Figure 1 presents the schematic diagram of the filter by embedding hybrid cavity between dual DBRs. The top DBR (DBR1) consists of M layers of GaN/SiO2 pairs (with thicknesses d1 = 91 nm and d2 = 67 nm respectively, M = 2). The bottom DBR (DBR2) consists of N layers of Si3N4/SiO2 pairs is set (with thicknesses d3 = 70 nm and d4 = 40 nm respectively, N = 3).
The refractive index of SiO2 is considered as n0 = 1.47, and the optical properties of Si3N4 and GaN are described by the following dispersion formulas, respectively [25, 26]:
$${\text{n}}_{\text{Si}\text{3}\text{N}\text{4}}\text{=}\sqrt{\frac{\text{2.8939}{\text{λ}}^{\text{2}}}{{\text{λ}}^{\text{2}}\text{-}{\text{0.13967}}^{\text{2}}}\text{+1}}$$
1
$${\text{n}}_{\text{GaN}}\text{=}\sqrt{\frac{\text{1.75}{\text{λ}}^{\text{2}}}{{\text{λ}}^{\text{2}}\text{-}{\text{0.256}}^{\text{2}}}\text{+}\frac{\text{4.1}{\text{λ}}^{\text{2}}}{{\text{λ}}^{\text{2}}\text{-}{\text{17.86}}^{\text{2}}}\text{+3.6}}$$
2
\(\text{λ}\) is the considering wavelength.
Ag layer is embedded in the middle with thickness of dAg=30 nm. This design can simulate the TPP mode effectively. The dielectric constant of Ag is expressed by Drude model [27]:
$${\text{ε}}_{\text{r}}\text{=}{\text{ε}}_{\text{∞}}\text{- }\frac{{\text{ω}}_{\text{p}}^{\text{2}}}{{\text{ω}}^{\text{2}}\text{+}\text{iωγ}}\text{-}\frac{\text{Δ}{\text{×}\text{Ω}}^{\text{2}}}{\left({\text{ω}}^{\text{2}}\text{-}{\text{Ω}}^{\text{2}}\right)\text{+}\text{iΓ}}$$
3
The related parameters are listed in Table 1.
Table 1
Parameter values for the Drude model.
Parameter | Value |
ε∞ | \(\text{2.4064}\) |
ωp | 2π⋅2214.6⋅1012 Hz |
γ | 2π⋅4.8⋅1012 Hz |
Δ | 1.6604 |
Ω | 2π⋅620.7⋅1012 Hz |
Γ | 2π⋅1330.1⋅1012 Hz |
Thickness of LayerA and LayerB is changeable to maintain different color through adjusting the designed central wavelengths of the dual DBRs. There is a great need on the balance between color intensity and purity influenced by the bandwidth for color filter when designing the structure. Thickness of each layer and the number of pairs in DBRs can be optimized so that high-intensity transmission with proper bandwidth can be obtained.
To calculate the optical properties of the structure, the simulation in this paper uses the transfer matrix method (TMM) due to simplicity and flexibility of the method. The transfer matrix \(\text{M}\) is constructed firstly with the individual interface matrices \({\text{I}}_{\text{j}}\) and propagation \({\text{P}}_{\text{j}}\), the index \(\text{j}\) is identifier of the interface in discussion [28]
$${\text{r}}_{\text{j}}\text{=}\frac{{\text{μ}}_{\text{j+1}}{\text{n}}_{\text{j}}\text{-}{\text{μ}}_{\text{j}}{\text{n}}_{\text{j+1}}}{{\text{μ}}_{\text{j+1}}{\text{n}}_{\text{j}}\text{+}{\text{μ}}_{\text{j}}{\text{n}}_{\text{j+1}}}$$
4
$${\text{τ}}_{\text{j}}\text{=}\frac{\text{2}{\text{μ}}_{\text{j+1}}{\text{n}}_{\text{j}}}{{\text{μ}}_{\text{j+1}}{\text{n}}_{\text{j}}\text{+}{\text{μ}}_{\text{j}}{\text{n}}_{\text{j+1}}}$$
5
$${\text{I}}_{\text{j}}\text{=}\frac{\text{1}}{{\text{τ}}_{\text{j}}}\left(\begin{array}{cc}\text{1}& {\text{r}}_{\text{j}}\\ {\text{r}}_{\text{j}}& \text{1}\end{array}\right)$$
6
$${\text{P}}_{\text{j}}\text{=}\left(\begin{array}{cc}{\text{e}}^{{\text{-}\text{ϕ}}_{\text{j}}\text{i}}& \text{0}\\ \text{0}& {\text{e}}^{{\text{ϕ}}_{\text{j}}\text{i}}\end{array}\right)$$
7
where \({\text{r}}_{\text{j}}\) and \({\text{τ}}_{\text{j}}\) are the reflection and transmission coefficient, respectively. \({\text{μ}}_{\text{j}}\) and \({\text{n}}_{\text{j}}\) are the (relative) magnetic permeability and refraction of index respectively. \({\text{ϕ}}_{\text{j}}\) is the layer phase thickness corresponding to the phase change of incident light from top of the structure when traverses layer j. Fields at incidence and substrate side can be described by the transfer matrix M,
$$\text{M}\text{=}\left(\begin{array}{cc}{\text{M}}_{\text{11}}& {\text{M}}_{\text{12}}\\ {\text{M}}_{\text{21}}& {\text{M}}_{\text{22}}\end{array}\right)\text{=}\prod _{\text{j=1}}^{\text{L}}{\text{I}}_{\text{j}}{\text{P}}_{\text{j}}{\text{I}}_{\text{L+1}}$$
8
$$\left(\begin{array}{c}{\text{E}}_{\text{0}}^{\text{-}}\\ {\text{E}}_{\text{0}}^{\text{+}}\end{array}\right)\text{=}\text{M}\left(\begin{array}{c}{\text{E}}_{\text{L+1}}^{\text{-}}\\ {\text{E}}_{\text{L+1}}^{\text{+}}\end{array}\right)$$
9
Along the z axis, the incident electric field is \({\text{E}}_{\text{0}}^{\text{-}}\), the reflected field is \({\text{E}}_{\text{0}}^{\text{+}}\), the transmitted field is \({\text{E}}_{\text{L+1}}^{\text{-}}\). In the demonstration, transmission is dependent on the polarization of incident light and incident angle. The initial condition for the calculation is under normal incidence with TE polarization before discussing the influence of the incident angle and polarization.