Open Access
Add to BibSonomy
Abstract
The camouflage for developed hyperspectral detection technology, which can accurately distinguish the spectrum between object and background, has emerged as an important unsolved challenge. In this study, a biomimetic film (Ge/ZnS multilayer structure) for optical camouflage of hyperspectral and laser with color simulation has been proposed and experimentally demonstrated. By taking advantage of the wavelength selective property of Ge/ZnS multilayer through film interference, the biomimetic film which can simulate the reflection spectral characteristics of vegetation background and eliminate laser signal has been realized based on inverse design. The selective narrowband absorption can manipulate the contrary condition for hyperspectral camouflage (high reflectance in 0.8-1.3 µm) and laser camouflage (low reflectance at 1.06 µm) in the same waveband. The planarized biomimetic multilayer film presents several distinct advantages: (1) elaborate simulation of vegetation reflectance spectrum for hyperspectral camouflage (the spectral similarity coefficient of 92.1%), and efficient absorption at 1.06 µm for laser camouflage (reflectance of 17.8%); (2) tunable color chrominance of various vegetation types for visual camouflage; (3) thermally robust camouflage performance (up to 250 °C) due to temperature endurable property of Ge and ZnS. The hyperspectral-laser camouflage film expands the design strategy of optical camouflage application.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The rapid development of advanced optical detection technologies, such as hyperspectral and laser, has enhanced the ability to identify and track ground military targets [1,2]. The hyperspectral detection technology can accurately measure the spectral characteristics of each pixel area in the reconnaissance image. It analyzes the reflectance spectrum in 0.4-2.5 µm waveband (visible light to near-infrared) of a target and its surrounding background [3]. As a result, the traditional colored camouflage targets, although which has realized the same color as background through visible spectral regulation (such as pigments, photonic crystals, metasurfaces, and so on), exhibit significantly different reflectance spectrum with backgrounds. The green vegetations are typical backgrounds for visible camouflage [4–8]. Considering the ground targets under green vegetation background, the hyperspectral camouflage material should simulate the reflectance spectrum of green vegetation to counter hyperspectral detection. The hyperspectral camouflage material mainly regulates the following spectral characteristics: Firstly, the green reflection peak at 550 nm, with a reflection intensity of 15%-25%, is used to simulate the color of green plants; Secondly, the near-infrared reflection plateau at 800-1300 nm has an internal reflection intensity of 50%-60%, which is used to simulate the reflection of plant surface structure to light; Thirdly, the low reflection valleys at 1450 nm and 1950nm are caused by water absorption in plant leaves [9,10].
Many works based on chemical approaches have been implemented to simulate reflectance spectrum of green vegetation from a bionics perspective. Xie et al. developed a microcapsule colorant with a chloroplast-like structure and chlorophyll-like absorption, and designed a generic bilayer coating to provide high spectral similarity to leaves with different growth stages, seasons and species. The microcapsule colorant is fabricated by monomeric zinc phthalocyanine (ZnPc), which provides highly selective absorption like the in vivo chlorophyll Q-band for simulation of green peak. The monomeric ZnPc is compartmentally encapsulated in micro-scale polymer shell to simulate green peak [11]. Qin et al. designed a poly(urea-formaldehyde) microcapsule containing chlorophyll and water, which could simulate the solar spectrum reflectance of natural leaf [12]. The chlorophyll tends to be photodegraded in vitro, it has an adverse effect on the endurance of the bionic material under sunlight, limiting its practical application. Gao et al. proposed a bionic membrane containing hygroscopic material and Cr2O3 pigment to simulate the solar spectrum reflectance of natural leaves [3]. The absorption of Cr2O3 pigment and that of water contribute to “green peak” and water absorption band in the reflection spectrum of the bionic membrane. The strong scattering of Cr2O3 results in high reflectance to simulate the “red edge” and “near-infrared plateau” in 800-1300 nm of natural leaf. However, the materials of above works easily lose moisture under thermal effects, leading to the disappearance of water absorption characteristics. Lu et al. selected Cr2O3 as a color simulation material and Palygorskite as a water absorption simulation material, combined with polyurethane (PU) to form a hyperspectral camouflage coating [13]. It demonstrates a good spectral similarity coefficient of 0.9124. There is a certain disparity in water absorption characteristics between the coating and the green leaves. Although these materials achieved a similar reflectance spectrum to green vegetation, the inferior stability, dehydration, and easy decomposition limit their practical application. At the same time, due to different species and seasons of vegetation backgrounds, the hyperspectral camouflage materials are required to achieve different colors to match the vegetation background.
