Metafilms for visible and infrared compatible camouflage of high-temperature targets

With the rapid development of multispectral detection, infrared and visible compatible camouflage becomes necessary. Metafilms with dielectric/metal/dielectric (D/M/D) structures can be highly transparent in visible band (380 ∼ 780 nm) and highly reflective in infrared atmospheric windows (3 ∼ 5 μm, 8 ∼ 14 μm). The metafilm can be deposited on the equipment surface, and the high visible transmittance can make the original camouflage coating continue to achieve visible camouflage, while the low infrared emissivity can inhibit the infrared signal to achieve infrared camouflage. Compatible camouflage is urgently needed by high-temperature targets such as exhaust pipes and engine cabins. Therefore, the thermal stability of multilayer structure is very important. In this study, a D/M/D-structured metafilm with improved thermal performance is proposed. Al-doped zinc oxide (AZO) is selected as the material of the dielectric layers due to good thermal stability, and high visible transmittance is realized through the mechanism of admittance matching. Ag is selected as the material of the metal layer to increase infrared reflectance. The metafilm with the structure of AZO/Ag/AZO is rigorously designed and fabricated. The results from Fourier transform infrared spectrometer and spectrophotometer show that the integrated visible transmittance and infrared emissivity at room temperature is higher than 0.87 and lower than 0.05, respectively. The camouflage performance of the metafilm is demonstrated on a flexible polyethylene terephthalate (PET) substrate. The camouflage performance of metafilm samples at 20 ∼ 140 °C is tested on a model cabin. The metafilm does not affect the original camouflage coating, so it can achieve visible camouflage. The radiation temperature of the metafilm is approximately 80 °C lower than that of the control surface, and the infrared signature is significantly attenuated. In order to further investigate the thermal stability and thermal fatigue resistance of the metafilm, metafilm deposited on quartz substrate is continuously heated and periodically heated at different temperatures. It is found that the sample can withstand continuous heating at 450 °C for 4 h or repeated heating for 20 cycles. SEM (scanning electron microscope) and EDS (energy dispersive spectrometer) scanning shows that if heated at higher temperature or for more cycles, the AZO layer becomes blocky, and the proportion of Ag and O changes significantly. This leads to the decrease of visible transmittance and the increase of infrared emissivity of samples.


Introduction
With the rapid development of photoelectric technology, advanced infrared and visible detection systems have been widely used in the military field [1]. The infrared detector recognizes and tracks the infrared radiation in infrared atmospheric windows (3 ∼ 5 μm and 8 ∼ 14 μm) spontaneously emitted by the target and converts it into measurable signals, while the visible detection works by identifying the visual contrast characteristics of the Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. target reflected in visible band (380 ∼ 780 nm). Visible camouflage reduces the visual difference between the target and background [2]. Common visible camouflage means, such as pattern painting, make the visible signature of targets consistent with surrounding environment through combination of a variety of color blocks. Infrared camouflage suppresses the infrared radiation difference between the target and the background. For high-temperature targets, there are two ways to conceal the infrared characteristics, i.e., controlling temperature and decreasing surface emissivity. Infrared camouflage material mainly considers decreasing the emissivity in infrared detection band [3,4]. According to the fact that absorbance is equal to emissivity, to opaque objects, low emissivity corresponds to high reflectance. Camouflage technologies individually for infrared or visible band cannot meet the needs of modern battlefield, and infrared-visible compatible camouflage technology has received increasing attention [5]. Infrared-visible compatible camouflage requires visible reflectance that is close to the environment and low infrared emissivity. At present, there are mainly two methods to achieve compatible camouflage, i.e., composite painting containing particles and functional coatings with artificial structures. Welldesigned coatings, also known as photonic crystals or metafilms, can display various colors and extremely low emissivity [6,7], making them promising materials for compatible camouflage. For example, the 5-period Ge/ ZnS one-dimensional photonic crystal designed by Zhang et al [8] and the Ag/Ge based multilayer structure designed and prepared by Peng et al [9], can achieve spectral selection in the infrared band by changing the structure.
