogous to penguin feathers, the helical surface traps air pockets beneath water, prevents complete wetting, and enables droplets to detach with minimal adhesion. This dual functionality reflects the bioinspired combination of light absorption and water repellency observed in natural barbules, rather than implying direct structural equivalence. Given the structural complexity, we adopt a sequential design strategy, first optimizing the plasmonic response in planar metal@oxide composite films (Fig. 3a and Supplementary Figs. 1 and 2), then extending the optimized composition into the three-dimensional helical geometry. To identify the optimal plasmonic absorber in the visible range, we perform finite-difference time-domain (FDTD) simulations of metal NPs (Al, Mg, Cu, Ag, Au) embedded in an oxide matrix (initially SiO₂), under identical structural conditions (15% doping concentration, 50 nm thickness) (Fig. 3b and Supplementary Fig. 1). The results show strong material dependence, because Al and Mg confine absorption to the ultraviolet with low visible absorptance (< 25%), while Cu, Ag, and Au exhibit pronounced visible absorption with average absorptance values of 34%. Considering both spectral selectivity and material cost, we select Cu as the plasmonic material. We then examine the role of the dielectric host matrix by comparing SiO₂ (n ≈ 1.5), ZnO (n ≈ 1.9), and TiO₂ (n ≈ 2.5) (Fig. 3b and Supplementary Fig. 1). Higher-index hosts enhance visible absorption (from 35 to 42%), but simultaneously redshift the plasmonic resonance, significantly reducing NIR transmittance critical for LiDAR operation. For instance, transmission at 905 nm drops from 85% in SiO₂ to 25% in TiO₂. To preserve both visible absorption while maintaining NIR transparency above 85% (a practical threshold for LiDAR)35, we select Cu@SiO₂ as the composite material system. Design optimization of plasmonic nanocomposite helices. a Schematic of planar Cu@SiO2 plasmonic nanocomposite film, illustrating embedded metallic nanoparticles (~ 5 nm diameter) within the dielectric glass matrix (top). Corresponding optical near-field distributions at visible and NIR wavelengths, showing strong field confinement only in the visible range, confirming wavelength-selective plasmonic absorption (bottom). b Transmission spectra of (i) plasmonic nanocomposite films with various metallic nanoparticles embedded within SiO2 (left). (ii) Effect of host glass matrix with Cu nanoparticles (right). Corresponding 905 nm transmission and average visible absorptance (380–700 nm) vs. permittivity of metal nanoparticles and oxide material (each bottom). c Simulated 905 nm transmittance (top) and visible absorptance (bottom) maps as functions of Cu nanoparticle’s volume fraction (2.5–25%) and film thickness (50–400 nm). d Extension of the optimized composition into 3D helical geometries forming plasmonic nanocomposite helices (left). Corresponding optical near-field distributions in the visible (600 nm) and NIR (905 nm) range (right). e Simulated transmission spectra of plasmonic helices vs. planar films, revealing enhanced NIR transparency due to high porosity and antireflective effects. f Schematic representation of solid fraction within the helical array (left) and analytically calculated contact pressure of the plasmonic nanohelices and penguin feather (right). Next, we investigate the effects of film thickness (50–400 nm) and Cu NP doping concentration (2.5–25%) (Fig. 3c and Supplementary Fig. 3 for full spectra). Increasing either parameter strengthens visible absorption, for example, the 400 nm thick Cu@SiO2 film with 20% Cu NPs yields 86.5% visible absorptance, but also reduces NIR transmittance to 65%, compromising LiDAR compatibility. Conversely, low Cu NP doping (< 5%) results in insufficient absorption, requiring impractically thick films. Balancing these trade-offs, we determine an optimal condition at 10% Cu NP doping, which ensures selective visible absorptance (≥ 30%) while maintaining NIR transparency (≥ 85%) across a practical thickness (100–300 nm). To extend the optimized composition into mechanically functional architectures, we design helical nanostructures, which can be realized with fabrication conditions described below (Fig. 3d). Based on the optimized planar Cu@SiO2 nanocomposite, we further perform numerical simulations to examine the effects of helix height, number of turns, and Cu NP concentration in the three-dimensional plasmonic nanocomposite helix architecture (Supplementary Fig. 4). These parameters define a trade-off between visible photothermal absorption and LiDAR-compatible transmittance at 905 nm, yielding a practical design window of helix heights below ~1000 nm, Cu concentrations below ~13 vol.