Figure 1(d) displays the absorption spectra in the band from 0.9 µm to 4 µm for normal incidence electromagnetic waves. It can be noted that under the optimized structure parameters, which were introduced in the previous section, we obtained four maximum absorption peaks around the wavelengths of 1064 nm, 1550 nm, 2080 nm, and 3000 nm with absorption values of 99.9%, 99.9%, 99.1%, and 99.9%, respectively. These values of absorptivity indicate perfect absorption at the four wavelengths.
To explain the nature of these four absorption peaks, absorption spectra are simulated for the sequential steps of construction growth direction (i.e., the effect of adding each Au ring to the structure). This process is depicted in Fig. 2. Figure 2(a) introduces the absorption spectra for the first Au ring, which has a radius of \(Rg1\). It can be noted that the structure with this ring has a single absorption peak at a wavelength of 2965.5 nm with an absorptivity of 97.3%. As shown in Fig. 2(b), after adding the second ring, which has a radius of \(Rg2\), another absorption peak appears at a wavelength of 1982 nm with an absorptivity of 94%. Also, it can be noted that the addition of the second ring causes a red shift to the first peak from 2965.5 nm to 3000 nm and enhance the value of its absorption to be equal to 99.9%. Figure 2(c) illustrates the effect of adding the third ring, which has a radius of \(Rg2\), on the absorption spectra. Another absorption peak appears at a wavelength of 1428 nm with an absorptivity of 90.85%. Also, the addition of the third ring leads to a red shift to the second peak to a wavelength around 2080 nm with enhanced absorptivity to become 99.1% while it does not affect the first peak. Figure 2(d) shows the effect of adding the fourth ring to the structure. This addition results in the appearance of the fourth peak at the wavelength of 1010 nm with an absorptivity of 96.5% and a red shift to the third peak to a wavelength around 1550 nm with an absorptivity of 99.9% while it does not affect the first and second peaks. Finally, to adjust the fourth peak to be at a wavelength of 1064 nm with absorptivity of 99.9%, a golden disk of radius \(Rg5\) is added on the top of the fourth ring as shown in Fig. 2(e). From the description of Fig. 2, we can conclude that adding a new ring adds a new resonance and adjust the previous resonance to a desired wavelength with perfect absorbability.
For a deep explanation of the concept of perfect absorption at the designed four wavelengths, we introduce in Fig. 3 the real and imaginary components of the relative input impedance which are retrieved from the S-parameters by using the following equation
$$Z\left(\omega \right)=\sqrt{\frac{{\left(1+{S}_{11}\right)}^{2}-{S}_{21}^{2}}{{\left(1-{S}_{11}\right)}^{2}-{S}_{21}^{2}}}$$
1
It is clear from this figure, the real component of the relative impedance at the four resonances is approximately equal to one (i.e., Re (z) ≈ 1), while the imaginary component is near zero (i.e., Im (z) ≈ 0). These results indicate that the impedance of the QPMA is in a good matching with the free space characteristic impedance (i.e., \({Z}_{0}=120\pi\) Ω) at the four resonance wavelengths, hence, no reflection occurs and perfect absorption is achieved.
To investigate the concept of absorption of QPMA, we introduce the absolute value of the electric field, \(\left|E\right|\), and the magnetic field, \(\left|H\right|\), distributions of the absorber, at (a) \({\lambda }_{1}\)=3000 nm, (b) \({\lambda }_{2}\)=2080 nm, (c) \({\lambda }_{3}\)=1550 nm, and (d) \({\lambda }_{4}\)=1064 nm, in Figs. 4 and 5, respectively. Figure 4 shows the y–z plane cross-section view for \(\left|E\right|\) field distribution while Fig. 5 shows x–z plane cross-section view for \(\left|H\right|\) field distribution. It is clear from Fig. 4(a) and Fig. 5(a), corresponding to the wavelength of 3000 nm, that \(\left|E\right|\) field and \(\left|H\right|\) field distributions are mainly concentrated inside the first dielectric disk, D1, between the rings M1 and M2, with some little distribution inside the substrate. The distribution of the electric field between these rings, M1 and M2, indicates an equivalent capacitance while the magnetic distribution indicates an inductance behavior of this region. The interaction between the capacitance and inductance results in resonance at the first band i.e., 3000 nm. This behavior confirms the result shown in Fig. 2(b) in which the resonance at 3000 nm comes mainly due to the ring M1 with fine tuning from ring M2. Similarly, at a wavelength of 2080 nm, it can be noted from Fig. 4(b) and Fig. 5(b) that \(\left|E\right|\) field and \(\left|H\right|\) field distributions are mainly concentrated at the dielectric D2 between the ring M2 and M3, which results in a peak absorption at that frequency. Also, the same behavior carries out between rings M3 and M4 at the wavelength of 1550 nm as indicated in Fig. 4(c) and Fig. 5(c). Finally, for the peak absorption at a wavelength of 1064 nm, Fig. 4(d) and Fig. 5(d) shows the \(\left|E\right|\) field and \(\left|H\right|\) field distributions are mainly concentrated inside the top dielectric disk, D4, between the ring M4 and the top metal disk.
