1. INTRODUCTION

Energy consumption in buildings is reported to account for 20%–40% of the total energy consumption in developed countries [1]. According to the U.S. Department of Energy, more than 40% of the total primary energy in buildings is spent on heating, cooling, and ventilation [2]. In particular, energy consumption for space cooling is expected to increase significantly due to global warming and urbanization [3]. In recent years, the shift toward carbon neutrality and electric vehicle use has been accelerating worldwide. In electric cars, the battery is the only power supply for the driving and electrical components of the car. As a result, the reduction in air conditioner load is directly linked to the increase in traveling distance. Additionally, the dangers of heatstroke and death due to confinement in a car in midsummer have become social problems. A common cause of these two social problems is that the temperature environment inside the car largely depends on the near-infrared radiation from the sun entering through the window. Thus, the development and commercialization of heat shield films for automobile windows meet the actual needs of society.

A heat shield film is a material that transmits visible light while reflecting and absorbing infrared light. Therefore, it is possible to prevent the entry of infrared radiation and lessen the increase in the in-automobile temperature by applying a heat shield film to the window glass. This reduces the need for air conditioners and other appliances and therefore saves energy.

For heat-shielding coatings, the transmission of solar visible light is crucial. Because the coating is attached to the window, the field of view must be kept clear. Additionally, the near-infrared regime, which has one of the highest infrared solar photon energies and a wavelength range of 750–1400 nm, as shown on the spectral power distribution of solar radiation in Fig. 1(a), must be shielded. If the primary heat-shielding method is absorption, some of the absorbed heat is reradiated into space, reducing efficiency. Reflection is thus the preferred heat-shielding method.

 

Fig. 1. (a) The AM1.5 solar irradiation spectrum exhibits the Sun’s radiation at sea level (for more information, see: [56]). (b) Schematic illustration of the heat shield CSMM for application in 6G communication systems. (c) Dimensions of the CSMM unit cell.

Download Full Size | PDF

Various other types of heat shield materials have been studied, such as indium-tin-oxide-based coatings [4], dielectric coatings [5], and electrochromic [6] and thermochromic materials [7]. For instance, because environmental factors can change the characteristics of electrochromic and thermochromic window reflection, transmission, and absorption, these windows can be used in both warm and cold climates. For such materials, it is still challenging to regulate the optical characteristics of reflection and transmission precisely and independently within a chosen wavelength range. Heat shield films based on typical dielectric/metal/dielectric coatings have a high infrared reflectance of over 80% and a fairly high visible light transmittance of over 60% for wavelengths between 400 and 700 nm [8]. However, the near-infrared reflectance is frequently less than 80% for wavelengths between 800 and 1300 nm [9]. Since the majority of the solar thermal energy delivered to the earth is concentrated in the near-infrared wavelength band of less than 1400 nm, a high reflectance in the near-infrared region is favored. Low reflectance (high transmittance) for wavelengths below 750 nm, high reflectance (low transmittance) for wavelengths above 800 nm, and a sharp shift in reflectance and transmittance for wavelengths near 700–800 nm are required to more effectively block solar infrared radiation [10]. Unfortunately, it is difficult to find natural materials that have both high transparency in the visible range and high reflectivity in the near-infrared range. However, by employing metamaterials, optical spectra can be artificially and systematically manipulated by varying structural parameters.

In real life, when applying the heat-shielding material as a window in moving vehicles with people inside, it should also have a high transmission of radio signal and millimeter waves (4G, 5G), especially, terahertz waves (6G and beyond) for telecommunication. Mankind is now using 5G communication and preparing the infrastructure for the revolution of 6G and beyond 6G communication, which utilizes terahertz waves. This revolution deals with increasing the frequency bandwidth by using shorter wavelengths of electromagnetic waves, which allows more and more devices to connect at the same time with high speeds of data downloading and uploading. Thus, heat-shielding windows in the future should have high visible light transmission, high near-infrared reflection, and high transmission of the terahertz wavelength range with low incident angle dependence for applications in 6G and beyond 6G communication systems.