%, and 2–3 turns. Within this range, a representative structure with 2 turns, 250 nm major diameter, 100 nm minor diameter, and 800 nm height is selected and arranged in a hexagonal lattice with 300 nm center-to-center spacing. The resulting array introduces ~90% volumetric porosity (air volume fraction) due to air gaps between adjacent helices, lowering the effective refractive index. Accordingly, numerical simulations show that the plasmonic nanocomposite helix array maintains visible absorption comparable to that of the flat film while increasing the NIR transmittance from 75% to 87%, owing to reduced reflectance (Fig. 3e). Effective medium approximation predicts that porosities above 50% suppresses Fresnel reflection at the air-solid interface36, supporting our simulation results. We also analyze the wetting properties of the plasmonic nanocomposite helices using the pressure stability model34, which relates the apparent contact time to the intrinsic nano-micro structure diameter and the solid fraction (Fig. 3f). For non-wetting behavior based on pressure stability model, the structure must suppress droplet adhesion time, with high solid fraction, small size, and hydrophobicity. Since SiO₂ is intrinsically hydrophilic37, we chemically modify the surface with hydrophobic self-assembled monolayer to impart water repellence38. While fabrication details are presented in a later section, we note here that our experimental measurements are consistent with the pressure stable state analysis. On planar nanocomposite films, the hydrophobic coating produces a static contact angle of ~110°. In contrast, helically-structured plasmonic nanostructure arrays reach an apparent contact angle of 143° with the area fraction of 45% solid (Fig. 3f), validating that the hierarchical geometry sustains the pressure stable state. Fabrication of plasmonic nanocomposite helices We fabricate metal@oxide nanocomposite films on low-iron glass substrates by physical vapor co-deposition (Fig. 4a)39. During deposition, the Cu volume fraction is fixed at 10%, while the film thickness is varied from 50 to 400 nm in 50 nm increments, monitored by quartz crystal microbalance (QCM). Electron microscopies confirm uniform dispersion of Cu NPs (ca. 5.1 ± 0.3 nm) within the SiO2 matrix and precise thickness control (Fig. 4a and Supplementary Fig. 5). Optical measurements show that increasing film thickness enhances visible absorption (400–700 nm) but reduces NIR transmittance at 905 nm (Fig. 4b and Supplementary Fig. 6). This trade-off occurs because thicker films contain a larger fraction of absorbing Cu NPs and longer optical path length, which increase visible absorption but also induce scattering and absorption losses in the NIR, consistent with simulations. For instance, at 400 nm thickness of the Cu@SiO2 nanocomposite film, NIR transmittance decreases below 40%, whereas a 100 nm film maintains 78% (Fig. 4c). Based on these results, we select 100 nm thickness with 10% Cu NP doping as the optimized condition, balancing sufficient visible absorption for photothermal heating with the NIR transparency required for LiDAR. Optical and photothermal properties of plasmonic nanocomposite films and helices. a Schematic of physical vapor co-deposition (left). Representative electron microscopic images of the fabricated plasmonic nanocomposite film (middle) and associated digital images with various film thicknesses (50–400 nm, right). b Corresponding transmission spectra, c 905 nm transmittance and average visible absorptance vs. film thickness. d Schematic of glancing angle co-deposition method (left). Representative electron microscopic images of plasmonic nanocomposite helices (middle) and associated digital images with various Cu nanoparticle’s doping concentration (right). e Corresponding transmission spectra, f 905 nm transmittance and average visible absorptance vs. Cu nanoparticle’s doping concentration. g Photothermal responses of plasmonic nanocomposite films (left) and helices (middle) under AM 1.5 G (980 W/m²) illumination, and comparison at identical visible absorptance (right). h NIR transparency and photothermal performance comparison with various photothermal materials (20 references). To further improve optical performance, we extend this design into three-dimensional plasmonic nanohelices using glancing angle co-deposition40, which is more compatible to fabricate three-dimensional structure standing on substrate than synthesis. Since at a highly tilted angle of the vapor flux (nearly parallel to the target substrate), the shadow regions formed behind randomly deposited nanoislands prevent further material growth, wafer-scale arrays of helical structures can be grown with a continuous azimuthal substrate rotation. In addition to this, when co-evaporating Cu and SiO2, i.e., inherently non-mixing materials, Cu NP-embedded SiO2 nanohelices (Cu@SiO2) are fabricated (Fig. 4d). Arrays are fabricated with doping concentrations of 7, 10, and 13% Cu NPs. Analysis of electron microscopic images reveals nanohelices with ~100 nm minor diameter, 250 