Traditional metafilms for compatible camouflage is designed to maintain low infrared emissivity while displaying different colors. For example, Qi et al [10]designed and prepared a composite heterostructure based on multilayers of SiO 2 /Ag/ZnS/Ag with average reflectance higher than 95% in infrared band. The sample could be yellow, navy and cyan by changing the structure. They [11]also designed a ZnS/Ge-based metafilm. By changing the thicknesses of the Ge and ZnS layers, the sample can be blue, brown and of other colors with emissivity as low as 5.4% in 3 ∼ 5 μm band. However, in order to achieve the visible camouflage, different structures are needed for different colors, which is difficult to form large-area pattern painting. We previously proposed a new approach to achieve infrared-visible compatible camouflage, which is to combine the pattern painting with the dielectric/metal/dielectric (D/M/D) metafilm with high visible transmittance and high infrared reflectance. The metafilm can be deposited on the equipment surface, and the high visible transmittance can make the original camouflage coating continue to achieve visible camouflage, while the low infrared emissivity can inhibit the infrared signal to achieve infrared camouflage. With this idea, we designed and prepared metafilms based on ZnS/Ag/ZnS and ZnO/Ag/ZnO, which can achieve transmittance in visible band and reflectance in infrared atmospheric windows both greater than 0.8 [12,13].
The parts requiring infrared camouflage are often located in high-temperature areas, such as exhaust pipes or engine cabins. For example, the temperature of a gun barrel in a tank can reach 700 K in the case of multiple rapid fire [14]. The thermal stability of materials for compatible camouflage needs to be studied. However, the compatible camouflage performance under high temperature is lack of research. On the other hand, the ZnSand ZnO-based metafilms proposed by our research group are insulating materials and cannot be deposited with a direct current (DC) magnetron sputtering system. The coating rate of radio frequency (RF) magnetron sputtering is relatively slow and not suitable for mass production. It is necessary to reselect materials with better conductivity to facilitate deposition.
Al-doped zinc oxide (AZO) is widely used in optical films because of its rich reserves, innocuity, good conductivity, good thermal stability and high visible transmittance [15,16]. Metafilms constructed by combining AZO with Ag or Sn-O x have high infrared reflectance [17][18][19]. Van Eek et al [20]applied the AZO/ Ag/AZO structure to the field of solar cell film manufacturing and studied the effect of Ag thickness to the transmittance of the structure in visible and near-infrared bands. Ren et al [21] designed a transparent electrode based on ITO/Ag/AZO structure and achieved transmittance of 0.98 at the wavelength of 486 nm. It can be found that by adjusting the thickness of each layer, the metafilm based on AZO and Ag can achieve high transmittance in visible band (380 ∼ 780 nm) and high reflectance in infrared atmospheric windows (3 ∼ 5 μm, 8 ∼ 14 μm). Therefore, applying this structure to visible-infrared compatible camouflage is worth studying.
For compatible camouflage of high-temperature targets, this work based on the dielectric/metal/dielectric (D/M/D) structure selects AZO as the dielectrics and Ag as the metal to design and prepare metafilm samples. The spectral signature is simulated by transfer matrix method and compared with the test results. The visibleinfrared compatible camouflage performance is demonstrated by a model cabin, and the thermal stability of the metafilm is discussed.

Model and preparation method
Transfer matrix method (TMM) is widely used to design D/M/D metafilms because it can calculate the reflection and transmission spectra of multilayers efficiently and precisely. The basic idea of TMM is to divide the whole structure into multiple layers. In each layer, the physical properties are isotropic, and the relationship between adjacent layers can be represented by a transfer matrix. Using the transfer matrix, the electromagnetic field distribution of the whole structure can be extrapolated from the electromagnetic field of a single layer, and then reflection coefficient and transmission coefficient of the film structure can be calculated.
