1. Introduction

Camouflage is an adaptive strategy involving various physical and behavioral techniques to conceal an object in its environment in order to enhance the survival chances or success of the object [1–4]. To confront the threats posed by detectors operating both actively and passively in different spectral ranges, a pressing need exists for multi-band compatible camouflage [5–7]. Modern multi-band camouflage technologies involve modulation of infrared (IR) radiation in the mid-infrared (MIR) band [8,9], color control in the visible band [10,11], and suppression of reflective signals for lasers and microwaves [12–14]. Notably, due to the prevalence of IR detection, IR camouflage research has emerged as a focus area [15,16]. According to the Stefan-Boltzmann law, the way to modulate the IR signal of targets is either controlling the temperature or tuning the target emissivity [8,17]. Conventional metal-based camouflage materials (Al, Au, Ag, etc.) have broadband low IR emissivity, but undesirably block the radiative cooling of high temperature targets [18,19]. Consequently, contemporary MIR camouflage materials require low emission in the detection bands and high emission in the non-detection band to cope with radiative heat dissipation [20,21]. Along with passive detection in the MIR band, active detection lidars, predominantly the Nd:YAG lidar operating at 1.06 µm and CO₂ lidar at 10.6 µm, should also be considered into the design of camouflage materials [22,23]. Further compatibility with the visible band necessitates that camouflage materials embody various structural colors like green, blue and yellow in order to match typical backgrounds [24,25]. Therefore, the four requirements for the design of camouflage materials are: (i) low averaged emissivity (smaller than 0.3 [7,13,25]) in two MIR atmospheric window bands where the atmosphere is nearly transparent, i.e. the 3∼5 µm mid-wavelength-infrared (MWIR) and the 8∼14 µm long-wavelength-infrared (LWIR) bands [26–29]. The MWIR and LWIR bands are the operating bands for IR detection, so targets with low averaged emissivity in these bands can avoid being detected; (ii) relatively high averaged emissivity in the non-atmospheric window band (5∼8 µm) [30–32]. While satisfying the premise of IR stealth in the atmospheric windows, a relatively high averaged emissivity in the non-atmospheric window (5∼8 µm) can radiate more thermal energy into the atmosphere to create a cooling effect; (iii) high absorptivity at the operating wavelengths of lidar (most frequently being 1.06 µm and 10.6 µm) [6,12,33,34]; (iv) tunable structural colors in the visible over a wide color gamut [35–38].

However, to date, no reported surface structures simultaneously satisfy all the four camouflage requirements mentioned above. Metasurfaces are often employed in the design of multi-band compatible camouflage. Pan et al. [33], Kang et al. [34], and Yu et al. [22] designed camouflage metasurfaces for the MIR dual-windows and the 10.6 µm CO₂ laser, concurrently facilitating radiative cooling. Wu et al. [39] and Lee et al. [40] designed metasurfaces achieving thermal camouflage and radiative thermal management in MIR, while maintaining optical transparency in the visible band. The metasurfaces designed for camouflage typically have complicated structures. In general impression, the structure complexity increases with incorporation of more operating camouflage bands. In contrast, multilayer films receive great concern in multi-band camouflage due to their simple geometries. Qi et al. [35,36] and Zhang et al. [37,38] implemented MIR camouflage and radiative thermal management based on one-dimensional photonic crystals, which achieved structural color control by varying the thickness of the top layer. Jiang et al. [13] designed a 7-layer metamaterial, achieving IR camouflage, radiative thermal management and stealth for 1.06 µm, 1.55 µm and 10.6 µm lasers. The research on camouflage materials closest to fulfilling the four requirements is conducted by Zhu et al. [7], who designed a Ge/ZnS 12-layer film structure on a microwave metasurface that is compatible for visible, infrared, lasers (1.55 µm and 10.6 µm) and microwave camouflage with radiative cooling. However, the 1.55 µm laser is much less applied in practice than the 1.06 µm laser. Additionally, ZnS might be unstable under humid circumstances.

Therefore, camouflage materials reaching a balance between multi-band compatibility and structural simplicity is still in lack. In this work, we report a simple 9-layer film structure on quartz substrate which achieves compatible camouflage obeying the above four requirements [see Fig. 1(a)]. Combining rigorous coupled-wave analysis (RCWA) [41] and genetic algorithm (GA) [42], we optimize multilayered films with no predefined number of layers. The materials of each layer are chosen from SiO₂, Ge and TiO₂, which are chemically stable and high-temperature resistant. The designed structure achieves low emissivity in the MWIR and LWIR bands, high emissivity in the non-atmospheric window band, high absorptivity at the wavelengths of 1.06 µm and 10.6 µm, and tunable structural colors in the visible band. The underlying physics are analyzed and the overall camouflage performance (OCP) is evaluated.