nm major diameter, 800 nm height, and 300 nm center-to-center spacing, corresponding to 90% porosity (Fig. 4d and Supplementary Fig. 7). While visible absorptions of these nanohelices are comparable to the optimized thin films, the NIR transmittance is significantly higher due to the antireflective effect of porous geometry (Fig. 4e)41. This physical effect maintains the NIR transmittance at 905 nm above 80% when visible absorption rises from 7.5 to 29.2% as the doping concentration of Cu NPs within the given helical shapes increases from 7 to 13% (Fig. 4f). The minor discrepancy between simulated and measured transmission arises from substrate differences (ideal semi-infinite SiO2 in simulation vs. finite-thickness glass in experiment), which introduce additional reflection and interference, while the extracted absorption spectra remain in good agreement (Supplementary Figs. 4 and 6). Importantly, this transmittance is preserved over incident angles of ±60°, demonstrating angle-robust optical performance sufficient for LiDAR operation (Supplementary Fig. 8). As a result, we identify the 800 nm-high plasmonic nanocomposite helices with 10% Cu NP doping as the optimal condition, balancing strong visible absorption with minimal LiDAR signal distortion. We then analyze the photothermal performance of the planar films and nanohelices using a solar simulator calibrated to 1 sun (980 W/m²), corresponding to direct sunlight under clear sky conditions (Supplementary Fig. 9a). For planar films, the 400 nm Cu@SiO2 film shows a maximum temperature rise of 16.4 °C within 12 min (Fig. 4g), owing to enhanced visible absorption, but its low NIR transmittance renders it unsuitable for LiDAR. By contrast, the optimized 100 nm film maintains 78.2% NIR transmittance and produces a moderate temperature rise of 4.9 °C within 14 min, sufficient for antifogging26. More crucially, the plasmonic nanocomposite helices display superior performance. With 10% Cu NP doping, Cu@SiO2 nanocomposite helices achieve a maximum surface temperature rise of 4 °C within 9 min (Fig. 4g), surpassing planar films under the same illumination and even showing repeatable temperature rise of 8.8 °C under 1000 W/m2 irradiance (Supplementary Fig. 9b–d). Although the temperature rises of both planar films and nanohelices proportionally increase as a function of visible absorptance (Supplementary Fig. 9), according to classical heat transfer theory42, as well as thin film approximation26, the arrays consistently produce higher surface heating than planar films at comparable visible absorption (Fig. 4g). This enhancement might be attributed to three possible geometrical effects: (i) the increased surface-air interface enhancing optical absorption efficiency43, (ii) the porous architecture suppressing reflection losses, resulting in more localized heat generation, and (iii) suppressing radiative losses in atmospheric windows by tailored nanostructure44. These results demonstrate that Cu@SiO₂ nanocomposite helices function as visible-selective photothermal surfaces that simultaneously maintain NIR transparency required for LiDAR. Compared with reported photothermal coatings based on polariton (metal oxide), plasmon (metal absorbers26,28,31,45–48,) exciton (conducting polymer, semiconductor, carbon wire12–14,49) and electron (MXene15–17,50) that typically compromise NIR transmission, our plasmonic nanocomposite helices achieve temperature rises above 9.2 °C, sufficient to trigger antifogging, while preserving NIR transmittance above 81.8% (Fig. 4h and Supplementary Table 1). This unique combination of selective visible absorption and high NIR transparency highlights the advantage of integrating plasmonic nanoparticles within bioinspired barbule-like nanostructures. Clearance of multiscale water droplets To validate the dual functionality of our plasmonic nanocomposite helices, we evaluate their ability to clear both microscopic condensation (fog) and macroscopic droplets, which together represent the primary optical challenges for LiDAR operation in humid and rainy conditions4,5. For antifogging tests, we construct a sealed chamber with low iron-glass substrates serving as a window (Fig. 5a and Supplementary Fig. 9a). The plasmonic nanocomposite helices face the University logo placed inside the chamber, which is maintained at temperature 24 °C and relative humidity (RH) 40%. When hot vapors are injected to increase the internal humidity, on bare glass, used as a control, condensed droplets persist for > 8 min, strongly scattering the light and obscuring the logo in both visible and NIR images (Fig. 5b and Supplementary Fig. 10). Even after 8 min of white light illumination, only 37% (visible) and 48% (NIR) of image visibilities are recovered (Fig. 5c). In contrast, the glass coated with plasmonic nanocomposite helices initiate condensation removal after 4 min of illumination, and by 6 min both