According to whether there is electric field or magnetic field component in the propagation direction, electromagnetic wave can generally be divided into TE wave and TM wave. In the propagation direction of TE wave, there is only magnetic field component and no electric field component, while in the propagation direction of TM wave, there is only electric field component and no magnetic field component. For a k-layer metafilm as shown in figure 1(a), TM wave is considered first, where n i and d i are the refractive index and physical thickness of layer i, respectively. The subscripts of 0 and k + 1 refer to air and substrate. E ii and E ri are the electric field intensity of incident light and reflected light of layer i, respectively. The overall transmission equation of the metafilm is given as follows: where ε i and μ i are the relative permittivity and magnetic permeability of layer i, α i is the refraction angle of the incident ray in the layer i, λ 0 is the wavelength of light in vacuum, E is the electric field intensity, and H is the magnetic field intensity. Taking metafilm and substrate as an equivalent layer, its admittance Y is There is only forward travelling wave but no backward one in the substrate so we get (4) and (5) into equation (1), we have

Substituting equations
cos sin sin cos According to the calculation formula of amplitude reflection coefficient of single interface, the reflection coefficient r and reflectance R of the film are as follows: The transmittance T and absorbance A of the film are as follows: where h 0 is the effective admittance of air, and h + k 1 is the effective admittance of substrate. h + k 1 can be calculated by material properties from equation (3). Superscript * stands for conjugate. In the above process, only the transmittance and reflectance formula of the metafilm itself has been obtained. In practical application, the metafilm is always deposited on the substrate, so the overall transmittance and reflectance of the metafilm and substrate need to be calculated, as shown in figure 1(b). Considering a substrate with thickness d s and the complex refractive index + n ik , s s the overall transmittance and reflectance of the metafilm and substrate can be obtained according to ray tracing method: where R a and T a are the reflectance and transmittance of the metafilm to air, R b and T b are the reflectance and transmittance of the film system to substrate, R s and T s are the reflectance and transmittance of the substrate to air and T c is the transmittance of light in the substrate. The formula is as follows: Based on induced transmission theory [21], the transmittance of the film is affected not only by its own optical properties and film thickness but also by adjacent media. In order to maximize the transmittance of the metafilm, it is necessary to make the effective admittance of the metafilm equal to the admittance of the outgoing medium. The D/M/D metafilm can achieve the effect of visible anti-reflection by adjusting the admittance matching of the middle metal layer and dielectric layers on both sides and achieve high infrared reflectance through the metal layer. The metal layer as the intermediate layer is the main functional layer which is required to have the optical characteristics of high infrared reflectance and high visible transmittance. The absorbance of Ag in visible band is low, the visible absorbance of a 13-nm-thick Ag layer is less than 5%, and the reflectance in infrared band is more than 90% [22]. Ag is an ideal metal layer material. The role of the dielectric layer in the metafilm is to adjust the admittance matching and protect the intermediate metal layer from oxidation. In order to achieve compatible camouflage, the dielectric layer also needs to have low absorbance in both visible and infrared bands. AZO has been selected as the material of internal and external dielectric layers because of its rich reserves, innocuity, good conductivity, good thermal stability and high visible transmittance. The optical constants of Ag and AZO are taken from literature [23,24].
AZO/Ag/AZO metafilms were deposited by magnetron sputtering on a 0.5-mm-thick double polished quartz substrate and polyethylene terephthalate (PET) flexible substrate respectively. The substrates needed to be cleaned before sputtering. First, the substrate was cleaned by ultrasonic wave with ethanol for 10 min and with deionized water for 10 min, and it was dried with high-pressure nitrogen. In this work, Kurt J. Lesker LAB18 magnetron sputtering coater was used to fabricate the metafilm. The coating power was set as 150 W, and the air pressure was 3 mTorr. Three-inch Ag target (99.99% pure) and AZO target (ZnO: Al 2 O 3 = 98:2 wt%, 99.99% pure) were used. In order to determine the deposition rate, the substrate was evaporated with Ag target and AZO target respectively for 15 min, and then the film thickness was measured with a step meter. The calibrated deposition rates of Ag and AZO were 6.11 Å s −1 and 0.49 Å s −1 , respectively.