 

Fig. 1. (a) The application scenarios of camouflage for visible, infrared and lasers with radiative cooling and the schematic diagram of multilayer film structure (MFS) composed of SiO₂, Ge and TiO₂ films; (b) Schematic diagram of the workflow of GA; (c) Schematic diagram of the optimized MFS; (d) The ideal emission spectra (red line), the emission/absorption spectra of the optimized MFS (blue line) and the atmospheric transmission spectra calculated by MODTRAN (light blue area).

Download Full Size | PDF

2. Model and methods

The multilayer film structure (MFS) is illustrated in Fig. 1(a). The MFS is composed of SiO₂, Ge, and TiO₂ layers stacked alternately on a quartz substrate. The main reasons for selecting SiO₂, Ge, and TiO₂ as layer materials of the MFS are as follows [43,44]: (i) they are stable at high temperatures, so the performance will not deteriorate significantly with increasing temperature of the object; (ii) they have distinct refractive indices and different IR transparent zones (i.e. 0.9∼8 µm for SiO₂, 1.7∼23 µm for Ge and in 0.4∼10 µm for TiO₂), which will lay the foundation for the MIR selective absorption/transmission of the structure [43,44]. In addition, quartz is chosen as the substrate material due to its high absorptivity/emissivity in the MWIR band [7,44]. The MFS above the substrate will be designed to selectively transmit the thermal emission of the substrate in the band of 5∼8 µm. By fixing the thickness of the quartz substrate at 2000µm, it is possible to absorb almost completely the transmitted electromagnetic waves in the 5∼8 µm band, which is equivalent to a high emission characteristic in the band. Thinner quartz substrates do not satisfy the condition well in the 5∼8 µm band. The radiation properties of quartz films of different thicknesses are discussed in the Supplemental material.

The ideal spectra of the MFS are shown as the red line in Fig. 1(d). Considering the peak shift resulting from processing errors, the ideal spectra should have a relatively broader peak at 1.06 µm [45]. The narrow peak centered at 10.6 µm can reduce the radiation intensity in the LWIR band as much as possible.

Recently, deep learning becomes very popular in nanophotonics design [45–48]. However, the radiation characteristics of MFS can be calculated with high speed by RCWA. Therefore, a combination of RCWA and GA to optimize the number of layers, thickness and material of each layer in the MFS is relatively more efficient and accurate. As shown in Fig. 1(b), GA starts with the initialization of the population and each layer contains two attributes, i.e., material and layer thickness, which are randomly selected from the corresponding library. After the initialization of the population, GA enters the process of fitness calculation. If the termination criteria are not reached, GA will perform selection, crossover, and mutation (the genetic operator) to produce individuals of the next generation and reenter the fitness evaluation. The optimal individual is finally obtained by selecting from generation to generation. In the application of GA, we define a fitness function to evaluate the individuals as:

3. Results and analysis

3.1 Multi-band camouflage

After 1000 iterations of GA, an individual with the fitness value of 15.14 was finally selected, generating a 9-layer MFS. The corresponding values of ${\mathrm{\bar{\alpha }}_{\textrm{1}\mathrm{.0\sim 1}\mathrm{.1}\mathrm{\mu}\textrm{m}}}$, ${\mathrm{\bar{\varepsilon }}_{\textrm{MWIR}}}$, ${\mathrm{\bar{\varepsilon }}_{\mathrm{5\sim 8}\mathrm{\mu}\textrm{m}}}$, α_10.6µm and ${\mathrm{\bar{\varepsilon }}_{\textrm{LWIR}}}$ are 0.98, 0.12, 0.66, 0.99, and 0.21, respectively. The geometric structure of the optimized MFS is shown in Fig. 1(c), and the thickness of each layer is given in Table 1. Note that the layers are numbered from top to bottom.

