Artificially prepared electromagnetic metamaterial equipment has been widely computed and experimentally investigated since it has been confirmed theoretically and experimentally.^[1–3] In practice, the exotic properties of electromagnetic metamaterial equipment usually attain from the unit cell structure designed scheme rather than constituent materials. Based on the exotic property, metamaterial equipment was used in a wide range, such as superlens, and absorber.^[4–6] Especially, the metamaterial absorber has attracted a lot of research attention, since the metamaterial absorber was experimentally demonstrated by Landy et al.^[7] To achieve absorption as high as possible, metamaterial absorber usually consists of many functional layers, including the electric resonator, dielectric interlayer, and metal reflector layer. The electric resonator and the metal reflector layer will couple separately in order to absorb as much incident electromagnetic wave as possible at resonance frequency.^[8–14] Some research indicates that the absorption magnitude can be maximized when the impedance of the metamaterial absorber matches well to that of the free space air layer according to the effective mediums theory:^[7,15]

In this paper, a simple structure metamaterial absorber is designed and fabricated, which consists of a circular-holes structured electric resonator, an MgF₂ dielectric layer, and continuous metal films. The perfect absorbing property of metamaterial absorber is studied by the effective medium theory and the computed currents. The metamaterial absorber shows high absorption rates in a wide incident angle range (0 –60 ) in both transverse electric (TE) and transverse magnetic (TM) cases. Moreover, an absorption sensor scheme is suggested based on combining variation of the surrounding medium and the concept of the perfect absorber. The measured results show that the suggested absorption sensor scheme has a high sensitivity to the variation of the refractive indices of different surrounding media.

Figures 1(a) and 1(b) show the design of the proposed matematerial absorber, and figure 1(d) displays the optical images. On the one hand, the proposed metamaterial absorber is fabricated through circular hole arrays in a square lattice. First, a silicon wafer is selected as a temporary substrate, and then a 400-nm thick SU-8 layer will be spun on the temporary substrate. A 45-nm thick gold layer is evaporated at a rate of 2.2 Å s-1 at the -atm (1 atm = 1.01325 Pa) working pressure on the SU-8 layer as a bottom metal reflector by using the electron beam evaporation (Peva-600E). A 90-nm-thick MgF intermediate medium layer is deposited on the first gold layer. Then, another 45-nm-thick top gold layer is evaporated on the MgF₂ intermediate medium layer (also using Peva-600E). The fabricated gold/MgF₂/gold three-layers structured metamaterial (semi-finished products, not the final sample) is removed from the temporary Si substrate subsequently through adopting the buffered oxide etchant in order to etch the SU-8 layer away and liftoff the three-layers structured metamaterial. The gold/MgF₂/gold three-layers structured metamaterial is dried by using a C-MAG HP10 Hot plate (110 degrees Celsius, for 13 minutes). The purpose is to obtain an independent substrate absorber, which can be applied to a variety of sensor platforms conveniently. At the end, the proposed hole arrays are defined through adopting regular electron-beam lithography. The area of each sample is 2 mm mm. The measured absorption result is obtained through the Bruker Optics Equinox spectrometer 55 Fourier transform infrared spectrometer. The optical images of samples are obtained through using a Leica DM2700M. On the other hand, the proposed metamaterial absorber is computed numerically by adopting Ansofts HFSS 11.0. The bottom gold layer is thick enough, which is larger than the skin depth of the incident light and therefore forbids the transmitted light through the samples. Therefore, the absorption rate can be obtained in simulations and experiments as follows: . In simulations, the permittivity of the MgF layer is set to be 1.9,^[23] while the permittivity of bulk gold is set by the Drude model. In the Drude model, the damping constant is s , and the plasma frequency is s .^[24]

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (color online) (a) Top view of the unit cell on the XOY plane. (b) Side view of the unit cell on the XOZ plane. The orange part represents gold and the green part refers to MgF2. (c) The measured absorption spectra. (d) Optical photos of samples.

In this paper, the physical mechanism behind the absorption peak is explored by the simulation method. However, the difference between computed result and measured result naturally occurs, especially the damping constant of the Drude model plays an important role in the optical property of metamaterial. This is because the damping constant of the gold film is higher than that of the bulk gold due to the surface scattering and the grain boundary effect in thin metal film.^[25] It is important to optimize the damping constant of bulk gold in simulation, in order to obtain the most realistic computed results. To obtain a most reasonable damping constant in the targeted scope (150 THz–300 THz) in this paper, simulated absorption peaks as a function of damping constants are shown in Fig. 2. It can be found that different amplitude absorption peaks are achieved with different damping constants. For the 1.0 time damping constant case (the black curve in Fig. 2), an absorption peak close to 95% at 212 THz is obtained. However, for the 1.06 times damping constant case (the red curve in Fig. 2), a maximum absorption rate close to 99.6% at 212 THz is obtained. This simulated result indicates that the damping constant equal to 1.06 times that in Ref. [24] can yield a high absorption peak resonance. Therefore, a simple structure designed perfect absorber is achieved. For 1.2 and 1.5 times cases (the green and blue curves in Fig. 2), the maximum absorbance rates are reduced to 87% and 65%, respectively. For different damping constants, the resonance frequencies are almost unchanged, while a maximum absorption rate can be obtained. It suggests that the optimization damping constant is important in simulation. Comparing these results of Figs. 1(c) and 2, it can be found that the computed result which is obtained under the 1.06 times damping constant is well consistent with the measured result in the 150-THz 300-THz targeted spectrum. In the next results, the damping constants are all 1.06 times damping constants.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. (color online) Computed absorption peaks with different times damping constants of bulk gold.