In the optical frequency band, 1.06 µm is a typical wavelength of lidar for detecting objects [14]. Laser camouflage can be realized by absorption or scattering at corresponding laser wavelength [15–18]. These methods are mostly implemented through broadband low reflectance. However, the low reflectance of 1.06 µm laser is contradictory to the characteristics of “near-infrared plateau” with broadband high reflectance in 0.8-1.3 µm waveband, which is required for hyperspectral camouflage. The reason is that the broadband absorption or scattering for laser camouflage will destroy the broadband reflection of the “near-infrared plateau” [16]. Therefore, it is difficult to achieve the two requirements at the same time, and hyperspectral-laser compatible camouflage lacks an effective approach. For optical multispectral camouflage against other detection approach, like visible-infrared camouflage, spectral selectivity can be achieved by multilayer films or metasurfaces [2,5]. At the same time, many selective reflectors or emitters have been designed for camouflage and radiative cooling utilizing inverse design based on optimization algorithm recently [19–21]. It has the potential for hyperspectral-laser camouflage by combining artificial photonic structure and inverse design.
In this work, a planarized biomimetic film (multilayer structure) for hyperspectral-laser compatible camouflage has been proposed based on inverse design. The multilayer stacking structure consists of three materials (Ge, ZnS, and YbF3) with a total thickness of 1.708 µm. For inverse design process, the genetic algorithm is used for accurately simulating the reflectance spectrum of green vegetation and realizing low reflectance for 1.06 µm laser, which demands sophisticated spectral features over a wide wavelength region (0.4-2.5 µm). The visual color can also be regulated by altering the interference layer thickness of ZnS and YbF3. The reflectance spectrum is measured experimentally, which shows good agreement with the calculated result. The hyperspectral-laser camouflage film maintains high performance after thermal annealing (250 °C), indicating well heat resistance due to its inorganic constitution. Moreover, the proposed biomimetic film is fabricated by planarized deposition of electron beam evaporation technology, which has great potential for large-scale preparation. This work which simulates the spectrum of natural vegetation provides a novel approach for the application in optical camouflage technology.
2. Experiment
2.1 Fabrication process
The multilayer films are fabricated by electron beam evaporation (EBE) technology (Rankuum, ZZS1100-8/G) on a 2-inch SiO2 substrate. The multilayer films are deposited at a base pressure of 5 × 10−4 Pa and the baking temperature of the substrate is set as 150 °C during the fabrication process. The quartz crystal oscillator is used to monitor the thickness and deposition rate of the films. The deposition rates of Ge, ZnS, and YbF3 are 0.4 nm/s, 0.8 nm/s, and 1.2 nm/s, respectively.
2.2 Characterization of multilayer films
For microstructure characterization, the cross-section of the fabricated Ge/ZnS multilayer films is characterized by a field emission scanning electron microscope (FESEM, ZEISS Sigma300). For optical measurement, the refractive index data of Ge and ZnS are derived from spectroscopic ellipsometer (VASE Mark II) across a wavelength range of 0.3-2.5 µm. The reflectance spectrum in the visible and near-infrared wavebands of green leaves and multilayer films is measured by a UV-Vis-NIR spectrophotometer (Shimadzu UV3600-Plus). The spectral matching performance of the fabricated multilayer films with green vegetation is evaluated by the spectral similarity model. It indicates the spectral similarity between the object and reference in the 400-2500 nm waveband. The spectral similarity coefficient is calculated as Eq. (1) [13]:
3. Results and discussion
3.1 Principle of hyperspectral-laser camouflage
The conceptual scene of the planarized biomimetic film is depicted by shrouding a car with two types of films in Fig. 1(a). Both the conventional green film and biomimetic film can blend into the vegetation background by displaying green color to realize camouflage against visible imaging. However, the conventional green film is unambiguously detected by hyperspectral imaging technology, which measures the reflectance of each wavelength in a 400-2500 nm waveband. On the contrary, the biomimetic film can mimic the reflectance spectrum of vegetation background and achieve camouflage against hyperspectral imaging.
Fig. 1. (a) Concept of the hyperspectral-laser compatible camouflage film. (b) The reflectance spectrum of green vegetation. (c) Schematic diagram of the multilayer film structure.