visible and NIR images recover full clarity. These results confirm that the plasmonic nanocomposite helices enable passive antifogging under ambient illumination. Antifogging and water repellence performance of plasmonic nanocomposite helices. a Experimental setup for evaluating antifogging performance under 1 sun illumination. The sample window, maintained at 24 °C and 40% RH, is exposed to hot vapor inside a sealed chamber while visible and NIR cameras record real-time recovery. b Sequential visible and NIR images of bare glass (left) and plasmonic nanocomposite helices (right) under vapor condensation, showing rapid clearance on the plasmonic nanohelices. c Corresponding pixel-intensity recovery curves at visible and NIR wavelengths. d Water-repellence test comparing bare glass and PFOTS-treated plasmonic nanohelices tilted at 45° under rainfall (Gwangju, Republic of Korea, June 13, 2025, left). Time-series photographs show that water remains pinned on glass but rapidly slides off the plasmonic nanohelix surface (right). e Time-resolved images of droplet dynamics, illustrating pinning behavior on bare glass (top) and bouncing/rolling motion on plasmonic nanohelices (bottom), confirming pressure-stable hydrophobic wetting behavior. We also assess the water repellent performance for larger droplets. Because Cu@SiO₂ nanocomposite helices are intrinsically hydrophilic, we modify the surface with a hydrophobic molecular monolayer to further induce pressure-stable hydrophobicity (see “Methods”). Static contact angle measurement on the planar nanocomposite film shows ~110° of the water contact angle, whereas the plasmonic nanocomposite helices exhibit 143°, confirming the contribution of hierarchical geometry (Supplementary Fig. 11a). To evaluate dynamic repellence, outdoor tests with substrates tilted at 45° confirm that the plasmonic nanocomposite helices remove raindrops effectively under inclined geometries (Fig. 5d). Furthermore, individual droplet behaviors on surfaces are monitored by tilting 20° under 19.8°C temperature and RH 57.6% and dispensing 10 µL water droplets onto the surface (Supplementary Fig. 11b and Supplementary Movie 1)51,52. On bare glass and plasmonic nanocomposite films, droplets remain pinned and fail to detach (Fig. 5e and Supplementary Fig. 11c). In contrast, droplets on the plasmonic nanocomposite helices bounce and roll off within 30 msec, leaving no residual water (Fig. 5e). This result demonstrates that the hierarchical plasmonic nanocomposite helices simultaneously provide efficient photothermal antifogging and robust pressure-stable hydrophobic water repellence, enabling effective clearance of multiscale water droplets. Outdoor LiDAR tests under various weather conditions To further confirm real-world applicability, we evaluate LiDAR signal stability under various moisture-related environmental conditions (Supplementary Fig. 12). Note that these outdoor LiDAR tests are performed using the as-fabricated plasmonic nanohelices (3 turns), in order to assess the intrinsic functional performance of the structure. We first examine a vapor-induced fogging scenario under mild outdoor conditions (Fig. 6a and Supplementary Movie 2). A target object is placed 5 m away from the LiDAR sensor, with the sensor window covered by either bare glass or plasmonic nanohelices. We inject hot vapors toward the sensor window for 5 s to induce condensation under clear afternoon conditions (ambient temperature 8 °C, RH 40%) with natural irradiance of 280 W/m². Although this irradiance is lower than the standard 1 sun condition (980 W/m²), the plasmonic nanocomposite helices still enable rapid antifogging, removing moisture within 6 sec, potentially effective under cloudy conditions (Fig. 6b). LiDAR return profiles are reconstructed through image processing to compensate for background noise (Supplementary Fig. 13). On bare glass, condensation attenuates the returned signal, requiring 23 s to recover to 80% of its initial level (Fig. 6a). In contrast, the plasmonic nanocomposite helices maintain continuous signal transmission and achieve immediate recovery within 1 s, demonstrating that plasmonic heating suppresses condensation and prevents signal loss. We next evaluate performance under moderate rainfall conditions (3 mm/h), corresponding to a low temperature and high-humidity environment (3 °C, RH 94%, irradiance 50 W/m2) (Fig. 6c and Supplementary Movie 2). Under these conditions, bare glass accumulates raindrops, resulting in gradual signal degradation (~ 20% decay after 20 min, Fig. 6d). Meanwhile, the plasmonic nanocomposite helices effectively repel incoming droplets, keeping the LiDAR intensity constant throughout the whole measurement. Notably, despite the limited photothermal temperature rise (~ 2 °C) under low irradiance (50 W/m², Supplementary Fig. 9d), which is below conventional antifogging thresholds26, effective moisture