Optimization of radiation properties
To achieve infrared-visible compatible camouflage, the D/M/D metafilms, which covers the pattern painting, should be highly transparent in visible band and lowly emissive in infrared region. According to the fact that emissivity is equal to absorbance, for opaque objects, low emissivity corresponds to high reflectance. To further optimize the metafilm structure, a metafilm performance optimization function Z is defined here: where l T ( ) is the spectral transmittance of the metafilm structure, l D is the relative spectral power distribution of D65 standard light source, and l V ( ) is the visual coefficient of the human eye [25]. Including the emissivity of 5 ∼ 8 μm band to the calculation, ε is integrated infrared emissivity at room temperature in 3 ∼ 14 μm band, the formula is as follows:  = t K 293 . It is assumed that the spectral transmittance and reflectance do not change with temperature.
In order to find the optimal structure, the steepest descent method is used for optimization. The basic idea is to find the extreme value along the descending or ascending direction of the function gradient. Through optimization calculation, the structural parameters are shown in table 1, and the calculated transmittance and reflectance are shown in figure 1(c). The highest calculated transmittance of the film is at 0.55 μm and higher than 0.91. The visible integrated transmittance of film is 0.9. The reflectance in infrared band is higher than 0.9, and the infrared integrated reflectance is 0.92. The samples on different substrates are prepared by magnetron sputtering coating, as shown in figure 1(d). The thickness of each layer of samples is confirmed by SEM, as shown in figure 1(e). It can be found that the thickness of each layer is 45.6 nm, 14.4 nm and 48.2 nm respectively, close to the designed values.
A UV-visible spectrophotometer (SHIMADZU Solid-3700) and a Fourier transform infrared (FTIR) spectrometer (Bruker VERTEX 80) were used to measure the normal incidence hemispherical transmittance and the normal incidence hemispherical infrared reflectance of metafilms on a quartz substrate and a PET substrate, as well as bare quartz substrate and PET substrate. The results are shown in figure 1(c). The measured integral transmittance of quartz substrate metafilm is 0.87, and that of PET substrate film system is 0.80. The highest transmittance of quartz base film is 0.88 at λ = 0.55 μm and the highest transmittance of PET base film is 0.81 at λ = 0.54 μm. The integral transmittance of uncoated PET sheet is 0.88 and that of uncoated quartz sheet is 0.90. It can be found that the highly transparent metafilm exerts little effect on the transmittance of the substrate itself in visible band. The integral reflectance of the metafilm of quartz substrate and PET substrate in infrared band is more than 0.95, which is much higher than the uncoated quartz sheet and PET sheet, indicating that the film structure has high infrared reflectance.
The distributions of electric and magnetic fields of quartz substrate at λ = 0.55 μm and λ = 5 μm simulated by the finite difference time domain (FDTD) method are shown in figure 2. This method directly solves the Maxwell rotation equation in time domain by converting it into a finite difference equation. By establishing a progressive sequence with discrete time, the electric field and magnetic field can be calculated alternately in the interlaced grid space. It can be found from figure 2(a) (b) that when the electric field intensity is maximum, the magnetic field intensity is minimum. The separation of the electric and magnetic fields at λ = 0.55 μm implies the occurrence of standing wave pattern. The reflected waves will superpose with one another to form a destructive interference leading to this standing wave pattern. This effect is also known as the Fabry-Perot (F-P) resonance effect. This leads to the high transmittance of the metafilm in visible band [26]. The distribution of electric field and magnetic field at λ = 5 μm is mainly concentrated on the upper AZO layer and Ag layer, indicating that almost all electromagnetic waves emitted by the light source are reflected by the upper structure, which leads to the high reflection characteristics of the metafilm in infrared band. The transmittance and reflectance of the metafilm at different incident angles are also calculated, and the results are shown in figure 1(f). It can be found that when the incident angle of TE wave and TM wave is less than 47°, the visible integrated transmittance of the multilayer film is greater than 0.8. For any incident angle of TE wave or when the incident angle of TM wave is less than 55°, the infrared integrated reflectance is greater than 0.9. It shows that the sample can realize compatible stealth in visible and infrared bands in a wide range of angles.