For hyperspectral imaging, mimicking the reflectance spectrum of green plants can fuse the objects with the vegetation background. Figure 1(b) shows the reflectance spectrum of green vegetation leaves in 400-2500 nm waveband (green solid line). The “green peak”, “near-infrared plateau”, “red edge”, and “water absorption valleys” are the main features of the vegetation reflection spectrum in 400-2500 nm. The “green peak” shows a reflection peak around 550 nm, which is generated by the absorption characteristics of chlorophyll in the visible waveband. The “near-infrared plateau” shows a reflectance of around 55%−60% in the 800-1300 nm wavelength, which is caused by the scattering process in the multilayer cell structure. The collimated incident radiation transforms to approximately uniform diffuse radiation through the scattering process in the leaves, thus forming a highly reflective region. The “red edge” is located approximately in the 690-790 nm waveband, which is attributed to chlorophyll absorption and scattering in the multilayer cells. The “water absorption valleys” are located around 1450 and 1940nm, which are generated by the absorption characteristics of water in the leaves [22].
Furthermore, considering the development of NIR lidar, the 1.06 µm laser is a typical wavelength, which is often used for laser detection. Therefore, the proposed biomimetic film also realizes laser camouflage by absorbing the light (low echo signal) at 1.06 µm.
3.2 Modeling of Ge/ZnS multilayer films
We design a planarized multilayer biomimetic film due to its easy fabrication and the need for large-scale application. The structure of the green biomimetic films is illustrated in Fig. 1(c). The film comprises alternate deposition of thin films of Ge and ZnS, and YbF3 film is the top layer of the structure. We select Ge and ZnS as the materials of multilayer films for their good mechanical compatibility and large refractive index contrast. The Ge is a lossy material with a high refractive index in the visible range, enabling large refractive index contrast for interference and strong resonance with reasonable structure design [23], which is a benefit for simulating the “green peak” and “red edge”. The ZnS is served as a low refractive index to provide a thin-film constructive interference for high reflection of “near-infrared plateau”. The measured refractive index n and extinction coefficient k values of Ge and ZnS are plotted in Fig. 2. Meanwhile, the optical constants of YbF3 are extracted from ref. 24. Both ZnS and YbF3 are transparent in visible to near-infrared wavebands.
Fig. 2. Material dispersion of two dielectric materials, (a) Ge and (b) ZnS in the visible-NIR region.
The biomimetic films discussed here are composed of aperiodic multilayers consisting of Ge and ZnS alternating layers grown on SiO2 substrates. The initial structural parameter is ZnS (220 nm)/Ge (50 nm)/ZnS (240 nm)/ Ge (50 nm)/ZnS (160 nm)/Ge (50 nm)/SiO2 (substrate), which is set manually and empirically to make its spectrum as close as possible to the vegetation spectrum. The calculated reflectance spectrum of the initial structural parameter is shown in Fig. 3(a). The reflectance intensity is higher than that of the vegetation target spectrum. To determine the dimensions of the structure with the best spectral property, the optimal design proceeds.
The layer number and individual layer thicknesses are all considered as design parameters. The genetic algorithm (GA) is employed to minimize the difference between the reflectance spectrum of the initially designed structure and the target spectrum [25,26]. The GA we chose is a commonly used optimization algorithm. It can achieve a relatively high optimization efficiency and great results for such multilayer films via combing with transfer matrix method (TMM). Other optimization algorithms like the needle algorithm, which is also commonly used method, optimizes the spectrum by continuously introducing ultra-thin layers with abrupt refractive index. The method leads to the increase of layer number. The fabrication process is too complex due to the ultra-thin thickness of single layer and large layer number. Figure 3 shows the detailed workflow of the optimal design process with the GA. Figure 3(b) depicts the target function of optimization, which is basically consistent with the green vegetation reflectance spectrum shown in Fig. 1(b). The YbF3 film is set as the top layer for antireflection, as shown in Fig. 3(c). At the beginning of the optimization process, a total of 60 cases with materials of Ge, ZnS, and YbF3 randomly arranged in a 6-layer film are generated as the first generation. The reflectance spectrum of the corresponding structure is calculated by TMM and compared to the target spectrum, outputting an error factor [27]. The error factor is the residual error between the designed spectrum and the target spectrum, minimizing the error factor is the ultimate goal of the optimal process. Here we set the function of error factor as Eq. (2).