mitigation is still achieved in the outdoor environment. This indicates that antifogging is not solely dictated by absolute temperature increase, but strongly influenced by coupled environmental factors such as humidity, airflow, and droplet dynamics, highlighting the importance of real-world performance validation. Outdoor LiDAR demonstrations of antifogging and water-repellent performance. a Schematic of the outdoor LiDAR antifogging test, where a LiDAR sensor monitors a target 5 m away while hot vapor is injected toward the sensor window under natural sunlight (top). Photographs of bare glass and plasmonic nanocomposite helices before and after fog injection (bottom). The pixel-recovery ratio of LiDAR signals for bare glass and plasmonic nanohelices during antifogging (right, inset: quantitatively scored environmental condition). b Corresponding time-resolved LiDAR images. c Schematic of the rainfall test performed under rain and reduced irradiance (top). Photographs of bare glass and plasmonic nanohelices during natural rainfall, indicating persistent water adhesion on glass vs. continuous self-clearing on the plasmonic surface (bottom). Relative LiDAR signal intensities for bare glass and plasmonic nanohelices under rainfall (right, inset: environmental condition). d Corresponding time-resolved LiDAR images. Outdoor LiDAR e, f antifogging test and g, h rainfall test result under harsh conditions (irradiance ~50 W/m2 for antifogging test and 5 mm/h precipitation for rainfall test). Beyond these mild and moderate outdoor scenarios, we further examine this cover under more challenging low-irradiance and high-precipitation conditions. Under low solar irradiance (~ 100 W/m², Fig. 6e and Supplementary Movie 3), the antifogging time required to recover 80% signal increases to 24 s compared to sunny conditions, yet the LiDAR system can still resolve the target within 10 s, significantly faster than bare glass (~ 30 s) (Fig. 6f). Similarly, under heavier rainfall (5 mm/h precipitation), bare glass exhibits rapid signal degradation (~ 20% decay after 5 min), whereas the plasmonic nanocomposite helices maintain stable signal transmission (Fig. 6g, h and Supplementary Movie 4). These outdoor field tests demonstrate that integrating plasmonic nanocomposite helices allows LiDAR systems to adapt to weather-induced moisture, previously hindered operation, by combining photothermal antifogging and pressure-stable hydrophobic droplet repellence. As a result, LiDAR-based computer vision and object recognition can remain reliable under diverse environmental conditions, ensuring consistent performance and enhanced safety in applications such as self-driving vehicles and autonomous robotics53,54. Durability and mechanical robustness While the outdoor LiDAR tests above are performed using the as-fabricated nanohelices to validate the intrinsic antifogging and water-repellent functionality, practical deployment in real-world environments additionally requires mechanical robustness against abrasive and impact conditions. We therefore further explore durability and structural reinforcement as an engineering extension of the nanohelical platform. To address the potential mechanical vulnerability associated with the high porosity and slender geometry of the nanohelical structures, we introduced a conformal Al2O3 passivation layer via atomic layer deposition (ALD) as a mechanically reinforcing strategy (Fig. 7a). The passivation material is selected based on its balanced mechanical robustness, intrinsic wettability, and chemical stability (Supplementary Fig. 14)55–58. A 20 nm-thick Al2O3 layer is deposited onto the plasmonic nanohelices, followed by PFOTS-based hydrophobic functionalization (Fig. 7b). Importantly, this passivation strategy preserves the original optical and wetting performance of the nanohelical structures. The passivated nanohelices maintain high transmittance at 905 nm while exhibiting an approximately 20% increase in visible absorbance (Fig. 7c), with negligible changes in contact angle and photothermal response (Fig. 7d). In addition, the PFOTS coating remains stable even after 446 days of ambient exposure (Supplementary Fig. 15a). Mechanical robustness and durability of plasmonic nanocomposite helices with ALD passivation. a Schematic illustration of conformal Al2O3 passivation of plasmonic nanocomposite helices via atomic layer deposition (ALD). b Digital photographs of the helices before (top) and after (bottom) passivation, with corresponding side- and top-view electron microscopy images. c Transmission spectra and d contact angle and photothermal response of 3-turn nanohelices before and after passivation. e High-pressure water exposure test (0.3 MPa), including schematic, digital images (top), and transmission spectra for 2-turn and 3-turn nanohelices (bottom). f Mechanical rubbing test under controlled pressure. g Particulate impact (sand abrasion) test following ASTM D968 conditions. h Wettability and photothermal response of nanohelices after the durability tests. To evaluate robustness under realistic operating conditions, we perform durability tests that mimic representative LiDAR-relevant stresses, including high-pressure water exposure (e.g., rainfall or car washing, Fig. 7e and Supplementary Movie 6), mechanical rubbing (e.g., wiping or cleaning, Fig. 7f and Supplementary Fig. 15b), and particulate impact (e.g., sand or dust, Fig. 7g and Supplementary Movie 7). Under a water pressure of 0.3 MPa, both 2- and 3-turn nanohelices maintain stable optical properties after 30 min exposure, with transmittance at 905 nm remaining above 85% and only minor changes in contact angle (Fig. 7h). Under mechanical and particulate stresses, however, a clear dependence on structural geometry is observed. The 3-turn nanohelices, which provide higher-aspect-ratio structures, exhibit partial detachment under repeated rubbing and show degradation in visible absorption after sand impact, indicating that the present 20 nm Al2O3 passivation is insufficient to fully stabilize the higher-aspect-ratio geometry. Although the 3-turn structures retain functional photothermal response after partial damage, performance degradation follows the loss of structural integrity. In contrast, the 2-turn nanohelices demonstrate significantly improved mechanical stability. Even after repeated rubbing and up to 400 g sand impact, the structures remain firmly attached to the substrate (Supplementary Fig. 15c), with minimal changes in optical, wetting, and photothermal performance. These results confirm that ALD passivation effectively enhances mechanical robustness, and that under the current reinforcement condition, the lower-aspect-ratio geometry provides superior resistance to mechanical and abrasive damage. These results reveal a clear functionality–durability trade-off governed by the helix geometry. While the 3-turn nanohelices offer superior initial hydrophobic and photothermal performance, the 2-turn nanohelices provide substantially greater mechanical robustness. Importantly, these findings do not alter the baseline LiDAR demonstrations obtained with the as-fabricated structures, but rather show that the nanohelical platform can be further engineered for mechanically robust deployment. Hence, these results highlight that the nanohelical platform is both intrinsically functional and engineerable for robust real-world deployment. We develop bioinspired plasmonic nanocomposite helices that combine visible-light-driven photothermal heating with pressure-stable hydrophobic water repellence to mitigate LiDAR signal degradation caused by multiscale droplets. Inspired by penguin feather barbules, our Cu@SiO2 helical arrays, fabricated by glancing angle co-deposition, introduce two innovations: the first photothermal antifogging strategy optimized for LiDAR, where visible absorption is harnessed while preserving NIR transparency, and the stabilization of Cu nanoparticles by embedding them in a silica matrix, enabling their use as durable and cost-effective plasmonic absorbers. The optimized plasmonic nanohelix arrays exhibit a water contact angle of 143°, maintain >80% NIR transmittance at 905 nm, and remove condensation within 6 s while keeping LiDAR signals stable under rainfall. In addition, the Al2O3 capped plasmonic nanohelix endures rubbing (9.4 MPa) and high-pressure water (0.3 MPa) without degrading its optical and mechanical properties. Looking ahead, several directions remain to translate this platform into broader deployment. Scalability is key consideration: the current growth system enables coatings on substrates up to 4 inches, but larger optics will require scale-up using bigger chambers59 or roll-to-roll deposition60. Performance can be further advanced by co-doping with complementary plasmonic or broadband absorbers to increase heating efficiency or achieve tailored spectral selectivity, potentially extending functionality toward longer infrared regimes (e.g., short-wave infrared)61,62 or more selective band-type responses depending on application requirement40. Because the fabrication method is adaptable to diverse substrates, integration onto curved lenses63, LiDAR housings64, or transparent displays39 is also feasible. In summary, our results establish plasmonic nanocomposite helices as a stable and scalable platform for weather-adaptive LiDAR, while also pointing toward future multifunctional optical coatings that combine antifogging, water repellence, and spectral filtering for applications in autonomous vehicles, robotics, outdoor imaging, and smart architectural windows. Methods Numerical simulation Optical properties of plasmonic nanocomposite films and helices with metal nanoparticles are calculated using a commercial finite-difference time-domain (FDTD) solver (Ansys Lumerical 