The most basic type of heat-shielding film is created by sputtering a certain thin film on a metal foil [11]. However, this film also reflects visible light, which causes window glare and makes windows difficult to see through, and inhibits radio and terahertz waves, which makes it difficult to use mobile phones and other devices. Dielectric multilayer films consisting of a stack of nano-thin films with various refractive indices are recognized for reflecting electromagnetic waves at particular wavelengths and having a high level of terahertz wave transparency [12]. However, the large incident angle dependency of the transmission and reflection spectra makes it difficult to achieve uniformity over a large area. A material based on silver nanoplates has been created as a near-infrared reflecting material that addresses the issues of radio wave transparency and angle dependency [13]. This material was created utilizing the photographic film manufacturing process of tabular silver halide grain formation. Surface plasmon resonance was employed to boost the reflection in the near-infrared region by dispersing nanosized silver tabular grains with a particular areal density. However, the maximum reflectance was low (${\sim}{53}\%$).

Metamaterials are materials that have been artificially endowed with electromagnetic properties by using subwavelength microstructures. The optical properties arise from their structures and arrangements. Therefore, it is possible to obtain materials with desired optical properties that do not exist in nature. Metamaterials with various shapes have been used widely for applications, such as perfect absorbers, filters, and force and refractive index sensors [14–30]. Metamaterials have been found in the field of near-infrared shielding [31,32]. However, the optical properties in the terahertz range of the heat-shielding metamaterials have not been studied.

Among various metamaterials, cross-shaped metamaterial (CSMM) has been used for interaction with electromagnetic waves in some frequency ranges, such as modulation of a terahertz bandpass [33,34], absorbers in microwave and near-infrared ranges [17,35]; and infrared resonators [36]. However, engineering functional CSMM in a large area for heat-shielding window applications with high transmission and low incident angle independence of terahertz waves is still challenging.

For effective reflection in the infrared range, the metamaterial structure should have subwavelength dimensions of tens to hundreds of nanometers, which is beyond the diffraction limit of traditional photolithography. Thus, other advanced lithography techniques should be used for controllable fabrication. Various advanced lithography methods have been employed for the fabrication of nanometer-scale metamaterials, such as deep-UV lithography [37], electron beam (EB) lithography [38–42], nanoimprint lithography [43], focused ion beam lithography [44], and scanning probe lithography [45]. Among them, EB lithography is a reliable, high- resolution (down to 10 nm) technique that was employed in this research to produce a heat shield metamaterial with high performance.

 

Fig. 2. (a) Boundary conditions applied for one CSMM unit cell in the simulation model. (b) Simulated transmission band, reflection band, and absorption spectrum of the CSMM in the visible and near-infrared regimes. (c) Distribution of the tangential electric field vector ${\boldsymbol {\vec E}}$ (red color vectors) around the cross-shaped structure in the $y - z$ plane at the center of the ${y}$-bar; electric dipole ${\boldsymbol {\vec P}}$ (cyan color vectors) across the ${y}$-bar; the inset image displays the distribution of the electric dipole vector in the upper surface of the bars; the color depicts the magnitude of ${{E}_y}$.

Download Full Size | PDF

By using EB lithography for the fabrication of CSMMs at the nanometer scale, we demonstrate the ability to engineer the size and shape of the metamaterial array for tailoring the optical properties of the metamaterials precisely and achieving high-performance heat-shielding windows. The low throughput of the EB technique narrows its applications, and one may be concerned about its high cost, small sample size, and ease of surface damage rise. However, when EB lithography is employed to fabricate a mold for the nanoimprint technology, it can be used for mass production of the shielding window at a lower cost. Furthermore, in some specific fields, high-performance heat-shielding windows outweigh the high price, and high-speed communication (6G and beyond 6G) is required. For example, in aerospace, where surface damage is minimized, and a clear view of space is the most desired, the heat-shielding window may have its application.