Table 1. Material and thickness for each layer of the proposed multilayer film.
For the visible range, the reflectance spectrum can be regulated by altering the thickness of the top ZnS and YbF3 layers, which can blend the object into various vegetation backgrounds. By varying the ZnS thickness from 183 nm to 203 nm, the reflection peaks are formed at different wavelengths. Figure 4(a) shows the simulated reflectance spectrum for different ZnS thicknesses and the corresponding colors with CIE-1931 chromaticity coordinates (x, y) [28]. Different chromaticity determined by ZnS thickness can make the biomimetic film suitable for different different vegetation species and seasons. Meanwhile, by varying the YbF3 thickness from 52 nm to 72 nm, the reflection peak intensity changes significantly with the same wavelength (Fig. 4(b)), which is associated with the luminance change. We can regulate the chromaticity and luminance of the biomimetic film simultaneously by changing the thickness of ZnS and YbF3. The near-infrared spectra of the green biomimetic film are influenced slightly by the top YbF3 layer by its negligible thickness compared to wavelength. Furthermore, the chromaticity coordinates of numerous surface colors with various ZnS thicknesses (from 173 to 213 nm, at an interval of 5 nm) and YbF3 thicknesses (from 42 to 82 nm, at an interval of 5 nm) are also calculated (shown in Fig. 4(c)). The continuous track of chromaticity coordinates further confirms the flexible color configuration and strong expansibility of the biomimetic film, which is beneficial for background adaptability. We can adjust the color in advance by changing the thickness of the film, according to the requirements of different vegetation backgrounds and seasons. The customization of thin films applied to different backgrounds can been achieved.
Fig. 4. (a) Simulated reflectance spectrum in the visible-NIR region for different ZnS thicknesses. Inset is the calculated colors corresponding to the thicknesses of ZnS. (b) Simulated reflectance spectrum in the visible-NIR region for different YbF3 thicknesses. Inset is the calculated colors corresponding to the thicknesses of YbF3. (c) Simulated chromaticity coordinates with various ZnS (white crosses) and YbF3 (black crosses) thicknesses.
In many instances, the angular independent performance is important for the objects under directional detection. The simulated reflectance spectra of s-polarized and p-polarized light at different incident angles are shown in Fig. 5. The reflectance spectrum of s-polarized light increases with incident angle increasing, and that of p-polarized light decreases. This is consistent with the theoretical variation trend of the film reflectance at different incident angles. Then, following the next formula: Runpolarized = (Rs-polarized + Rp-polarized)/2, the reflection intensity of unpolarized light remains unchanged, which is still close to the standard reflection spectrum of green vegetation (Fig. 5(c)). Therefore, it maintains hyperspectral camouflage ability within incident angles of 60° under detection of non-polarized light. The reflectance of 1.06 µm laser at the incident angle of 0°, 15°, 30°, 45°, and 60° is 12.4%, 12.9%, 16.49%, 31.51%, and 43.88%, respectively. It maintains great laser camouflage performance (R < 20%) at incident angle below 10°. Fortunately, laser camouflage just requires low reflectance at normal incidence as the oblique incidence of the laser would be reflected to other directions which are beyond the detection range of lidar [5].
Fig. 5. (a, b) Reflectance spectra for s-polarized and p-polarized light at oblique incidence angles of 15°−60°, respectively. (c) Reflectance spectra for unpolarized light at oblique incidence angles up to 60°.
3.3 Mechanism of Ge/ZnS multilayer films
The reflectance, transmittance and absorptance of the structure in the 400-2500 nm waveband are calculated, and the results are shown in Fig. 6. The spectral data shows that there is no transmission in the 300-800 nm waveband due to the strong absorption of Ge. The strong selective absorption leads to low reflectance. The green reflection peak appears by adjusting the thickness of the ZnS layer; The decrease of absorptance at 700-800 nm leads to the increase of reflectance, which simulates the increase of reflectance (“red edge”) of green plants due to their cell structure. At 800-1300 nm, due to the decline of extinction coefficient of Ge, the absorptance continue to decline, and the transmittance begin to rise. Therefore, the reflectance is maintained about 50%, which can simulate the near-infrared reflection plateau of green plants. Strong absorption and partial transmission occur at 1.06 µm, so the reflection at this wavelength is weak. Because Ge has none absorption after 1300 nm, the absorption gradually decreases to zero, the transmittance at 1300-1450 nm continues to rise, and the reflectance also decreases again, forming a reflection valley (water absorption). In the 1500-2500 nm waveband, the reflectance sharply declines to form reflection valleys, thus simulating the reflectance change caused by the water absorption of green plants.
