Broadband absorber using ultra-thin plasmonic metamaterials nanostructure in the visible and near-infrared regions

In this work, an ultra-thin plasmonic metamaterial nanostructure absorber is simulated using the finite difference time domain method in the visible and near-infrared regions. A periodic square titanium-silica cap coated with glass medium is mounted on top of a silver substrate. The presence of the glass host enhances the absorption bandwidth by 276%. With an almost perfect metal–insulator-metal absorber, over 90% absorbance has been obtained for wavelengths from 440 to 1850 nm, producing an absorption bandwidth of 1410 nm. The impact of changing dimensions and using different materials on the absorption spectra has been investigated in both visible and near-infrared regimes. The considered metals for the top layer are titanium, nickel, silver, aluminum, and gold; however, the insulators are silica, quartz, vanadium dioxide, methyl methacrylate, and aluminum dioxide. In addition, aluminum, silver, copper, and gold are then simulated as a substrate. The optimum structure, which produces the maximum absorber bandwidth, 1410 nm, with a higher absorption, over 90%, is Glass-Ti–SiO2–Ag. The finding illustrates that the optimum dimensions of the Ti–SiO2 cab and the square base unit cell of the silver substrate are 250 nm and 200 nm, respectively. Finally, the absorption bandwidth is calculated using different polarization angles ranges from 10° to 70° with a step10°.


Introduction
Plasmonic metamaterials absorber has drawn attention in the last decade for various applications. Such as solar cells, nano-antennas, ultra-sensitive sensors, photovoltaic, thermal emitters, optical devices, optical switches, photoresistors, filters, and modulators (Pradhan et al. 2020;Maier and Brueckl 2010;Wang et al. 2012;Wu et al. 2012;Li et al. 2012;Askari 2021;. Recently, light absorption bandwidth enhancement of plasmonic nanostructures is the trend that needs to be achieved Dragoman and Dragoman 2008;Segovia et al. 2015;Shin et al. 2012;Ni et al. 2015;Cui et al. 2014). Many techniques have been used to obtain perfect metamaterial absorbers, such as cylinder array, hole array, and multilayer structure (Watts et al. 2012, Ra'Di et al. 2015Lei et al. 2018). However, these techniques have some problems, such as need noble metals or produce narrow bandwidth. A metal-insulator-metal nanostructure is the most common structure used as a perfect absorber for its simplicity and efficiency, especially in the ultraviolet, visible, and infra-red regions (Lei et al. 2018;Fann et al. 2019;Landy et al. 2009). Metamaterials are structured as a top metal periodic layer followed by an insulator as a spacer layer on the top of a perfect refractor metal (Cui et al. 2014). Many up-to-date works have been carried to obtain a wideband absorber using plasmonic metamaterials. Hao et al. introduced periodic structure of plasmonic metamaterial using three layers as Ag-Al 2 O 3 -Ag. They achieved absorbed light is in the visible region with very low bandwidth (less than 50 nm) (Hao et al.2011).
An ultra-broadband absorber has been obtained using Ti-SiO 2 -Al structure by L. lie and et al. (Ni et al. 2015). It provides a bandwidth of 712 nm that extends from visible to near-infrared regions. They used different materials for the top cap, such as Ni and Al, which produced a lower bandwidth than Ti material. C. Fann et al. introduce a multiband plasmonic metamaterials absorber in the infrared region using Ti-SiO2-Al structure (Fann et al. 2019). The proposed structure has been numerically simulated and experimentally tested at two prominent bands 4.8-7.5 µm and 9.7-10.5 µm with high absorption. Hedayati et. al. introduced a plasmonic nanocomposite metamaterial absorber structure with different concentrations . A gold nanoparticle with different concentrations is distributed on the top of the SiO2 spacer. The maximum obtained bandwidth in the visible region is around 500 nm at a gold concentration of 40%. Gao et al. provide a plasmonic absorber based on the insulator-metal-insulator-metal structure of SiO2-TiN-SiO2-TiN. The absorption bandwidth is almost 1 µm, from 200 to 1200 nm, from ultraviolet to near-infrared band for absorption over 90% (Gao et al. 2019).
In this work, a plasmonic metamaterial unit cell has been simulated using Finite Difference Time Domain method FDTD using a Lumerical tool to calculate the absorption from visible to near-infrared region, 0.3-3 µm. The absorption of the proposed structure is simulated with and without a glass cover coating. The optimum base dimensions of the bottom substrate, L, and the cap insulator, W, are also calculated. The absorbance is tested for different top layer metals; Nickel (Ni), Titanium (Ti), Aluminium (Al), and Silver (Ag). A Silicon Dioxide (SiO 2 ), Aluminium Oxide (Al 2 O 3 ), Quartz, ploy (methyl methacrylate), and Vanadium Dioxide (VO 2 ) are used as spacer material. At the same time, Al, Ag, Au, and Cu are tested as a bottom layer of the proposed structure. Finally, the absorption of the broadband absorber is simulated at different polarization angles of the incident light.