In this paper, the design, fabrication, and simulation of a heat shield metamaterial are realized with nanometer-scale cross-shaped structures using EB lithography. The metamaterial structure was designed to achieve a high-performance heat-shielding window by reducing the absorption and increasing reflection in the near-infrared range while maintaining visible transparency. High visual transparency (77.2%), near-infrared reflectance (86.1%), and low near-infrared absorption (${\lt}{20}\%$) over a wide range of incident angles were achieved. Especially, high terahertz transmittance (up to 82%) was also achieved for 6G communication system applications. Numerical simulation was performed to optimize and design the metamaterial structure.

2. DESIGN AND NUMERICAL CALCULATION

A. CSMM Design

A functional heat shield film was designed to combine high terahertz transmission, high visible transparency, and high near-infrared reflection. In particular, terahertz transmission is required for use in vehicles that are equipped with 6G communication systems. A metamaterial device comprises a dielectric substrate and a metasurface array on top. Metasurfaces can be made of metal (Au, Ag, Al, and so on), dielectric, or a combination. From the design point of view for the heat-shielding window, the substrate must have high transmission of visible light and terahertz waves, and the metasurface array should reflect near-infrared efficiently. The dielectric metasurface has lower loss owing to its non-plasmonic nature (no absorption due to surface plasmon resonance), which benefits from high near-infrared reflection. However, the reflection band of the dielectric metasurface is narrow and dependent on the incident angle [46–48]. The design and fabrication of dielectric metasurfaces of combining metal with dielectric metasurfaces with a large reflection bandwidth of near-infrared are complicated [31,49]. The fabrication in a large area is still challenging for dielectric metasurfaces. Furthermore, dielectric metasurfaces may suffer from low transmission in the terahertz range [50].

Considering the pros and cons, we employed aluminum, which is a cheap metal for fabrication of CSMM in a large area; metal metasurfaces at the nanometer scale can be realized precisely using VLSI or MEMS technology, which has been well established and widely used in industry. Furthermore, by well designing the dimensions of the CSMM, the reduction in near-infrared reflection due to the plasmonic effect can be minimized [shown in the absorption curve in Fig. 2(b)].

A schematic illustration of the CSMM is shown in Fig. 1(b). A quartz (${{\rm SiO}_2}$) substrate was employed because of its visible and terahertz transparency. The cross-shaped metamaterial (CSMM) structure comprised two aluminum bars in the $x$ and $y$ directions (${x}$-bar and ${y}$-bar) that are perpendicular to each other with periodicity, which allowed the metamaterial structure to interact with the electric field components of incident light with polarization independence for normal incident light. $L$, $w$, $p$, and $t$ denote the length, width, pitch, and thickness of the CSMM, respectively. The dimensions of the CSMM unit cell are depicted in Fig. 1(c). The thickness of the ${{\rm SiO}_2}$ substrate was 532 µm. Aluminum was chosen instead of noble metals due to its cost effectiveness.

The near-infrared reflectance was increased by varying the length of the bar (${L}$) and designing the resonant wavelength to be in the near-infrared region. By adopting the cross structure, the localized surface plasmons were less affected by the surrounding surface plasmons, and the incident angle dependence was reduced [51]. In addition, the visible transmission was increased by reducing the area occupation ratio of the metal.

B. Numerical Simulation

The optical properties of the CSMM were calculated in the frequency domain using a finite element method based on Maxwell equations via COMSOL Multiphysics (version number 6.1). Numerical simulation was employed to design the CSMM with dimensions that satisfied the requirements of heat-shielding windows. The aluminum cross shape was modeled as a 3D structure in the visible and infrared regions. In the visible and infrared regimes, the refractive indices of thin aluminum film and the quartz substrate are frequency dependent. The refractive index (real part and imaginary part) of aluminum was referenced from the experimental data of Mathewson and Meyers in [52]. The refractive index (real part and imaginary part) of the quartz substrate was referenced from the experimental data of Ghosh in [53]. For simulation of the CSMM array, periodic boundary conditions with the Floquet periodicity type were applied to one unit cell of the CSMM as shown in Fig. 2(a). The periodic ports are utilized for electromagnetic wave excitation and deriving transmission as well as reflection responses. One may use other boundary conditions, such as a perfect electric condition (PEC) and perfect magnetic condition (PMC) for simulation of one unit cell. In that case, it is necessary to determine how many periods are necessary to achieve a collective effect; one should determine that the number of periods should be tuned as reported in [54].