Fig. 6. Simulated reflectance, transmittance, and absorptance spectra of the structure at normal incidence.
To understand the physical origin of the reflectance spectrum of the designed structure, we simulated the electric field (|E|) and resistive heat loss (Q) distribution of the structure at typical wavelengths by finite element method (FEM). The FEM conducts frequency domain calculation, which has higher computational accuracy for a single frequency. Figure 7 depicts the electric field and resistive loss distribution at wavelengths of 550, 700, 900, 1060, 1300, and 1450 nm, which correspond to each feature of the green peak, red edge, near-infrared plateau, water absorption band, and laser, respectively. The green region refers to ZnS layer, and that with blue and red colors refers to Ge and YbF3 layers. Note that the wave is incident from the right boundary, and the substrate is adjacent to the left boundary. For the “green peak” at 550 nm, the energy of incident light is attenuated by Ge layers, resulting in no light transmission through the structure. For the characteristic absorption band of water at 1450 nm, the electric field does not decay within the Ge/ZnS multilayers, leading to high transmittance (low reflectance) through the film. For a laser wavelength of 1060 nm, the reflectance is realized by the coordination of absorption and transmission, and the incident wave is mainly decayed within the Ge layers.
Fig. 7. The electric field |E| and resistive loss Q in the structure for wavelengths of 550 nm, 700 nm, 900 nm, 1060 nm, 1300 nm, and 1450 nm.
3.4 Measured reflectance spectrum of Ge/ZnS multilayer films
As shown in Fig. 8(a), the biomimetic film is observed visually as green, which is in accordance with the calculated color, allowing the covered objects to visually blend into the vegetation environment. The cross-section of the biomimetic film is shown in the inset of Fig. 8(b), with a scale bar of 500 nm. It can be seen that the Ge and ZnS layers are stacked alternately and the boundaries are smooth and clear, indicating the successful fabrication of the designed biomimetic multilayer film.
Fig. 8. (a) Image of the sample with ZnS thickness of 193 nm and YbF3 thickness of 62 nm. (b) Reflectance spectrum of the film measured by a spectrophotometer.
The reflectance spectra of the green biomimetic film is measured by a UV-Vis-NIR spectrophotometer. As shown in Fig. 8(b), it indicates good agreement with the calculation. It realizes the four main characteristics of vegetation spectrum, containing “green peak”, “near-infrared plateau”, “red edge”, and “water absorption valley”. The measured green peak is located around 550 nm, dominating to form a green color. The reflectance in 800-1300 nm waveband is around 50-60%, forming the feature of “near-infrared plateau”. The low reflection characteristics around 1450 and 1950nm are consistent with water absorption. The “red edge” around 760-860 nm waveband is formed by the strong spectral selectivity. The reflectance of the normal incident angle at 1.06 µm is measured as 17.8%, which is mainly in accordance with the calculated value (12.4%). The slight discrepancy between the calculated and measured results might have originated from an unexpected thickness error during the fabrication process, the refractive index deviation, and normal error in measurement. The performance can be further improved through optimization of structure and fabrication parameters. The spectral similarity coefficient between the measured sample spectrum (except 1000-1100 nm waveband) and the measured reflection spectrum of green vegetation is 92.1%, showing excellent hyperspectral camouflage performance.
The samples were annealed on a heating stage in the air atmosphere to verify their thermal stability. We used a thermocouple to measure the surface temperature of the samples. The annealing process lasted 30 minutes after reaching the setting temperature. The whole process of each annealing process took about 45 minutes. The hyperspectral camouflage is mainly applied to land equipment, which temperature is generally as high as around 100-200 °C [29]. Therefore, we conducted the temperature test up to 250 °C according to our application scene, which covers the serving temperature range of land equipment. The reflectance spectra of the green biomimetic film are characterized after annealing in air atmosphere at different temperatures. As shown in Fig. 9, the reflectance spectrum of the green biomimetic film varies slightly after annealing at different temperatures and is still maintains well hyperspectral and laser camouflage performance. In particular, the 1.06 µm laser reflectance of the green biomimetic films after annealing at 100 °C, 200 °C, and 250 °C for 30 minutes are 17.8%, 18.1%, and 18.6%, respectively. The reduced laser camouflage performance may be attributed to the slight change in thickness of the biomimetic film after annealing. It is remarkable that the hyperspectral camouflage performance of the green biomimetic film can be maintained well with a heating temperature of up to 250 °C. This demonstrates the thermal stability of green biomimetic films in harsh conditions, such as the vehicle hood and other heat-affected parts.