Compatible camouflage demonstration
In order to demonstrate how the metafilm works for compatible camouflage, a testing apparatus was built. A model cabin with side length of 4.5 cm was coated with thermal-stable pattern painting. Each surface of the cabin was heated by strip heaters. The photo of the model cabin is shown in figure 3(a). During the demonstration, one surface of the cabin was coated with aluminum sheet, and one with flexible PET substrate metafilm sample. A third surface stayed uncoated to serve as a control. The infrared reflectance of each surface is provided in figure 3(b). Both the aluminum sheet and the metafilm are highly reflective in infrared band, but the aluminum sheet covers the underneath painting. The cabin surface with pattern painting displays low infrared reflectance. In the experiment, the sample temperature was controlled by adjusting the power of heating strips. The actual temperatures of the surfaces were measured by thermocouples, and the thermal images and radiative temperatures were taken by two infrared thermal imagers with wavelength ranges of 3 ∼ 5 μm and 8 ∼ 14 μm. During the experiment, the ambient temperature was 20°C. The surface temperature of the model cabin was heated from room temperature to 140°C. The built-in emissivity of thermal imagers was set to 1, and thermal images were taken every 20°C. The change of thermal radiation temperature after obtaining thermal balance is shown in figure 3(b) and the infrared thermograms when thermal balance was reached are shown in figure 3(c). At each actual temperature, the radiative temperature of the control surface is the highest, and those of the metafilm and aluminum surfaces are close. Specifically, at 140°C, the radiation temperature of sample in 3 ∼ 5 μm band is 43°C, which is similar to the radiation temperature of aluminum sheet and 82°C lower than the radiation temperature of control surface. While the radiation temperature of sample in 8 ∼ 14 μm band is 34.6°C , which is close to the radiation temperature of aluminum sheet and 85.8°C lower than the radiation temperature of control surface. This is because the infrared emissivity of sample is far lower than the emissivity of the control surface, which is close to the emissivity of aluminum sheet. The radiation temperature difference of the metafilm sample is less than 5°C in 3 ∼ 5 μm and 8 ∼ 14 μm band, which is because the integrated emissivity of the sample is basically the same in different wave bands. In general, the radiation temperature of metafilm is lower than the radiation temperature of actual cabin surface and similar to that of aluminum foil, which is in line with expectations. And its radiation characteristics are closer to the surrounding environment, which shows the good infrared camouflage characteristics of the sample. Because of the high visible transmittance, the sample has infrared-visible compatible camouflage effect.