To prevent additional responses due to interference waveforms, perfectly matched layers (PMLs) that could absorb all scattered waves and scattering boundary conditions were set at the top and bottom boundaries of the model. The plane-wave electromagnetic wave was incident perpendicular to the surface of the CSMM array and polarized in the $y$ direction. For the simulation in the terahertz regime (0.2–2 THz), a transition boundary condition was applied with a depth equivalent to the thickness of the cross-shaped aluminum structure, and the complex refractive index of the ${{\rm SiO}_2}$ substrate was taken from our experimental data as ${n} = {1.96} + {j*0.001}$; the refractive index value of aluminum was one. The electrical conductivity of aluminum was ${3.5} \times {{10}^7}\;({\rm S}/{\rm m})$ in both cases, and the incident electromagnetic wave power was set to 1 W.

The simulated reflection band (${\gt}{50}\%$) in the near-infrared region ranges from 0.9 to 1.4 µm with a resonant reflection wavelength of 1.08 µm corresponding to 86% of reflectance, as shown in Fig. 2(b), in which $t$, $p$, $w$, and $L$ are 91, 460, 57, and 352 nm, respectively; the substrate thickness is 300 µm. This resonant peak is caused by the collective oscillation of the free electrons, which are driven by the $y$-polarized electric field component (${{E}_y}$) of the incident light, leading to the redistribution of electric charges in the ${y}$-bar, which is evidenced by the formation of an electric dipole ${\vec P}$ [as shown in Fig. 2(c)]. This electric dipole has a maximum value at the resonant reflection wavelength that is four-fold higher than that at the wavelength of 2 µm. These electrons re-emit light almost in the reflective direction due to the collective oscillation.

The simulation transmission band in the visible region ranges from 0.4 to 0.75 µm with a transmittance higher than 50%. This transmission band originates from the low absorption and reflection (${\lt}{20}\%$), as shown in Fig. 2(b). The optical characteristics simulation proves that the CSMM at the nanometer scale can serve as a heat-shielding window.

In the visible and near-infrared regimes, the wavelength is comparable to nanostructure dimensions of some tens and hundreds of nanometers, greatly affecting light interaction with the CSMM. The optimization simulation was performed by parametric sweeping to determine the influence of each CSMM dimension parameter on the optical responses: $t$ was varied in the range of 40–80 nm, $w$ in the range of 70–130 nm, $p$ in the range of 400–550 nm, and $L$ in the range of 250–450 nm. The simulation results of the transmission and reflectance spectra are shown in Fig. 3. The initial dimensions of the structure used in the calculation are as follows: $t$, $p$, $w$, and $L$ are 91, 460, 57, and 352 nm, respectively; the substrate thickness is 300 µm. As shown in the graph, by alternately arranging structures with different dimensions, peaks appear for each structure at the resonant reflection wavelengths in the near-infrared regime, and the position and transmittance of the transmission band change.

 

Fig. 3. Reflection and transmission spectral dependence on: (a), (b) aluminum thickness; (c), (d) width; (e), (f) pitch; and (g), (h) length of the CSMM.

Download Full Size | PDF

Figures 3(a)–3(d) show the reflectance and transmittance when $t$ and $w$ are varied. As $w$ and $t$ increase, the resonant reflection wavelength blue shifts, and the visible transmission band shifts to a lower transmittance. The behavior of the transmission response in the visible range on the transmission curve shown in Fig. 2(b) could be explained for the energy redistribution by the excitation of ($\pm {1},0$) SPP resonances and Wood’s anomalies in the periodic array [51,55].

When $p$ increases, the resonant reflection wavelength almost does not change, but the reflection bandwidth narrows in the near-infrared regime, and the visual transmission band widens and shifts to a higher transmittance, as shown in Figs. 3(e) and 3(f). This can be explained by the decreased coupling distance between the two dipoles of nearby CSMM units, which results in a smaller bandwidth of the reflection band and a widening of the transmission band.