Fig. 9. The measured reflectance spectrum of the green biomimetic film of annealing temperatures from ambient temperature to 250 °C.
4. Conclusion
In summary, a planarized biomimetic film of Ge/ZnS multilayer structure with broadband spectral selectivity is proposed and demonstrated for hyperspectral-laser camouflage applications based on inverse design. Using genetic algorithm-assisted design, the structure of Ge/ZnS multilayer can be efficiently optimized for vegetation spectrum in 0.4-2.5 µm along with low reflectance at 1.06 µm. The green biomimetic films fabricated by EBE technology exhibit excellent hyperspectral-laser compatible camouflage performance, showing high comparability with vegetation (>90%) in 0.4-2.5 µm and low reflectance of below 20% at 1.06 µm. The thermal stability of reflectance spectrum is experimentally measured, representing that the proposed biomimetic film can maintain excellent hyperspectral camouflage performance with serving temperature from room temperature to 250 °C. Furthermore, portable or wearable applications can be enabled as the films are fabricated on the surface of a thermostable flexible substrate such as polyimide films [30,31]. Phase change materials (e.g., VO2, Ge2Sb2Te5, Sb2S3) or liquid crystal also can be utilized to realize adaptive spectrum control [32–38], which can dynamically change colors and merge with various vegetation backgrounds. This work provides a novel approach for simultaneous control of hyperspectral and laser camouflage through planarized multilayer design, and has broad applications in efficient and delicate spectral regulation.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. K. L. Roth, A. Casas, M. Huesca, S. L. Ustin, M. M. Alsina, S. A. Mathews, and M. L. Whiting, “Leaf spectral clusters as potential optical leaf functional types within California ecosystems,” Remote Sens. Environ. 184, 229–246 (2016). [CrossRef]
2. M. Pan, Y. Huang, Q. Li, H. Luo, H. Zhu, S. Kaur, and M. Qiu, “Multi-band middle-infrared-compatible camouflage with thermal management via simple photonic structures,” Nano Energy 69, 104449 (2020). [CrossRef]
3. Y. Gao and H. Ye, “Bionic membrane simulating solar spectrum reflection characteristics of natural leaf,” Int. J. Heat Mass Transfer 114, 115–124 (2017). [CrossRef]
4. X. Chai, D. Zhu, Y. Liu, Y. Qing, Z. Ren, F. Luo, W. Zhou, Z. Huang, and P. Li, “Silver-modified chromium (III) oxide as multi-band compatible stealth materials for visual/ infrared stealth and radar wave transmission,” Compos. Sci. Technol. 216, 109038 (2021). [CrossRef]
5. H. Zhu, Q. Li, C. Tao, Y. Hong, Z. Xu, W. Shen, S. Kaur, P. Ghosh, and M. Qiu, “Multispectral camouflage for infrared, visible, lasers and microwave with radiative cooling,” Nat. Commun. 12(1), 1805 (2021). [CrossRef]
6. D. Qi, F. Chen, X. Wang, H. Luo, Y. Cheng, X. Niu, and R. Gong, “Effective strategy for visible-infrared compatible camouflage: surface graphical one-dimensional photonic crystal,” Opt. Lett. 43(21), 5323–5326 (2018). [CrossRef]
7. Y. Huang, Y. Zhu, B. Qin, Y. Zhou, R. Qin, P. Ghosh, M. Qiu, and Q. Li, “Hierarchical visible-infrared-microwave scattering surfaces for multispectral camouflage,” Nanophotonics 11(16), 3613–3622 (2022). [CrossRef]