24.2). For metal@oxide nanocomposite films, 5 nm metal nanoparticles are randomly distributed within the dielectric matrix with volume fractions of 2.5–25%, and the film thickness varies from 50 to 400 nm. For single helix, the geometry is defined with a length of 400–1000 nm, a pitch of two turns, an average helix diameter of 100 nm, an outer diameter of 250 nm, and an array spacing of 400 nm. A SiO₂ substrate is used as the base material, and the simulation domain employs a uniform mesh size of 1 nm with periodic boundary conditions in the x and y plane and a perfectly matched layer in the z direction. A broadband plane wave source ranging from 300 to 1500 nm illuminates the structures under normal incidence. The optical constants of metals (Au65, Ag65, Cu66, Mg67, Al66) and dielectrics (SiO₂66, HfO₂68, TiO₂69) are taken from literature. The simulations provide optical near-field enhancement, reflection, transmission, and absorption spectra. Growth of thin film and plasmonic nanohelices Cu@SiO₂ nanocomposite thin films are fabricated by the co-deposition method with two electron beam evaporators in a high vacuum chamber (2 × 10⁻⁷ torr). The deposition rates are fixed at 0.24 nm/s for SiO₂ and 0.05 nm/s for Cu, with the total thickness monitored by a quartz crystal microbalance (QCM) and controlled in the range of 50–400 nm. The substrates are mounted without tilting (0°), and no rotation is applied during thin film growth. For the fabrication of plasmonic nanocomposite helices, the glancing angle co-deposition method is applied in a high-vacuum chamber (2 × 10−⁷ torr). Nanocomposites are deposited on 5 cm × 5 cm (or 12 cm × 6 cm) glass substrates under a tilt angle of 85° with a substrate rotation speed of 2 rotations per hour. The substrate temperature is maintained at −36 °C during the deposition process. The SiO₂ deposition rate is fixed at 0.24 nm/s, while the Cu deposition rate is varied to 0.017 nm/s (7%), 0.026 nm/s (10%), and 0.037 nm/s (13%) to achieve different volume fractions, when fixing their overall thickness at 1500 nm. Hydrophobic molecule coating The Cu@SiO2 nanocomposite thin film and helices are treated with O2 plasma (pressure: 40 sccm, power: 100 W, CUTE Plasma, Femtoscience) for 10 min to remove organic impurities and activate hydroxyl groups on the surface. The samples are immersed in 50 mL of 26 mM perfluorooctyltrichlorosilane (PFOTS) dispersed in 99.9% ethanol. After 8 h at room temperature, samples are rinsed with ethanol several times and annealed on a hotplate at 50 °C for 20 min to fully dry out the solvent. Electron microscopic analysis The plasmonic nanocomposite thin film thicknesses and helical geometries are obtained using a scanning electron microscopy (SEM) system (Hitachi S4700) operated at 10 kV and transmission electron microscopy (TEM) performed using a Tecnai G2 F30 S-Twin instrument (FEI) operated at 300 kV. For thin film imaging, the samples are directly grown onto TEM grids. For nanohelix imaging, the samples grown on Si wafers are cut into ~1 cm² pieces and immersed in 1 mL of deionized water inside a glass vial. The suspension is treated in an ultrasonic bath for 1 min to release the helical structures into solution. Subsequently, 10 μL of the dispersed suspension is drop-cast onto TEM grids (EMS CF400-CU, UL) and air-dried. This drop-casting and drying process is repeated five times to achieve sufficient structural coverage for TEM observation. Optical spectroscopy Transmission spectra of the plasmonic nanocomposite thin films and helices are measured using a fiber-optic–coupled spectrometer (Ocean Optics QEpro). A halogen lamp (PHILIPS 7724, 4.33 mW) provides the broadband white light. The light illumination and collection are performed through a bifurcated optical fiber aligned normal to the sample surface. Reflection spectra are then collected through a reflection probe (QR400-7-UV-VIS, Ocean Optics) in which the illumination path from the halogen lamp and the collection path to the spectrometer are coupled within a single fiber assembly. All spectra are collected under identical illumination conditions to allow direct comparison between thin film and nanohelix samples. Photothermal analysis Photothermal measurements are carried out under simulated solar irradiation using a Xenon lamp (66984-300XF-R1, Newport) equipped with an AM 1.5 G filter (81094, Newport), providing a power density of 980 W/m² (1 sun). The sample temperature is monitored throughout all in situ heating experiments using K-type thermocouples (5SC-TT-K-40-36, OMEGA) connected to a commercial data logger (OM-CP-QUADTCTEMP-A2, OMEGA). Rubbing test The mechanical durability of the plasmonic nanocomposite helices is evaluated using a 1-axis abrasion tester (Daesung Precision, South Korea) equipped with a rubber contact probe. The coated glass samples were mounted on the sample stage, and repeated linear rubbing is applied under