On the other hand, as $L$ increases, the reflection in the near-infrared regime changes dramatically: the resonant reflection wavelength in the near-infrared regime has a large red shift, and the reflection bandwidth is also broadened. In the visible regime, the transmission band shifts to a higher transmittance, as shown in Fig. 3(h).

Considering the influence of the parameters on the transmission band and reflection band shown in Fig. 3, we propose a structure that has high near-infrared reflectance, high visible transparency, and a sharp transition between the two bands. The CSMM structure chosen for fabrication was designed with the following dimensions: $t = {60}\;{\rm nm}$, $p = {460}\;{\rm nm}$, $w = {100}\;{\rm nm}$, and $L = {360}\;{\rm nm}$.

3. CSMM FABRICATION

The fabrication process of the CSMM is illustrated in Fig. 4. Four-inch quartz was used as the substrate, which was cleaned of organic and metallic particles [Fig. 4(a)]. After treatment in acetone with ultrasonication for 10 min, the substrate was immersed in isopropanol for 10 s prior to drying gently by air blowing.

 

Fig. 4. Fabrication process flow: (a) wafer cleaning; (b) photoresist spin-coating; (c) E-beam writing and development; (d) aluminum deposition; (e) lift-off.

Download Full Size | PDF

The resist coating was applied to the substrate by a spin-coater [Fig. 4(b)] with three layers: OPA (HMDS) thin film as the adhesion agent between the resist (ZEP 502A) and the substrate, and a conductive thin film of Espacer 300Z to promote the absorption of the electron beam. EB drawing (Advantest F7002S) was carried out on an area of ${4.0} \times {3.3}\;{{\rm cm}^2}$ with a total dose of ${136}\;\unicode{x00B5} {{\rm C/cm}^2}$ to define the negative resist pattern [Fig. 4(c)].

The thin aluminum film [Fig. 4(d)] was deposited by ultrahigh vacuum vapor deposition equipment (NSP2). The CSMM array on the quartz substrate was formed after a lift-off process. Next, the residual resist was removed by immersing the sample in piranha solution with ${{\rm H}_2}{{\rm O}_2}$ (35%) and concentrated ${{\rm H}_2}{{\rm SO}_4}$ (98%) with a ratio of 1:2 for 10 min at 80°C [Fig. 4(e)]. Finally, the sample was cleaned by immersion in DI water three times for 10 min followed by dry spinning and kept in a vacuum desiccator before performing structural and optical characterization.

4. RESULTS AND DISCUSSION

A. Structural Characterization of the CSMM

Metamaterial patterning at the nanometer scale over large areas ranging from a few square centimeters is a difficult challenge that requires highly productive exposure schemes. EB lithography was employed to fabricate the CSMM. Nevertheless, it is suggested that scalable large-area patterning techniques, such as deep-UV or extreme-UV lithography, be used to create nanostructures over a larger area. The EB dose dramatically affects the shape of the patterned structure. The optimal dose was used with a base dose of ${100}\;\unicode{x00B5} {{\rm C/cm}^2}$ and a dose ratio of 1.36 (total dose of ${136}\;\unicode{x00B5} {{\rm C/cm}^2}$) for a design bar width of 60 nm. With this electron irradiation condition, the shape of the as-developed resist pattern was the most uniform.

After the lift-off step, the Al nanostructure of the CSMM was fabricated on the ${{\rm SiO}_2}$ substrate. The low-magnification field emission scanning electron microscopy (FESEM, JEOL JSM-6335F) image shown in Fig. 5(a) shows uniform periodicity over a large area of the CSMM array on the ${{\rm SiO}_2}$ substrate. The high-magnification FESEM image shown in Fig. 5(b) with a top view close to the CSMM structure was used to determine the average width, length, and pitch by using image analysis. The scanning probe microscope (SPM, NanoNavi IIs) image shown in Fig. 5(b) shows the 3D structure of one CSMM unit, which provides information on the thickness of the aluminum structure. The CSMM design and fabrication dimensions are shown in Table 1. The deviations between the design and fabrication dimensions of 3, 9, 8, and 1 nm were determined for the thickness, width, length, and pitch, respectively. As indicated by the parametric sweep simulation study above, these deviations only lead to a small difference in optical response. As a result, nanomachining using EB lithography can be used to engineer metamaterials with high accuracy, and thus, the optical response of the CSMM could be tuned precisely by varying the dimensions at the nanometer scale.

