利用传输矩阵法求解布拉格光栅的透射谱

利用传输矩阵法求解布拉格光栅的透射谱

采用传输矩阵法（TMM）计算具有任意折射率分布光栅结构的透射谱，TMM法描述如下：

能够计算折射率呈阶梯状分布的波导的反射和透射率，以及波导的传播常数。

在单模波导中，计算反射和透射率采用2×2的矩阵表示。

为了表示光栅（多个折射率突变界面），将矩阵乘成级联网络，

能够计算光栅针对每个波长的透射值和反射值。

1、均匀波导的传输矩阵

​ 传输矩阵定义如下：
$$\left[\begin{array}{c}{A}_1\\ B_1\end{array}\right]=\left[\begin{array}{cc}{T}_{11}& T_{12}\\ T_{11}& T_{12}\end{array}\right]\left[\begin{array}{c}{A}_2\\ B_2\end{array}\right]$$

传递矩阵的概念类似于散射参数矩阵，波导的均匀截面如图（a）所示的传输矩阵如下：
$$T_{hw}=\left[\begin{array}{cc}{e}^{j\beta L}& 0\\ 0& e^{-j\beta L}\end{array}\right]$$
其中，β为场的复传播常数，包括折射率和传播损耗：
$$\beta =\frac{2\pi n_{eff}}{\lambda }-i\frac{\alpha }{2}$$
计算均匀波导的传输矩阵法其MATLAB代码如下：

function T_hw=TMM_HomoWG_Matrix(wavelength,l,neff,loss) % Calculate the transfer matrix of a homogeneous waveguide. % Complex propagation constant beta=2*pi*neff./wavelength-1i*loss/2; T_hw=zeros(2,2,length(neff)); T_hw(1,1,:)=exp(1i*beta*l); T_hw(2,2,:)=exp(-1i*beta*l);

2、折射率呈阶梯状分布的波导

折射率呈阶梯状分布的波导的传递矩阵，如图b所示，为
$$T_{is-12}=\left[\begin{array}{cc}1/t& r/t\\ r/t& 1/t\end{array}\right]=\left[\begin{array}{cc}\frac{n_1+n_2}{2\sqrt{n_1{n}_2}}& \frac{n_1-n_2}{2\sqrt{n_1{n}_2}}\\ \frac{n_1-n_2}{2\sqrt{n_1{n}_2}}& \frac{n_1+n_2}{2\sqrt{n_1{n}_2}}\end{array}\right]$$
其中r和t是基于菲涅耳系数的反射率和透射率。计算折射率呈阶梯状分布的波导界面的传输矩阵的MATLAB代码如下：

function T_is=TMM_IndexStep_Matrix(n1,n2) % Calculate the transfer matrix for a index step from n1 to n2. T_is=zeros(2,2,length(n1)); a=(n1+n2)./(2*sqrt(n1.*n2)); b=(n1-n2)./(2*sqrt(n1.*n2)); %T_is=[a b; b a]; T_is(1,1,:)=a; T_is(1,2,:)=b; T_is(2,1,:)=b; T_is(2,2,:)=a;

3、布拉格光栅及反射和透射率

表述布拉格光栅的级联矩阵如下：
$$T_p=T_{tw-2}{T}_{is-21}{T}_{hw-1}{T}_{is-12}$$
其中下标1和2表示低和高折射率的区域。然后构造有N个周期的均匀布拉格光栅：
$$T_p=(T_{tw-2}{T}_{is-21}{T}_{hw-1}{T}_{is-12}{)}^N$$
考虑了相移均匀布拉格光栅，即带有两个布拉格光栅反射器的一级FP腔，传递矩阵如下：
$$T=[(T_p{)}^N]T_{hw-2}[(T_p{)}^N]T_{hw-2}$$
计算由波导截面和材料界面组成的波导布拉格光栅腔的传输矩阵MATLAB代码如下：

function T=TMM_Grating_Matrix(wavelength, Period, NG, n1, n2, loss) % Calculate the total transfer matrix of the gratings l=Period/2; T_hw1=TMM_HomoWG_Matrix(wavelength,l,n1,loss); T_is12=TMM_IndexStep_Matrix(n1,n2); T_hw2=TMM_HomoWG_Matrix(wavelength,l,n2,loss); T_is21=TMM_IndexStep_Matrix(n2,n1); q=length(wavelength); Tp=zeros(2,2,q); T=Tp; for i=1:length(wavelength)Tp(:,:,i)=T_hw2(:,:,i)*T_is21(:,:,i)*T_hw1(:,:,i)*T_is12(:,:,i);T(:,:,i)=Tp(:,:,i)^NG; % 1st order uniform Bragg grating% for an FP cavity, 1st order cavity, insert a high index region, n2.T(:,:,i)=Tp(:,:,i)^NG * (T_hw2(:,:,i))^1 * Tp(:,:,i)^NG * T_hw2(:,:,i); end

我们将光栅以折射率为n2的部分作为开始和结束。相移区域是采用高折射率材料n2来实现的。最后，生成透射T和反射R谱。通过对波长点的一维矩阵进行了计算。计算光栅的反射和透射率MATLAB代码如下：

function [R,T]=TMM_Grating_RT(wavelength, Period, NG, n1, n2, loss) %Calculate the R and T versus wavelength M=TMM_Grating_Matrix(wavelength, Period, NG, n1, n2, loss); q=length(wavelength); T=abs(ones(q,1)./squeeze(M(1,1,:))).^2; R=abs(squeeze(M(2,1,:))./squeeze(M(1,1,:))).^2;