Table 1. The thickness and material of each layer of the optimized MFS (from top to bottom)

View Table | View all tables in this article

We first inspect the IR emission/absorption properties of the optimized MFS. Figure 1(d) shows the IR emission/absorption spectra of the optimized MFS, along with the ideal spectra and the transmission spectra of the atmosphere calculated by MOTRAN. As shown in Fig. 1(d), the designed MFS exhibits low emissivity in the MWIR and LWIR bands, while displaying a broad emission peak in the non-atmospheric window band from 5 to 8 µm. At the temperature of 523 K, for instance, the averaged emissivities of the MFS in MWIR, non-atmospheric window and LWIR bands are 0.12, 0.67 and 0.21 [see Eq. (2)]. At room temperature (300 K), the averaged emissivities are 0.1, 0.6, and 0.21, respectively. At the Nd:YAG lidar wavelength of 1.06 µm, an absorptivity as high as 0.99 is observed by the optimized MFS. Moreover, high absorptivity is realized at wavelengths around 1.06 µm. At the CO₂ lidar wavelength of 10.6 µm, the MFS has narrowband absorptivity as high as 0.99 with a bandwidth of 0.31 µm (the bandwidth is defined as the full width at half-maximum). As a result, the optimized MFS not only has high absorption of the 10.6 µm CO₂ lidar light, but also remains low averaged emissivity in the LWIR band. Thus, the optimized MFS satisfies the requirements for camouflage in the multiple IR bands. Furthermore, practical MIR stealth requires low averaged emissivity of the structure in the two atmospheric window bands (smaller than 0.3 [7,13,25]). Although the spectral emissivity in 8∼10 µm of the optimized MFS is not very small, the overall low averaged emissivity in 8∼14 µm can satisfy the requirement of MIR stealth (${\mathrm{\bar{\varepsilon }}_{\textrm{LWIR}}}$=0.21). In addition, the small emission peak of the MFS in 9∼10 µm is attributed to the intrinsic absorption of SiO₂ in the layers. As for the radiative cooling effect in non-atmospheric window band, we note that notable radiative cooling effect were experimentally demonstrated in previous works like Refs. [7,13,34], the averaged emissivity in the non-atmospheric band of our optimized structure is comparable to or even larger than those in the previous works. Thus, the MFS, if properly fabricated, is also expected to have a notable radiative cooling effect.

In addition to the IR bands, the MFS can also possess various structural colors in the visible band. In fact, the top two layers (TiO₂ and Ge) are responsible for the visible structural colors. Since the Ge layer with a thickness of 0.693 µm is nearly opaque in the visible band, the spectral reflectivity of the MFS in the visible is determined by the thickness of the top TiO₂ layer (h_TiO2). Meanwhile, the top two layers are highly transparent at IR wavelengths, thus having minimal impact on the thermal emission from the multiple layers below. We calculated the visible-band spectral reflectivity of the MFS with different h_TiO2 (ranging from 70 to 200 nm with 10 nm intervals) and converted them into CIE-1931 chromaticity coordinates, as shown in Figs. 2(a) and (b). The spectral reflectivity refers specifically to the hemispherical reflectivity for normal incident wave. With the variation of h_TiO2, we demonstrate the different structural colors generated by the MSF from red to green. Figure 2(c) presents the IR emission spectra of the designed MFS for different colors obtained solely by altering the thicknesses of the top layer of the MFS. We can see that while the variation of the thickness of the top layer leads to different colors of the MFS, it does not affect the IR spectral emissivity/absorptivity much. The thickness of the top layer only has non-negligible effect on the absorptivity around 1.06 µm. To see the effect concretely, Fig. 2(d) illustrates the structural colors for different h_TiO2 and the corresponding absorptivity of the MFS at the wavelength of 1.06 µm. When h_TiO2 varies from 70 to 200 nm, the absorptivity of the MFS at 1.06 µm varies in the range of 0.57∼0.99. The averaged absorptivity at 1.06 µm across the different structural colors reaches 0.82. Thus, by adjusting the thickness of the first layer, the MFS can exhibit diverse colors while keeping good performance in the IR bands. These distinct colors can be used for visible camouflage in different environments, e.g. a thickness of 70/170 nm of the top layer can be suitable for ocean or sky, 120 nm for desert or mountains, and 200 nm for forest or meadow.

 

Fig. 2. Compatibility of visible structural colors with IR emission/absorption. (a) Reflected structural colors generated by the designed MFS for varying thickness of the top TiO₂ layer h_TiO2 in the CIE-1931 color space. The dashed line represents route of the color change as h_TiO2 varies from 70 to 200 nm with 10 nm intervals; (b) The visible reflection spectra of the designed MFS as h_TiO2 varies from 70 nm to 200 nm, the color of the line represents the structural color of the designed MFS with the corresponding h_TiO2; (c) The IR emission/absorption spectra of the designed MFS as h_TiO2 varies from 70 nm to 200 nm; (d) The structural colors generated by the designed MFS for different h_TiO2 and the corresponding absorptivity at 1.06 µm.