normal loads of 1.7 MPa and 9.4 MPa. For each applied load, 5 rubbing cycles are conducted. After the rubbing test, the optical properties and wettability are measured, and the coating integrity is examined. High-pressure water test The coated samples are exposed to continuous high-pressure water spray (~ 0.3 MPa) for 1, 5, 10, and 30 min. After each test, the optical properties, wettability, and coating integrity are evaluated. Sand abrasion test Sand abrasion tests are performed using silica sand with a particle size distribution of 0.3–0.7 mm in a gravity-driven impact configuration adapted from ASTM D968. The coated samples are mounted at an angle of 45°, and the sand is dropped from a height of 1 m through a tube with an inner diameter of 18 mm onto the sample surface. Two sand loading conditions, 20 g and 400 g, are applied. After the test, the directly impacted region and the peripheral region exposed to scattered particles are examined separately. Surface wettability, optical properties, and photothermal performance are measured before and after sand impact. Antifogging test Antifogging performance is evaluated in a homemade acrylic chamber under controlled humidity and temperature conditions (temperature: 24 °C, relative humidity 40%). Photothermal-based antifogging dynamics are monitored using a visible camera (Galaxy S21, Samsung) and an NIR (850 nm) vision system (Digital Night Vision Binoculars, R18, LIZHOUMIL, China), enabling simultaneous evaluation of image recovery. LiDAR outdoor test Outdoor LiDAR measurements are performed using a Velarray H800 (Ouster) LiDAR sensor. The condition of the LiDAR cover glass during operation is recorded with a smartphone camera (Galaxy S21, Samsung), while external objects (e.g., bicycle) are observed in the visible range using a digital camera (ILCE-6400, SONY). Solar irradiance during the outdoor test is also measured with a pyranometer (SR05-A1, Hukseflux) and recorded to account for variations in natural illumination. The antifogging test is performed on March 29, 2025, between 5:00 and 6:00 PM, and the rainfall test is carried out on June 13, 2025, during the same time interval. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Movie 1 Supplementary Movie 2 Supplementary Movie 3 Supplementary Movie 4 Supplementary Movie 5 Supplementary Movie 6 Supplementary Movie 7 Peer Review file Source data Source Data Acknowledgements We are grateful to the NanoSystems Laboratory at GIST for technical support. We also thank the GIST Advanced Institute of Instrumental Analysis (GAIA) for technical assistance with the SEM and TEM imaging and the GIST Nanoinfra for Compound Semiconductors (G-NICS) for access to the cleanroom. We also gratefully acknowledge the National Institute of Ecology (Chungcheongnam-do, Republic of Korea) for providing Pygoscelis papua feather samples. Author contributions J.L., G.K., and H.-H.J. designed experiments. J.L. performed the simulation. J.L. and G.K. performed the sample fabrication, with support from D.K., H.K. and J.-H.H., and all measurements, data analysis, and LiDAR experiments. J.L., G.K., H.-H.J. wrote the paper and all authors contributed editing. Peer review Peer review information Nature Communications thanks Haiping Zhang and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available. Funding This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (RS-2024-00352033), the InnoCORE program of the Ministry of Science and ICT (GIST InnoCORE KH0830), Global Value-Up 10X Project grant funded by the GIST in 2026, and Hyundai Motor Company. Data availability All relevant data supporting the findings of this study are available within the paper, Supplementary Information or Source Data file. Additional data are available from the corresponding author upon request. Source data are provided with this paper. Code availability The code used to reconstruct LiDAR profiles can be made available from the corresponding author upon request. Competing interests The authors filed one patent application related to the principle and fabrication method. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: JuHyeong Lee, Gyurin Kim. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-75037-1. References Associated Data Supplementary Materials Supplementary Information Description of Additional Supplementary Files Supplementary Movie 1 Supplementary Movie 2 Supplementary Movie 3 Supplementary Movie 4 Supplementary Movie 5 Supplementary Movie 6 Supplementary Movie 7 Peer Review file Source Data Data Availability Statement All relevant data supporting the findings of this study are available within the paper, Supplementary Information or Source Data file. Additional data are available from the corresponding author upon request. Source data are provided with this paper. The code used to reconstruct LiDAR profiles can be made available from the corresponding author upon request.