 

Fig. 5. Field emission scanning electron microscopy (FESEM) images with magnifications: (a) $\times {10000}$; (b) $\times {80000}$. (c) Scanning probe microscope (SPM) image of the CSMM.

Download Full Size | PDF

Table 1. CSMM Design and Fabrication Dimensions

View Table

 

Fig. 6. (a) Transmission and (b) reflection spectra in the visible and near-infrared regimes at different positions on the substrate area. The inset image in (a) displays the measured positions on the wafer.

Download Full Size | PDF

B. Optical Characterization of the CSMM

A UV-VIS system (JASCO V-570) was used for the optical characterization of the fabricated CSMM in the wavelength range from 0.4 to 2 µm. Three measurements were taken to average the optical data. Air and silver references were used for the transmittance and reflectance measurements, respectively.

To investigate the uniformity on a large scale, five measurements of the CSMM area were made at the center and four near-edge positions of the wafer. As shown in Fig. 6(a), the transmission was measured at various locations on the substrate, and the resulting spectrum of a sample area of ${3.3}\;{\rm cm} \times {4}\;{\rm cm}$ shows good homogeneity in the visible regime. The maximum transparency in the visible range is 77.2%. While the resonant reflection wavelength remains constant in the reflection spectra shown in Fig. 6(b), the reflectance varies, ranging from 86.1% at the highest to 71.4% at the lowest. According to the authors’ knowledge, a maximum reflectance and transmission of 86.1% and 77.2%, respectively, are the highest values reported in the literature for heat shield materials. The simulation results of the near-infrared reflectance and visible transmittance agree well with the experimental results, as shown in Figs. 7(a) and 7(b). The simulated resonant reflection wavelength is close to that measured experimentally, while the near-infrared reflectivity is higher and the visible transmittance is lower in the simulation, which results from the light polarization ($y$ polarization) in the simulation being different from that used in the experiment (random polarization). The differences also come from the unrealistic shape of the as-fabricated cross-shaped structures; the width and thickness of the bars decrease from the center to the end; the lengths of the bars are also not uniform; and all the sharp corners in the design where the strong electric field concentrates are rounded [Figs. 5(b) and 5(c)].

 

Fig. 7. Experiment and simulation in the visible and near-infrared regimes of the (a) reflection and (b) transmission. (c) Transmission in the terahertz regime.

Download Full Size | PDF

 

Fig. 8. Measurement and simulation of the oblique incident-angle-dependent transmission of the CSMM in the (a), (b) visible and near-infrared regimes and (c), (d) terahertz regime.

Download Full Size | PDF

A time-domain spectrometer (TeraProspector, Nippo Precision Co., Ltd.) was used to characterize the device in the terahertz regime. The terahertz transmission response of the CSMM was measured in the range of 0.2–2 THz [Fig. 7(c)]. As the terahertz frequency increases, the transmittance decreases gradually. In the terahertz frequency range, the maximum transmittance is 81% at 0.2 THz, and the transmittance decreases to 50% at 2 THz. In this frequency regime, the wavelength is much longer than the dimensions of the CSMM; therefore, the optical responses originate from the quartz substrate and the fill factor of the CSMM. The fill factor of the design is calculated to be 26.4%. The simulation result is higher than that of the experiment, which results from the absorption of the quartz substrate being higher in the experiment than in the simulation but remains in good agreement.

