1. Introduction

Materials that can absorb high-intensity light are highly desirable for many applications, including photonic radiation detectors, absorbers, efficient solar cells, and color filters [1–3]. In recent years, enhanced and complete optical absorption based on electromagnetic surface states has attracted considerable attention [4]. There are two typical electromagnetic surface states. The first is the conventional surface plasmon polariton (SPP), which can be excited by transverse magnetic (TM) waves using grating or prism structures [5,6]. The second is the optical Tamm state (OTS) [7], which is also called the Tamm plasmon polariton (TPP) [8,9], and it can be excited around the interface between a thick metal film and a truncated all-dielectric photonic crystal (PC). The electromagnetic (EM) fields of OTSs are highly localized near the interface between the metal film and the PC. Unlike conventional SPPs, OTSs can be excited directly within the light core and also in both transverse electric (TE) and TM polarizations [7–10]. These states have been discussed from the perspective of enhanced optical transmission [7]. Additionally, if zero reflection conditions at the entrance face are satisfied, the light of OTSs can be absorbed perfectly [11]. Such absorption behavior is desirable in many applications, including high-sensitivity sensors, optical switches [12,13] and high-efficiency lasers [14–17]. Multiple OTSs have been investigated theoretically in sandwich structures composed of a metal film, a top dielectric layer and a finite all-dielectric PC [18]. It was found that multiple OTSs could be realized by tuning of the top layer thickness. To date, many researchers have focused on these theoretical multiple optical absorption phenomena [18–24]. In this paper, we experimentally investigated multiple perfect optical absorption phenomena in the visible region in sandwich structures containing a thin metal film, a top dielectric layer and a truncated dielectric PC.

The remainder of this paper is organized as follows. In Section 2, we introduce the proposed sandwich structures and their experimental preparation, and then discuss the case of perfect absorption with a single peak. Subsequently, in Section 3, by varying the thickness of the top layer, we experimentally realize double and triple perfect absorptions. Finally, we draw our conclusions in Section 4.

2. Single near-perfect absorption in the sandwich structure

The sandwich structure is denoted by MP(LH)^NLS, as shown in Fig. 1, where M represents the thin metallic film whose thickness is larger than the skin depth, P denotes the top dielectric layer, and (LH)^NLS represents a truncated all-dielectric PC with periodic number N. The selected metal M is Cu, which has the refractive index given in [25]. L and H represent SiO₂ and TiO₂ layers with refractive indices of n_L and n_H and thicknesses of d_L and d_H, respectively. The top layer P is composed of TiO₂ and has the thickness d_P. S represents the substrate, which is BK7 glass and has a refractive index of n_S = 1.52. The sandwich structure is prepared using two processes. First, the P(LH)^NL structures are evaporated on the substrate by ion-assisted electron-beam evaporation under high vacuum conditions. The SiO₂ and TiO₂ layer thicknesses were monitored using a quartz crystal sensor. Second, the metal layer with target purity of 99.999% is deposited using the radio-frequency (RF) ion beam sputtering method. The thickness of Cu film is controlled using the deposition time based on the sputtering rate of the RF ion beam. The refractive indices of the SiO₂ and TiO₂ layers were calculated based on the measured transmission spectra of monolayer films of these materials [26] and values of n_L = 1.431 and n_H = 2.123, respectively, were obtained. In this paper, normal incidence conditions are considered, the periodic number N is 10, and the Cu layer thickness d_M is 30 nm. The transmission (reflection) spectra were measured using a Cary-100 ultraviolet-visible-near-infrared spectrophotometer. The absorptance can be calculated using the formula A = 1−T−R, where R, T and A are the reflectance, the transmittance and the absorptance, respectively.

 

Fig. 1 Schematic of the sandwich structure MP(LH)^NLS with a top dielectric layer between a metal film and a truncated PC, where M denotes the metal Cu, P and H represent TiO₂ with the refractive index n_H = 2.123, L is SiO₂ with n_L = 1.431, S denotes the BK7 substrate with n_S = 1.52. The thickness of M is d_M = 30 nm and the periodic number N is 10.