Table 1.

Table 1.

Table 1. Dimension parameters of the proposed absorber. . Parameter P R H h Value/nm 900 250 90 45

Table 1. Dimension parameters of the proposed absorber. .

The perfect simulated absorption peak in Fig. 2 (the red curve) indicates that the perfect impedance matching condition is achieved at the resonant frequency. Effective parameters (the effective permittivity and permeability) are extracted by using S parameters retrieval methods^[26,27] as shown in Fig. 3. The real part of μ decreases from 18 to in the absorption band, and crosses zero at THz. At the same time, the real part of permittivity increases from to 23 in the absorption band, and also crosses zero at the same resonant frequency. These computed results indicate the impedance match between the absorber and free air interface, which is similar to a previous result.^[28]

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (color online) (a) Imaginary parts of the effective parameters. (b) Real parts of effective parameters of the proposed metamaterial absorber.

In practice, the optical properties of the proposed absorber can be understood through the computed currents. Figure 4 shows the computed currents of the proposed absorber on the top and bottom gold layers at THz. On the top gold layer, many parallel currents towards the right are excited. At the same time, many anti-parallel currents are also excited on the bottom gold layer. These resonant currents distribute on the bottom and top gold layers, which constitute a circulating current system. The circulating current system then excites a magnetic resonance field between the top and bottom gold layers at 212 THz, which interacts and couples with the incident light magnetic field.^[29–31] The interacting and coupling phenomenon confines the incident light energy in the intermediate dielectric layer, which results in the absorption peak at 212 THz.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. (color online) Computed current distributions at 212 THz on (a) the top gold layer and (b) the bottom gold layer.

In this paper, the proposed metamaterials absorber is applied to various sensing platforms, such as photovoltaic cells, thermal emitters and detectors.^[32,33] It is important to absorb as much of the incident light as possible when the proposed metamaterial absorber is used. Figure 5 shows the absorption spectra for both TM and TE configurations at various incident angles. For the case of the TM polarization, the maximum absorption rate is nearly unchanged (from 99.6% to 97.3%) with incident angle increasing. As discussed above, the direction of the incident resonance magnetic field is almost unchanged when the incident angle is increased, which will drive these circulating currents effectively at a large incidence angle as shown in Fig. 5(a). For the TE polarization case, resonant frequencies of absorption peaks are almost unchanged, while the maximize maximum absorbance rate is reduced from 99.6% to 83%, as shown in Fig. 5(b). The abnormal decrease of the maximum absorption rate is due to the fact that the magnetic field cannot drive these circulating currents efficiently between two gold layers at a large incident angle.^[34]

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. (color online) Computed absorbance peaks for both TM and TE configurations.

In the following, the feasibility of the application to the sensor platform of the proposed metamaterial absorber will be discussed. The basic principle of the proposed absorption sensor is that the variation of the surrounding medium always plays an important role in the absorption rate. In practical applications, sensors always face different detection sensing environments. For instance, the surrounding medium on the sensor surface may change from air to water, or other media. When the sensing environment is changed, the refractive index of the surrounding medium on the sensor surface is thus changed, which causes the absorption rate to change. It is the opportunity for developing a sensor which is sensitive to detection and can sense the variation of the intensity in absorption when the refractive index of the surrounding medium is changed. Moreover, the proposed metamaterial absorber is independent of any substrate, which leads to the fact that the absorber can be applied to most sensor platforms.

To confirm the feasibility of the proposed absorption sensor above, three kinds of surrounding media (air, water, and glucose solution) are applied to the surface of the absorption sensor, and measured results are shown in Fig. 6(a). For the case of the air surrounding medium, the measured absorption peak reaches to 99.6%. When the air is replaced by the water surrounding medium, the measured absorption peak decreases to 97% (detection conditions are normal pressure and temperature). Finally, the water is replaced by the glucose solution surrounding medium (normal glucose solution for injection), and the measured absorption peak decreases to 93%. At the same time, the resonant frequency shows a shift toward low-frequency when the refractive index of the surrounding medium is increased. On the other hand, computed results are shown in Fig. 6(b). In these simulations, damping constants of gold layers are given as 1.06 times that of bulk gold. Moreover, the refractive index of surrounding media are given as , , , respectively. These computed results are well consistent with measured results.

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. (color online) (a) Measured and (b) computed absorption peaks with different surrounding media.

In order to evaluate the level of sensitivity on the variation of the refractive index of the surrounding medium, the figure of merit (FOM) is adopted to describe the performance of the absorption sensor in thefollowing. The FOM is given as

Fig. 7.

	Figure Option ViewDownloadNew Window
	Fig. 7. (color online) Computed FOM value versus frequency for different surrounding media.

In this work, a metamaterial absorber is designed, computed, and measured in a 150-THz 300-THz range. A perfect absorption peak is obtained at 212 THz. The measured absorber shows high absorption rates for the TM and TE polarizations at a large angle of incidence, for example 60 . Electromagnetic resonance properties of the measured absorber are analyzed through extracting effective parameters (the effective permittivity and permeability) and computing currents on the top and bottom gold layers at the resonance frequency. Moreover, the damping constant of bulk gold in simulation is optimized to obtain a most reasonable damping constant in the targeted scope (150 THz–300 THz) in this paper. Finally, an absorption sensor scheme is proposed through combining the measured absorber and the variation of surrounding media (including air, water, and glucose solution). Comparison among existing equipment shows that the proposed absorption sensor scheme is highly sensitive to variation of the surrounding media.