In the following subsections, the effects of the variation in polarization angle and incident angle on the performance of the proposed QPMA structure are studied.
3.1. Effect of polarization angle variation
The polarization insensitivity of the proposed QPMA for TE and TM modes is studied. Either TE or TM waves are incident vertically on the proposed absorber surface. The effect of varying the polarization angle of the TE or TM waves on the absorption of the proposed absorber is presented in Fig. 6. The obtained results reveal that when the polarization angle varies, the peaks of the absorption for the quad bands remain unchanged for both TE and TM modes. Thus, the proposed QPMA is insensitive to polarization angles.
3.2. Effect of incident angle variation
In practical applications, the effect of large-angle absorption plays a significant role in designating the performance of the absorber. In this subsection, we address the effect of incident angle for both TE and TM modes. Owing to the high sensitivity of the proposed absorber at 1064 nm wavelength to the variation of incident angle, as observed from simulation results, the results of this effect for TE and TM modes are introduced in Fig. 7(a) and (b) for 1064 nm wavelength, only, while Fig. 8(a) and (b) present the counterpart results for the remaining triple bands.
At θ = 0 (i.e., normal incidence), due to the symmetrical geometry of the proposed absorber, it can be noted that for both TE and TM, the proposed absorber can trap the incident infrared waves regardless of their polarization.
For the 1064 nm wavelength, Fig. 7(a) shows the absorption spectra in the band from 0.9 µm to 1.25 µm for TE mode. As can be observed, as the incident angle increases the peak absorption around 1064 nm tends to make redshift up to 1100 nm at θ = 50o then the absorption peak drops below 90% at a wavelength of 1120 nm and incident angle of θ = 60o. Also, other absorption peaks (harmonics) are observed at incident angles greater than θ = 30o. These harmonics, also, tend to make redshift with the increase of incident angle. The proposed structure can perform an absorption peak of 90% around 1064 nm at an incident angle of θ = 40o at TE mode. On the other hand, as indicated in Fig. 7(b) for the TM mode, peak absorption around 1064 nm tends to make a redshift with a continuously decreasing in its value. In addition, other absorption peaks can be observed at incident angles greater than θ = 10o which, also, tend to make redshift with the increase of incident angle. The proposed structure can perform an absorption peak of 85% around 1064 nm at an incident angle of θ = 30o at TM mode.
Furthermore, we investigate the absorber performances for the triple bands, 1550 nm, 2080 nm, and 3000 nm under several incident angles from 0° to 70° for both TE and TM modes are illustrated in Fig. 8(a) and (b), respectively. As shown in Fig. 8(a), in the case of TE mode, it can be noted that as the incident angle increases the absorption peak values decrease for all bands. Also, with the increase of incident angle value, the absorption peaks around the 3000 nm band tend to make a blue shift while the absorption peaks around the 1550 nm band tend to make a redshift. The proposed absorber can perform absorption greater than 90% for all three bands under consideration at incident angles up to 50o. For incident angles larger than 50°, the absorption intensity of the triple-band peaks degrades rapidly. This is because the magnetic field component in the x-y plane is \({H}_{x-y}=Hcos\theta\) which is required to excite the magnetic polaritons (Bai et al. 2015; Ye et al. 2010), where \(H\) is the incident magnetic field intensity. Because the intensity drops rapidly at θ > 50°, the magnetic resonance becomes weaker, hence, we observed lower absorption. On the other hand, in the case of TM mode, shown in Fig. 8(b), the triple absorption peaks intensity remains above 99%. This is due to the stability in the magnetic incident field component in the direction of magnetic polaritons oscillations. Also, it can be noted the absorption peaks around the center wavelengths of the triple bands tend to make, only, a blue shift with an insignificant decrease in its peak value. Yet, due to the blue shift from the center wavelengths, there is a reduction in the absorption values at the center wavelengths of the triple bands. The proposed absorber can perform absorption greater than 90% at the center wavelength of all three bands under consideration at incident angles up to 50o. The slight blue shift of the three absorption peaks arises, mainly, due to the off-phase oscillating effect (Ye et al. 2010).