Download Full Size | PDF

For the MP(LH)¹⁰LS sandwich structure, we initially set d_H = 83.9 nm, d_L = 116.7 nm, and d_P = 61.2 nm. At normal incidence from the left side of the sandwich structure, Figs. 2(a)-2(c) show the theoretical and measured spectra of the reflectance, the transmittance and the absorptance, respectively. The red lines represent theoretical results obtained using the transfer-matrix method [27] and the black lines give experimental values. In Fig. 2(c), the red line shows that one near-perfect absorption peak exists at 712.2 nm and the maximum absorptance in the experiment appears at 718.0 nm which is located in the forbidden band gap (602.2–803.5 nm) of the truncated PC. The results of Fig. 2 also show that the experimental spectra agree well with simulated values. The small discrepancies between theoretical and experimental spectra mainly come from two reasons. First, the bandwidth of the quasi-monochromatic light in the spectrophotometer can broaden the bandwidth of the absorption peak. Second, the differences originate from errors in layer thickness monitoring during the deposition process and refractive index discrepancies between the theoretical data and the experimental materials.

 

Fig. 2 (a) Reflectance (R), (b) transmittance (T), and (c) absorptance (A) of MP(LH)¹⁰LS for the light incidence from the left side. Red lines (black lines) represent simulated (experimental) values. The thicknesses of P, L and H are d_P = 61.2 nm, d_L = 116.7 nm, and d_H = 83.9 nm, respectively. All other parameters are the same as those in Fig. 1.

Download Full Size | PDF

To aid in revealing the physical origin of this perfect absorption phenomenon, the simulated intensity distributions of the electric field (${\left|\text{E}\right|}^2$) and magnetic field (${\left|\text{H}\right|}^2$) at the absorption wavelength 712.2 nm are shown in Fig. 3. We supposed that the intensity of the incident electric field was 1 and that the light was incident from the left side. The red and blue lines denote ${\left|\text{E}\right|}^2$ and ${\left|\text{H}\right|}^2$, respectively. The maximum values of ${\left|\text{E}\right|}^2$ and ${\left|\text{H}\right|}^2$ are localized near the interface between the metal layer and the top layer, which indicates that this type of the localized mode originates from the OTS [8,9,11]. Because an OTS leads to strong localization of ${\left|\text{E}\right|}^2$ and ${\left|\text{H}\right|}^2$ in the metal film at the tunneling wavelength, near-perfect absorption is realized.

 

Fig. 3 Simulated intensities of electric fields (red line) and magnetic fields (blue line) in the MP(LH)¹⁰LS structure at the absorption wavelength of 712.2 nm for the light incidence from the left side. All other parameters are the same as those used in Fig. 2.

Download Full Size | PDF

3. Multiple near-perfect absorptions in sandwich structures

While maintaining an invariant Cu layer thickness, we increased the thickness of d_P to 805.5 nm, and tuned the values of d_H = 86.1 nm and d_L = 119.8 nm. Figure 4 presents the reflectance, transmittance and absorptance for both the simulated (red lines) and experimental (black lines) results for the light incidence from the left side. In contrast to the previous single absorption peak, there are double near-perfect absorption peaks. In Fig. 4(c), the red line shows the theoretical maximum absorptances appear at the wavelengths of 655.7 nm and 770.3 nm, while the black line indicates measured maximum absorptances at 659.0 nm and 763.0 nm. The wavelengths of these peaks are also located in the forbidden band gap (618.7–830.1 nm) of the truncated PC. Though the measured data deviate from the theoretical results, the absorption trend remains basically consistent with that of the theory. Similar to the earlier analysis, the EM field intensity distributions are depicted at two near-perfect absorption wavelengths of 655.7 nm and 770.3 nm for the sandwich structure while again assuming normal incidence from the left side. The simulated results are shown in Fig. 5. Interestingly, the maximum values of ${\left|\text{E}\right|}^2$ and ${\left|\text{H}\right|}^2$ are clearly localized in the top layer and not at the interface between the metal layer and the top layer. Therefore, it was proposed that the physical mechanism of these double perfect absorptions mainly originated from Fabry-Perot (F-P) resonant cavity modes rather than from the OTSs. In terms of the complete structure, the top layer can be regarded as the F-P cavity and the photonic barriers were formed by the metal film and the truncated dielectric PC. These types of resonant cavity modes still lead to enhancement of the EM field intensities in the metal layer and thus the double absorption peaks were produced.