Characterization of thermal stability
Compatible camouflage is urgently needed by high-temperature targets such as exhaust pipes and engine cabins. Therefore, the thermal stability of multilayer structure is very important. Due to the temperature limit of softening point of PET, quartz is used as the substrate for further testing. In order to test the thermal stability of AZO/Ag/AZO metafilm, the samples were heated continuously in air at 350°C, 450°C and 550°C respectively with ambient temperature 24°C, relative humidity around 40% ∼ 50%, and the heating rate 1°C/s. The visible transmittance and infrared emissivity are shown in figures 4(a)-(c). After heating at 350°C for 8 h, the transmittance of the sample is basically unchanged with the highest transmittance at 0.9. After heating for 12 h, the highest transmittance is 0.88. After heating for 24 h, the highest transmittance is 0.85, which is reduced by 0.03. The emissivity of the metafilm sample does not changed basically after heating for 36 h. As the heating time increases at 450°C, the maximum transmittance of the sample decreases gradually. After heating for 8 h, the maximum transmittance decreases from 0.9 to 0.85. After heating for 24 h, the emissivity of the sample remains basically unchanged and increased by 0.02 after heating for 36 h. After continuous heating at 550°C for 4 h, the highest transmittance of the sample decreases from 0.9 to 0.7 and the transmittance does not change significantly after cooling, indicating that the sample structure has changed and cannot be recovered. After heating for 12 h, the emissivity of the sample increases significantly. After heating for 36 h, the emissivity has reached 0.5, indicating that the properties of the metafilm is functionally destroyed. Jeong et al [27]studied the effect of rapid thermal annealing (RTA) process on the optical properties of ITO/Ag/ITO structures. They found that the transmittance of metafilms annealed at 600°C decreased significantly. Through high resolution transmission electron microscope (HRTEM), they found that the originally continuous Ag layer became isolated islands due to the aggregation of Ag atoms. Also, they proposed that oxygen might diffuse into the Ag layer through the upper ITO layer, causing damage to the metafilm. It was found by XPS (x-ray photoelectron spectroscopy) that after annealing at 600°C, Ag diffused significantly into ITO at upper and lower layers. All these factors lead to the decrease of the film transmittance. The surface morphology of samples before heating and continuously heated at 550°C for 36 h were observed by SEM (scanning electron microscope), and their element distribution was analyzed by EDS (energy dispersive spectrometer). The results are shown in figure 5. Before heating the surface is continuous and element distribution is uniform. After continuous heating at 550°C for 36 h, the surface of sample is no longer continuous. Through EDS, it can be observed that the distribution of Zn is uneven, indicating that the original structure is damaged. It might be thermal stress that causes the damage [28,29].
The above results show that the sample can withstand continuous heating at 450°C for a long time. Here, the thermal fatigue resistance at 450°C was further tested. In the periodical test, the sample was repeatedly heated from room temperature to 450°C and stayed for 30 seconds before being cooled to room temperature. The visible transmittance and infrared reflectance after several cycles are shown in figure 4(d). It can be found that after 20 cycles of repeated heating at 450°C, the highest transmittance decreases from 0.89 to 0.82. The infrared reflectance does not change after 20 cycles and decreases by 0.05 after 50 cycles, indicating that the film has gradually degenerated. The surface element distribution of the sample after heating at 450°C for 50 cycles is analyzed by EDS. The proportion of elements before and after heating is shown in table 2. The proportion of Ag

Conclusion
In this work, an AZO/Ag/AZO metafilm with improved thermal performance was proposed for infrared-visible compatible camouflage. The metafilm could be deposited on the equipment surface. The high visible transmittance keeps the original visible camouflage coating work normally, while the low infrared emissivity could inhibit the infrared signal to achieve infrared camouflage. The thickness of the film layer (46 nm/12 nm/ 53 nm) was simulated by the transfer matrix method and optimized by the steepest descent method. The sample was fabricated on the quartz substrate and PET substrate using magnetron sputtering. The results from Fourier transform infrared spectrometer and spectrophotometer showed that the visible transmittance and infrared reflectance of quartz substrate metafilm were higher than 0.87 and 0.95 at room temperature, which meet the requirements of infrared-visible compatible camouflage. The demonstration experiment showed that the radiation temperature of the metafilm was approximately 80°C lower than that of the control surface, and the infrared signature was significantly attenuated. The visible transmittance and infrared emissivity of the samples remained basically unchanged when heated continuously at 450°C for 4 h or repeatedly for 20 cycles at most. SEM and EDS scanning showed that the originally smooth AZO layer becomes bumpy after heating, and the proportion of Ag and O changed significantly. The high infrared reflectance, high visible transmittance and excellent thermal stability make this metafilm promising in compatible camouflage applications.