Download Full Size | PDF

The angular dependence of the emission and absorption properties of MFS is also an important factor for camouflage. Figure 3 demonstrates the angular dependence of the emissivity/absorptivity in the IR bands. For lasers stealth, the key point is that the MFS can achieve relatively high absorptivity at the laser wavelengths around the normal directions. For large incident angles, the signal will be mainly reflected to other directions so that the detector will not receive the reflected signal. At 1.06 µm and 10.6 µm, particularly, as shown in Fig. 3 (a) and (b), the designed MFS has relatively large absorptivity for incident angles up to 60° (the averaged absorptivity is 0.91 at 1.06 µm and 0.52 at 10.6 µm for incident angle of 60°). For both TM and TE polarizations, as shown in Fig. 3(c) and (d), the emissivities in the two atmosphere window bands remain low while that in the non-atmospheric window band keeps high even at larger emission angles. In addition, the absorption peak of the optimized MFS at the laser wavelength shifts as the incident angle rises. Meanwhile, the MFS maintains a high absorptivity at 1.06 µm due to the wide absorption peak around the wavelength. Conversely, the absorption peak of the MFS at 10.6 µm is very narrow, and hence the absorptivity decreases rapidly when the absorption peak shifts. The absorptivity of the optimized MFS at 10.6 µm shows a peak again around 75°. This is related to the Brewster angle of the MFS.

 

Fig. 3. The spectral absorptivity for TM- and TE-polarizations, and the averaged value over the two polarizations as a function of incident angle at the wavelengths of (a) 1.06 µm and (b) 10.6 µm. Contour plot of the emissivity with respect to the wavelength and the incident angle for (c) TM-polarization and (d) TE-polarization.

Download Full Size | PDF

3.2 Physical mechanism

To reveal the underlying physics for the selective IR emission/absorption, the spectral reflectivity, transmissivity and absorptivity (R, T and A) of the 9 layers constituting of the designed MFS without quartz substrate are presented in Fig. 4(a). The distribution of the normalized electric field with respect to the wavelength and z-coordinate is presented in Fig. 4(b). As shown, the 9 layers form two distinct forbidden bands in the MWIR and LWIR, characterized by low transmissivity. In other words, most of the electromagnetic radiation within the forbidden bands are reflected out, and the 9 layers absorb little in these bands. In the non-atmospheric window band (5∼8 µm), however, a small proportion of the electromagnetic radiation is absorbed by the 9 layers, while a large portion can transmit through them. It can also be seen from the normalized electric field map distribution that the electric field intensity within the MWIR and LWIR bands is universally low, forming two distinct forbidden bands. The reason for the appearance of such forbidden bands is the destructive interference effect caused by the stack of dielectrics with different refractive indexes [44]. Consequently, when combined with a quartz substrate, the thermal emission from the quartz substrate can efficiently transmit the upper 9 layers only in the non-atmospheric band, the thermal emission in the forbidden bands is blocked. (see Section 1 of Supplement 1 for quartz’s radiation characteristics in different thicknesses). It can be observed from Fig. 4(b) that the electric field intensity within the 9 layers is indeed stronger in the non-forbidden band. Therefore, the combination of the 9 layers and the thick quartz substrate achieves selective emission in the MWIR and LWIR bands.

 

Fig. 4. (a) The spectral reflectivity, transmissivity and absorptivity of the upper 9 layers without the quartz substrate; (b) The normalized electric field intensity distribution as a function of wavelength from 3 µm to 14 µm in the upper 9 layers.

Download Full Size | PDF

As to the high absorptivity of the designed MFS at 1.06 µm, it is mainly caused by the intrinsic absorption of the second Ge layer (see Section 2 of Supplement 1 for a discussion of the materials’ optical constants) [49]. Figure 5(a) and (c) present the distribution of the electric field intensity and the normalized electromagnetic power loss density at 1.06 µm. As shown, the electric field intensity decays sharply in the second Ge layer and the electromagnetic losses are mainly located inside this layer. At the CO₂ laser wavelength of 10.6 µm, the narrowband high absorptivity can be attributed to two mechanisms: (i) the symmetric Fabry-Perot resonance in the cavity formed between the second and fourth layer [50]; (ii) the intrinsic absorption of the first and fifth TiO₂ layers. As shown in Figs. 5(b) and (d), the electric field at the wavelength of 10.6 µm is mainly confined in the third SiO₂ layer, indicating a localization of the electromagnetic power by a typical Fabry-Perot resonance in the cavity formed by adjacent layers. The Fabry-Perot resonance in the cavity causes a high aggregation of electric field intensity and large electromagnetic loss density. The electromagnetic power at 10.6µm is mainly dissipated in the first and fifth TiO₂ layer due to the intrinsic absorption of TiO₂. In short, the upper 5 layers mainly contribute to the formation of the Fabry-Perot resonance cavity and the intrinsic absorption of incident waves. Meanwhile, it can be seen from Fig. 4(a) that the upper 9-layer structure without the substrate has non-negligible absorption in the 5∼8 µm band, and we have verified that the bottom 4 layers can enhance the absorption in this band. Therefore, the repetitive layers are all physically significant.