8. D. Ding, X. He, S. Liang, W. Wei, and S. Ding, “Porous Nanostructured Composite Film for Visible-to-Infrared Camouflage with Thermal Management,” ACS Appl. Mater. Interfaces 14(21), 24690–24696 (2022). [CrossRef]
9. K. Xu and H. Ye, “Preparation and optimization of biomimetic materials simulating solar spectrum reflection characteristics of natural leaves,” J. Mater. Sci. 55(27), 12848–12863 (2020). [CrossRef]
10. D. Pauli, S. C. Chapman, R. Bart, C. N. Topp, C. J. Lawrence-Dill, J. Poland, and M. A. Gore, “The quest for understanding phenotypic variation via integrated approaches in the field environment,” Plant Physiol. 592, pp.00592.2016 (2016). [CrossRef]
11. D. Xie, M. Zu, M. Li, D. Liu, Z. Wang, Q. Li, and H. Cheng, “A Hyperspectral Camouflage Colorant Inspired by Natural Leaves,” Adv. Mater. 1, 1 (2023). [CrossRef]
12. R. Qin, G. Xu, L. Guo, Y. Jiang, and R. Ding, “Preparation and characterization of a novel poly(urea–formaldehyde) microcapsules with similar reflectance spectrum to leaves in the UV-Vis-NIR region of 300-2500 nm,” Mater. Chem. Phys. 136(2-3), 737–743 (2012). [CrossRef]
13. H. Lu, X. Bai, Z. Wang, Y. Guo, L. Zhang, X. Weng, J. Xie, D. Liang, and L. Deng, “Hyperspectral camouflage coating using Palygorskite to simulate water absorption of healthy green leaves,” Mater. Sci. Semicond. Process. 156, 107293 (2023). [CrossRef]
14. L.A. Montmorillon, P. Delaye, J.C. Launay, and G. Roosen, “Novel theoretical aspects on photorefractive ultrasonic detection and implementation of a sensor with an optimum sensitivity,” J. Appl. Phys. 82(12), 5913–5922 (1997). [CrossRef]
15. X. Jiang, H. Yuan, X. He, T. Du, H. Ma, X. Li, M. Luo, Z. Zhang, H. Chen, Y. Yu, G. Zhu, P. Yan, J. Wu, Z. Zhang, and J. Yang, “Implementing of infrared camouflage with thermal management based on inverse design and hierarchical metamaterial,” Nanophotonics 12(10), 1891–1902 (2023). [CrossRef]
16. Z. Deng, Y. Su, W. Qin, T. Wang, X. Wang, and R. Gong, “Nanostructured Ge/ZnS Films for Multispectral Camouflage with Low Visibility and Low Thermal Emission,” ACS Appl. Nano Mater. 5(4), 5119–5127 (2022). [CrossRef]
17. X. Feng, M. Pu, F. Zhang, R. Pan, S. Wang, J. Gong, R. Zhang, Y. Guo, X. Li, X. Ma, and X. Luo, “Large-Area Low-Cost Multiscale-Hierarchical Metasurfaces for Multispectral Compatible Camouflage of Dual-Band Lasers, Infrared and Microwave,” Adv. Funct. Mater. 1, 2205547 (2022). [CrossRef]
18. X. Liu, Z. Yan, H. Wang, Q. Zhao, H. Zhou, and T. Fan, “Bioinspired Meta-Reflection-Splitter for Near-Infrared Laser Stealth with Large Scattering Angle,” Laser Photonics Rev. 17(2), 2200635 (2023). [CrossRef]
19. W. Xi, Y. Lee, S. Yu, Z. Chen, J. Shiomi, S. Kim, and R. Hu, “Ultrahigh-efficient material informatics inverse design of thermal metamaterials for visible-infrared-compatible camouflage,” Nat. Commun. 14(1), 4694 (2023). [CrossRef]
20. J. Cho, Y. Lee, J. Kim, R. Hu, E. Lee, and S. Kim, “Directional Radiative Cooling via Exceptional Epsilon-Based Microcavities,” ACS Nano 17(11), 10442–10451 (2023). [CrossRef]
21. Z. Chen, S. Yu, B. Hu, and R. Hu, “Multi-band and wide-angle nonreciprocal thermal radiation,” Int. J. Heat Mass Transfer 209, 124149 (2023). [CrossRef]
22. D.A. Sims and J.A. Gamon, “Relationships between leaf pigment content and spectral reflectance across a wide range of species, leaf structures and developmental stages,” Remote Sens. Environ. 81(2-3), 337–354 (2002). [CrossRef]