Maximum absorption wavelength and absorber structure
A metamaterial is an artificial material that generates a subwavelength that can be tailored using different structures and unit cell dimensions. For insulator-metal stack, the resonance wavelength max at which maximum absorption occurs is given by (Fann et al. 2019).
where n m is the refractive index of the metallic substrate, n i is insulator refractive index, n 0 is the superstrate refractive index, and t is the dielectric material thickness.
The introduced plasmonic metamaterial absorber nanostructure is periodic structure caps of metal-insulator coated with a glass host and placed on a metallic substrate. The periodicity of the square titanium-silica caps in both x and y coordinates requires periodic boundary conditions while using the perfect matching layer in the z-direction serves the absorbing boundary condition.
We have used a plan wave source with Bloch signal in the simulation, and the operating wavelength is in visible and near-infrared regions. The light pulse has a wavelength bandwidth from 0.3 nm to 3 nm. Then, we comprehensively investigate the absorbed wave by the proposed structure using the Lumerical FDTD electromagnetic wave solver. The proposed structure is illustrated in Fig. 1.
The absorbance (A) spectrum was then obtained by where A is the absorbance, T is the transmittance and R is the reflectance. Different materials in the proposed structure are tested to obtain the maximum absorption bandwidth. Ni, Ti, Al, and Ag have been considered for the periodic metallic cap, while SiO 2 , Al 2 O 3 , Quartz, ploy, and VO 2 have been used for the insulator layer. Finally, various metals such as Al, Ag, Au, and Cu are used as a substrate. (