Transmission of light and terahertz waves through heat shield materials at various oblique angles is important for application in 6G communication systems. Figure 8(a) shows the change in the optical responses of the as-fabricated CSMM when the incident light angle (theta) varies from 0° to 60°. Theta is defined between the wave incidence with the normal direction of the substrate. The visible transmission spectrum decreases as the angle increases but remains almost above 50%, which is suitable for viewing through the CSMM wafer. The simulation transmission shown in Fig. 8(b) agrees with the trend of the experimental transmission spectrum with an increasing oblique angle.

In the terahertz regime, as shown in Fig. 8(c), the experiment transmission spectrum decreases as the oblique incident angle increases as shown in Fig. 8(c). For all cases, the transmission is lower than 82%, higher than 60% in the frequency range from 0.2 to 0.87 THz, and higher than 40% for frequencies lower than 1.3 THz, which is still acceptable for application in 6G communication systems under real conditions, i.e., for maintaining continuous signal connection under all conditions in moving vehicles equipped with 6G systems. The decrease in the terahertz transmittance due to the oblique angles originates from the quartz substrate because the fill factor is unchanged in this case. The terahertz transparency can be increased by improving the substrate quality. The experimental tendency of a decreasing transmittance with an increasingly oblique angle is confirmed by the simulation transmittance depicted in Fig. 8(d). From the result for terahertz wave transparency, it is speculated that the CSMM heat-shielding windows would transmit microwaves (5G communication) and RF waves (traditional communication) because the quartz substrate is transparent with microwaves and RF waves, while metamaterials at the nanometer scale do not interact with these waves due to their much longer wavelengths compared to the metamaterial’s dimensions.

To demonstrate visible transparency, the fabricated CSMM wafer was exposed to simple and complicated light conditions. When exposed to a single backlight [Fig. 9(a)], the observed image is slightly opaque at a normal angle and becomes darker at an incident light angle of 60°, as shown in Figs. 9(b)–9(d); however, the image can still be seen clearly. The data for the transmission measured by the UV-VIS measurement system corresponding to incident angles (theta) of 0° and 60° were used to calculate the color values using the CIE1931 diagram. The normalized tristimulus values calculated for incident angles of 0° and 60° were (${x} = {0.313}$; ${\rm y} = {0.324}$) and (${x} = {0.317}$; ${y} = {0.327}$), respectively, which are so close that their positions coincide on the CIE1931 diagram as shown in Fig. 9(i). The transmission color of the CSMMs wafer in Fig. 9(d) looks similar to the calculated color on the CIE diagram. Under complicated light exposure conditions [Fig. 9(e)], the sample is exposed to light sources from every direction. The fabricated CSMM wafer also overcomes this situation in that it shows good transparency at a normal incident angle, and its transparency becomes hazy at an incident light angle of 60°, but the observed image still exhibits good visibility through the CSMM wafer as shown in Figs. 9(f)–9(h).

 

Fig. 9. Illustration of light exposure under (a) simple and (e) complicated conditions. Optical images of objects under (b)–(d) simple exposures and (f)–(h) complicated exposures in three cases: without the CSMM; with the CSMM at a normal angle and 60°, respectively. (i) Color palette with a black marked dot at the color coordinates on the CIE1931 diagram for transmission color.

Download Full Size | PDF

5. CONCLUSIONS

A transparent heat-shielding window was successfully fabricated using EB lithography. High visual transparency (up to 77.2% and higher than ${\gt}\;{50}\%$) and infrared reflectance (up to 86.1%) and low infrared absorption (${\lt}\;{20}\%$) over a wide range of oblique incident angles were achieved using a nanometer-scale CSMM engineered by EB lithography. These visible and near-infrared features of the CSMM are suitable for application in heat-shielding windows. Furthermore, a high transmittance of higher than 40% and up to 82% from a low terahertz frequency to 1.3 THz at various angles of incidence was also achieved, indicating the heat shield metamaterial as a candidate for application in electric vehicles equipped with 6G communication systems in the future of the internet of things. Numerical calculation with the assistance of COMSOL software was used to design and simulate the CSMM, resulting in optical responses that were in good agreement with the experimental results. This research greatly contributes to the future practical applications of metamaterials.