 

Fig. 5. Distribution of the electric field intensity in the plane of y = 0 at the wavelengths of (a) 1.06 µm and (b) 10.6 µm. Distribution of the normalized electromagnetic power loss density in the plane of y = 0 at at the wavelengths of (c) 1.06 µm and (d) 10.6 µm.

Download Full Size | PDF

3.3 Overall camouflage performance

To better evaluate the overall camouflage performance (OCP) of the designed MFS, the OCP of the optimized MSF is evaluated through [33]:

As shown in Fig. 6(a), the OCP of the designed MFS decreases as the temperature rising from 293 K to 573 K. Notably, the highest OCP corresponds to earthy yellow of the MFS, indicating that the structure is most effective for stealth in the environment of desert or mountains. In addition, the MFS’s OCP is much higher than conventional full-band low-emissivity metal-based surfaces’ OCP (e.g. Au, Ag, Al, etc.) indicated by the gray dotted line in Fig. 6 (a). Figure 6(b) shows the radiation intensity of the MFS in the MWIR, 5∼8 µm and LWIR bands at different temperatures and their ratios included in Eq. (3). As the temperature increases, the radiation intensity of the designed MFS rises more sharply in 5∼8 µm compared with both MWIR and LWIR bands. This can be attributed to the peak-shift of the blackbody radiation intensity with temperature. The spectral radiation intensity of the MFS and the blackbody at the temperature of 300 K and 523 K are shown in Fig. 6(c). As temperature increases, the emission peak of the blackbody shifts towards shorter wavelengths. The designed MFS has higher emissivity in the non-atmospheric band compared with ordinary metal-substrate materials, which will lead to radiative cooling effect. Obviously, the radiated power of the designed MFS is spectrally selective.

 

Fig. 6. Overall camouflage performance (OCP) evaluation. (a) OCP of the designed MFS as a function of the temperature for different h_TiO2; (b) The radiation intensity calculated by Eq. (4) of the designed MFS in the MWIR, 5∼8 µm and LWIR bands and their ratio in Eq. (3) for temperatures varying from 293 K to 573 K; (c) The spectral radiation intensity of blackbody and the designed MFS in the MIR band at temperatures of 523 K and 300 K; (d) The signal reduction rate (SRR) calculated by Eqs. (5)∼(7) of the designed MFS in the MWIR and LWIR bands and the radiation intensity in the 5∼8 µm band as temperature varies from 293 K to 1093 K.

Download Full Size | PDF

MIR camouflage performance refers to the stealth capability of the structure in the passive detection bands or the two atmospheric windows. To further assess the MIR camouflage performance of the MFS, we evaluate the IR signal reduction rate (SRR) by the following equations [17]:

Herein, we list some recently reported results in the field of multi-band camouflage in Table 2. The majority of the reported works dose not achieve the compatibility of all the four camouflage bands. The simple 9-layer structure in this work, however, is compatible with all the four camouflage requirements and provides excellent performance in each band.

Table 2. Comparison of the recent reported results for multi-band camouflage with this work

View Table | View all tables in this article

4. Conclusions

In conclusion, a MFS composed of SiO₂, Ge and TiO₂ is designed to achieve multi-band camouflage for the IR, the visible and lasers with efficient radiative cooling by combining the methods of GA and RCWA. In terms of visible camouflage, the designed MFS shows tunable structural colors to match different backgrounds. For IR camouflage, low emissivities at the two atmospheric windows (${\mathrm{\bar{\varepsilon }}_{\textrm{MWIR}}}$=0.12 and ${\mathrm{\bar{\varepsilon }}_{\textrm{LWIR}}}$=0.21) are achieved. For lasers camouflage, high absorptivity can be achieved at both the 1.06 µm and 10.6 µm wavelengths. Meanwhile, the MFS has a broad emission peak in the non-atmospheric window, with an averaged value of 0.67. The camouflage characteristics of the designed MFS are applicable to large emission/incident angles. The underlying physics are the forbidden bands, the FP resonance formed by the multilayered structure together with the extra contribution from the intrinsic absorption by the materials. In addition, the overall camouflage performance of the designed MFS is evaluated, demonstrating its superior capabilities in camouflage applications.