Simulation results and discussion
Using the FDTD simulations, we thoroughly investigate the optical absorbance spectrum of the plasmonic MIM nanostructure absorber shown in Fig. 1. A bottom layer of silver with pitch L = 250 nm and thickness t = 200 nm is used as a substrate. Then, a periodic square cap of length W = 200 nm and thickness 100 nm is introduced on the upper surface. The cap consists of SiO 2 with height t 1 as a spacer and a top metallic layer of Ti with height t 2 = 20 nm. Firstly, conventional MIM plasmonic absorber in air surrounding is simulated. Then, we consider a glass coating layer for the Ti-SiO 2 cap in order to enhance bandwidth. The glass host thickness is t 3 = 100 nm. In this simulation, the refractive indices for the insulator layer and glass coating layer were gathered from the literature (Liu et al. 2018). The relative permittivity ε r (ω) of the dispersive Al nanoparticles array and back reflector film was determined using the Lorentz-Drude model (Palik 1985): where ε ∞ denotes the permittivity at infinite frequency, f m is a function of position specifying the oscillator strengths, and Γ m is the damping coefficient. The incident wave frequency and the resonant frequencies are respectively represented by ω and ω om . The refractive indices for the insulator layers and parameters used in Eq. (2) for the metallic layers are summarized in Tables 1 and 2 respectively. The absorbance, transmittance and reflectance spectra for both structures, with and without glass layer, are illustrated in Fig. 2. As shown, the transmittance is completely negligible by the thickness of the bottom metallic back reflector. Absorption spectra exhibit enhanced peaks opposite to highly attenuated troughs in the reflection spectra. Those absorption crests indicate less reflected optical energy due to perfect matching between the MIM absorber and surrounding. They may be associated with surface plasmon resonance at metal/insulator interfaces, metallic localized plasmon resonance, and dielectric Fabry-Perot cavity resonance (Lei et al. 2018;Fann et al. 2019). Broadening occurs due to low quality factor linked to highly lossy Ti. One may notice that the glass coating layer enhances the absorption bandwidth for A ≥ 90%. Over 0.9 absorbance, the bandwidth of the Ti-SiO 2 -Al structure is 510 nm. On the other hand, the glass: Ti-SiO 2 -Al structure exhibits extended bandwidth of 1315 nm. The plasmonic resonance at glass/Ti interface occurs at longer wavelength. The glass host layer enhances the incident light absorbance bandwidth by more than 257% for absorbance of more than 90% of the incident light.
We examine modifying the absorbance spectra of the reported absorber by changing the square cap side length W within a range of 180-220 nm with 10 nm step size. During this study, the other dimensions are kept fixed. Figure 3 shows the impact of changing cap dimensions on the absorbance. As the width increases, the absorption bandwidth gets more expansive, and the absorbance decreases. The maximum obtained absorption bandwidth is 1584 nm at W = 220 nm for absorbance higher than 87%. However, the minimum bandwidth is 1126 nm for absorption more than 90% at W = 190 nm.
(2) The optimum base dimension was 200 nm with absorbance bandwidth 1315 nm when the simulated structure was glass: Ti-SiO 2 -Al. Figure 4 illustrates the impact of changing pitch L of the sliver bottom layer on the absorbance spectra of the reported absorber. In this investigation, the periodicity L is modified within a range of 210-270 nm with a 20 nm step size while all other dimensions are kept unchanged. At L = 210 nm, the bandwidth is 1490 nm for absorption larger than 60% however bandwidth is 1610 nm at absorbance over 87% for L = 230 nm. For absorption over 90%, the absorbance bandwidth of 1350 nm and 1274 could be achieved at L = 250 nm and L = 270 nm, respectively. Hence, the base dimension L = 250 nm is more efficient and chosen for higher absorption and comparable bandwidth.  Figure 5 shows the absorbance spectra for the reported MIM plasmonic absorber with various cap top metals. The maximum absorbance, blue curve, is obtained when Ti material is used as a top metal of the cap. On the other hand, absorber with Au, Al, or Ag as a top layer allow a deficient absorption and high reflection in the visible and NIR regions. However, their absorbance is very high in the 0.3-0.6 nm band. In addition, the MIM plasmonic absorber achieves a high bandwidth when Ni metal is used as a top layer. But unfortunately, it provides a lower absorbance than obtained by Ti, especially in the near-infrared region.
We examine modifying the absorbance spectra of the reported absorber by changing the material of the insulator layer as it plays an essential role in the absorption bandwidth and absorbed light ratio. Figure 6 illustrates the absorbance spectra of the proposed nanostructure using a spacer of different materials. It is evident from this figure that using VO 2 as an insulating layer produced total absorption over 98%with bandwidth 814 nm. However, the bandwidth is 885 nm over 90% absorbance for Al 2 O 3 material. Also, the absorbance bandwidth of quartz is 577 nm over 88% absorption and 1125 nm for ploy for absorbance over 90%. The optimum material used as an insulator is the SiO 2 which gives 1315 nm absorption bandwidth over 90% of the absorbed light (Fig. 6). As most of the light either reflected or absorbed from the glass or cap structure, the portion of light transmitted to the bottom substrate is minimal, which means the effect of using different bottom substrates is also limited. Figure 7 illustrates the absorption when dissimilar metals are used as a bottom metal substrate.
The most common metals used as a bottom substrate are Al, Ag, Au, and Cu. Table 3 shows the absorption bandwidth, over 90%, for the given four used metals and the minimum absorption in the absorption range.
The maximum absorption bandwidth, 1410 nm with over 90% absorbance, is obtained when silver is used as substrate. The minimum absorption in the measured band is 92.5%, as shown in Fig. 8, while the maximum absorbance is 99% in the same range. Copper material has a very interesting comparable bandwidth, 1370 nm, with high absorbance. In addition, it is considered a cheap material to use in the fabrication process. Figure 8 illustrates the overall absorption bandwidth of the proposed nanostructure, Glass: Ti-Sio 2 -Ag, which gives a 1410 nm bandwidth over 90% absorbance and high absorbance in the visible and near-infrared band.  For the variation of the polarization orientation, a linearly polarized plane wave light source is attached either in TE mode (y-direction) or TM mode (z-direction) as illustrated in Fig. 9. We developed a study at angular incidence from 10° to 70° with a step10°. It was found that the structure is durable to the oblique incidence as shown in Fig. 10. The absorbance bandwidth is almost 2 µm for absorbance over 50%.   Table 4 gives an overall comparison between similar techniques and the introduced techniques. In Lei et al. (2018), the structure was Ag-Al 2 O 3 -Al with very low bandwidth, less than 50 nm, in the visible region with almost 100% absorption. The bandwidth increased to be 715 nm over 90% absorbed light in the visible and NIR regimes for Ti-SiO 2 -Al structure (Ni et al. 2015). In (Cui et al. 2014), authors introduce a similar design with different dimensions that provide a wide range of absorption in the Mid-IR region. They present two other bands, the first one from 4.8 µm to 7.5 µm and the second is from 9.7 µm to 1.05 µm. In Landy et al. (2009), the proposed structure was SiO2-TiN-SiO 2 -TiN with an absorption bandwidth of almost 1 µm, from 0.2 to 1.2 µm, in ultraviolet, visible, and near-infrared bands over 90% absorbance.
The proposed structure, Glass-Ti-Sio 2 -Ag, gives absorption bandwidth of 1410 nm for absorption over 90% in the visible and NIR regions.

Conclusion
A high absorber periodic ultra-thin plasmonic metamaterial nanostructure is produced in this work. The proposed structure consists of a regular cap of a metal-insulator structure, Ti-SiO 2 , placed on top of a metallic silver, Ag, coated and uncoated a glass environment. The absorption bandwidth is enhanced by 276%, from 510 to 1410 nm, using the glass host as a surrounding material over the substrate. Dimensions of both the bottom substrate and the periodic cap have been optimized to be 250 nm and 200 nm. Different top cap metal materials have been simulated, such as Ni, Ti, Al, and Ag. Titanium material gives the best performance as a cap metal and SiO 2 as an insulator material over Al 2 O 3 , Quartz, ploy, and VO 2 materials. On the other hand, silver material as a bottom substrate gives a superior bandwidth of 1410 nm, over Al, Au, and Cu for absorbance over almost 90%. In addition, we investigate the absorption spectra as a function of an oblique incident light polarization. Increasing the oblique angle increases the bandwidth and decreases the absorption to less than 50% of the incident light.