 

Fig. 4 (a) R, (b) T, and (c) A of the MP(LH)¹⁰LS structure for the light incidence from the left side. The thicknesses of P, L and H are d_P = 805.5 nm, d_L = 119.8 nm, and d_H = 86.1 nm, respectively. All other parameters are the same as those used in Fig. 1.

Download Full Size | PDF

 

Fig. 5 Simulated intensities of electric fields (red lines) and magnetic fields (blue lines) in the MP(LH)¹⁰LS structure at the double absorption wavelengths of (a) 655.7 nm and (b) 770.3 nm for the light incidence from the left side. All parameters are the same as those used in Fig. 4.

Download Full Size | PDF

Next, to obtain multiple perfect absorptions, keeping the thickness of the metallic layer M constant, we continued to increase the top layer thickness with d_P = 1371.6 nm, and set d_H = 77.6 nm and d_L = 103.4 nm. In the case of normal incidence from the left side of the sandwich structures, simulated (red lines) and experimental (black lines) results for reflectance (R_Left) and T are shown in Figs. 6(a) and 6(b), respectively, and Fig. 6(c) shows the corresponding absorptance (A_Left). In Fig. 6(c), the simulated values (red line) show that there are three near-perfect absorption peaks corresponding to the wavelengths of 573.1 nm, 626.9 nm and 691.4 nm in the forbidden band gap (544.45–734.62 nm) of the truncated PC, while the experimental maximum absorptances appear at 576.0 nm, 628.0nm and 691.0 nm. The measured results are in good agreement with the theoretical values. To study the absorption sensitivity to the direction of the light incidence, we also measured the absorptance (A_Right) for the light incidence from the right side of the structure, and the experimental results were shown in Fig. 6(d). Apparently, the absorptance is almost zero in the wide wavelength range of 550–720 nm in this case.

 

Fig. 6 (a) Reflectance (R_left), (b) transmittance (T), and (c) absorptance (A_Left), and (d) absorptance (A_Right) of the structure MP(LH)¹⁰LS, where the subscript Left (Right) represents the light incidence from the left (right) side. Red lines (black lines) represent simulated (experimental) results. The thicknesses of M, P, L and H are d_M = 30 nm, d_P = 1371.6 nm, d_L = 103.4 nm, and d_H = 77.6 nm, respectively. All other parameters are the same as those used in Fig. 1.

Download Full Size | PDF

Similarly, the EM field intensities at the three perfect absorption wavelengths of 573.1 nm, 626.9 nm and 691.4 nm were simulated in detail. Figures 7(a)-7(c) show the results for the case of light incidence from the left side. The maximum values of ${\left|\text{E}\right|}^2$ and ${\left|\text{H}\right|}^2$ occur in the top layer rather than at the interface between the metal layer and the top layer, which is similar to the results shown in Fig. 5. Therefore, three perfect absorption peaks are mainly realized because of the F-P resonances, and the top layer thus acts as the F-P cavity. However, there are still strong EM field intensities in the metal layer, which means that the structures can absorb the incident light almost completely at three different resonant wavelengths. In contrast, for the light incidence from the right side of the structure, Figs. 7(d)-7(f) show that weak localized EM fields occur in the structure at three resonant wavelengths. Therefore, the structures studied here have nonreciprocal absorption and reflection properties and could thus be used to fabricate multi-channel optical absorbers.

 

Fig. 7 Simulated intensities of electric (red lines) and magnetic (blue lines) fields in the MP(LH)¹⁰LS structure at three absorption wavelengths of (a) and (d) 573.1 nm, (b) and (e) 626.9 nm, and (c) and (f) 691.4 nm, where (a)-(c) corresponds to the light incidence from the left side and (d)-(e) corresponds to the light incidence from the right side. All parameters are the same as those used in Fig. 6.

Download Full Size | PDF

4. Conclusions

We have theoretically and experimentally demonstrated that single and multiple instances of near-perfect absorption can be achieved using sandwich structures composed of a metallic film, a top layer and truncated PCs. The physical mechanisms that lead to the realization of perfect absorption are analyzed in detail. It is found that OTSs play the main role in instances of single perfect absorption, while F-P resonances are more important for instances of multiple complete absorptions. These structures may be helpful in the design of new types of optical and photonic devices.