23. M. A. Kats and F. Capasso, “Optical absorbers based on strong interference in ultra-thin films,” Laser Photonics Rev. 10(5), 735–749 (2016). [CrossRef]
24. J.D. Traylor Kruschwitz and W.T. Pawlewicz, “Optical and durability properties of infrared transmitting thin films,” Appl. Opt. 36(10), 2157 (1997). [CrossRef]
25. X. Jiang, H. Yuan, D. Chen, Z. Zhang, T. Du, H. Ma, and J. Yang, “Metasurface Based on Inverse Design for Maximizing Solar Spectral Absorption,” Adv. Opt. Mater. 9(19), 2100575 (2021). [CrossRef]
26. J. Huang, H. Ma, D. Chen, H. Yuan, J. Zhang, Z. Li, J. Han, J. Wu, and J. Yang, “Digital nanophotonics: the highway to the integration of subwavelength-scale photonics,” Nanophotonics 10(3), 1011–1030 (2021). [CrossRef]
27. L. Wang, Y. Yang, X. Tang, B. Li, Y. Hu, Y. Zhu, and H. Yang, “Combined multi-band infrared camouflage and thermal management via a simple multilayer structure design,” Opt. Lett. 46(20), 5224–5227 (2021). [CrossRef]
28. D. Qi, Y. Cheng, X. Wang, F. Wang, B. Li, and R. Gong, “Multi-layer composite structure covered polytetrafluoroethylene for visible-infrared-radar spectral Compatibility,” J. Phys. D: Appl. Phys. 50(50), 505108 (2017). [CrossRef]
29. M. Khaled, C. Habchi, F. Harambat, A. Elmarakbi, and H. Peerhossaini, “Leakage effects in car underhood aerothermal management: temperature and heat flux analysis,” Heat Mass Transfer 50(10), 1455–1464 (2014). [CrossRef]
30. J. N. Wang, B. Xiong, R. W. Peng, C. Y. Li, B. Q. Hou, C. W. Chen, Y. Liu, and M. Wang, “Flexible Phase Change Materials for Electrically-Tuned Active Absorbers,” Small 17(31), 2101282 (2021). [CrossRef]
31. N. Lee, J. Lim, I. Chang, D. Lee, and H. H. Cho, “Flexible Thermocamouflage Materials in Supersonic Flowfields with Selective Energy Dissipation,” ACS Appl. Mater. Interfaces 13(36), 43524–43532 (2021). [CrossRef]
32. D. Guo, C. Ling, C. Wang, D. Wang, J. Li, Z. Zhao, Z. Wang, Y. Zhao, J. Zhang, and H. Jin, “Hydrothermal One-Step Synthesis of Highly Dispersed M-Phase VO2 Nanocrystals and Application to Flexible Thermochromic Film,” ACS Appl. Mater. Interfaces 10(34), 28627–28634 (2018). [CrossRef]
33. S. Yoo, T. Gwon, T. Eom, S. Kim, and C. S. Hwang, “Multicolor Changeable Optical Coating by Adopting Multiple Layers of Ultrathin Phase Change Material Film,” ACS Photonics 3(7), 1265–1270 (2016). [CrossRef]
34. J. He, M. Zhang, S. Shu, Y. Yan, and M. Wang, “VO2 based dynamic tunable absorber and its application in switchable control and real-time color display in the visible region,” Opt. Express 28(25), 37590 (2020). [CrossRef]
35. E. Buhara, A. Ghobadi, and E. Ozbay, “Adaptive visible and short-wave infrared camouflage using a dynamically tunable metasurface,” Opt. Lett. 46(19), 4777–4780 (2021). [CrossRef]
36. Y. Liu, H. Zuo, W. Xi, R. Hu, and X. Luo, “Flexible Janus Functional Film for Adaptive Thermal Camouflage,” Adv. Mater. Technol. 7, 1 (2022). [CrossRef]
37. K. V. Sreekanth, R. Medwal, Y. K. Srivastava, M. Manjappa, R. S. Rawat, and R. Singh, “Dynamic Color Generation with Electrically Tunable Thin Film Optical Coatings,” Nano Lett. 21(23), 10070–10075 (2021). [CrossRef]
38. Y. Liu, J. Song, W. Zhao, X. Ren, Q. Cheng, X. Luo, N. X. Fang, and R. Hu, “Dynamic thermal camouflage via a liquid-crystal-based radiative metasurface,” Nanophotonics 9(4), 855–863 (2020). [CrossRef]