4、光栅物理结构设计

接下来将物理结构（波导几何形状）与有效折射率联系起来。输出波导部分为500×220 nm的条形波导和氧化物包层组成，其中波导宽度发生变化构成了光栅。

使用本征模计算光栅段的有效折射率，通过计算了波导与波长和波导宽度之间的有效折射率，然后对其进行参数化。数据可以通过曲线拟合为两个函数：
$$\begin{gathered} n_{eff-\lambda }(\lambda )=a_0-a_1(\lambda -{\lambda }_0)-a_2(\lambda -{\lambda }_0{)}^2 \\ {n}_{eff-w}(w)=a_0-a_1(w-w_0)-a_2(w-w_0{)}^2 \end{gathered}$$
其中，lambda的单位值微米，w为µm中的波导的宽度，neff (w)为给定波导宽度w下的有效折射率相对于其在λ0处的值的偏差。

定义光栅的物理参数，并画出频谱的MATLAB代码：

function Grating %This file is used to plot the reflection/transmission spectrum. % Grating Parameters Period=310e-9; % Bragg period NG=200; % Number of grating periods L=NG*Period; % Grating length width0=0.5; % mean waveguide width dwidth=0.01; % +/- waveguide width width1=width0 - dwidth; width2=width0 + dwidth; loss_dBcm=3; % waveguide loss, dB/cm loss=log(10)*loss_dBcm/10*100; % Simulation Parameters: span=30e-9; % Set the wavelength span for the simultion Npoints = 10000; % from MODE calculations switch 1 case 1 % Strip waveguide; 500x220 nm neff_wavelength = @(w) 2.4379 - 1.1193 * (w*1e6-1.554) - 0.0350 * (w*1e6-1.554).^2; % 500x220 oxide strip waveguide dneff_width = @(w) 10.4285*(w-0.5).^3 - 5.2487*(w-0.5).^2 + 1.6142*(w-0.5); end % Find Bragg wavelength using lambda_Bragg = Period * 2neff(lambda_bragg); % Assume neff is for the average waveguide width. f = @(lambda) lambda - Period*2*(neff_wavelength(lambda)+(dneff_width(width2)+dneff_width(width1))/2); wavelength0 = fzero(f,1550e-9); wavelengths=wavelength0 + linspace(-span/2, span/2, Npoints); n1=neff_wavelength(wavelengths)+dneff_width(width1); % low index n2=neff_wavelength(wavelengths)+dneff_width(width2); % high index [R,T]=TMM_Grating_RT(wavelengths, Period, NG, n1, n2, loss);figure; plot (wavelengths*1e6,[R, T],'LineWidth',3); hold all plot ([wavelength0, wavelength0]*1e6, [0,1],'--'); % calculated bragg wavelength xlabel('Wavelength [\mum]') ylabel('Response'); axis tight;

计算结果如下图所示：

5、Lumerical求解

通过Lumerial的FDTD求解光栅透射率代码如下：

############################################### # script file: Bragg_FDTD.lsf # # Create and simulate a basic Bragg grating # Copyright 2014 Lumerical Solutions ############################################### # DESIGN PARAMETERS ############################################### thick_Si = 0.22e-6; thick_BOX = 2e-6; width_ridge = 0.5e-6; # Waveguide width Delta_W = 50e-9; # Corrugation width L_pd = 324e-9; # Grating period N_gt = 280; # Number of grating periods L_gt = N_gt*L_pd;# Grating length W_ox = 3e-6; L_ex = 5e-6; # simulation size margins L_total = L_gt+2*L_ex; material_Si = ’Si (Silicon) - Dispersive & Lossless’; material_BOX = ’SiO2 (Glass) - Const’; # Constant index materials lead to more stable simulations # DRAW ############################################### newproject; switchtolayout; materials; # Oxide Substrate addrect; set(’x min’,-L_ex); set(’x max’,L_gt+L_ex); set(’y’,0e-6); set(’y span’,W_ox); set(’z min’,-thick_BOX); set(’z max’,-thick_Si/2); set(’material’,material_BOX); set(’name’,’oxide’); # Input Waveguide addrect; set(’x min’,-L_ex); set(’x max’,0); set(’y’,0); set(’y span’,width_ridge); set(’z’,0); set(’z span’,thick_Si); set(’material’,material_Si); set(’name’,’input_wg’); # Bragg Gratings addrect; set(’x min’,0); set(’x max’,L_pd/2); set(’y’,0); set(’y span’,width_ridge+Delta_W); set(’z’,0); set(’z span’,thick_Si); set(’material’,material_Si); set(’name’,’grt_big’);addrect; set(’x min’,L_pd/2); set(’x max’,L_pd); set(’y’,0); set(’y span’,width_ridge-Delta_W); set(’z’,0); set(’z span’,thick_Si); set(’material’,material_Si); set(’name’,’grt_small’);selectpartial(’grt’); addtogroup(’grt_cell’); select(’grt_cell’); redrawoff; for (i=1:N_gt-1) { copy(L_pd); } selectpartial(’grt_cell’); addtogroup(’bragg’); redrawon;# Output WG addrect; set(’x min’,L_gt); set(’x max’,L_gt+L_ex); set(’y’,0); set(’y span’,width_ridge); set(’z’,0); set(’z span’,thick_Si); set(’material’,material_Si); set(’name’,’output_wg’);# SIMULATION SETUP ############################################### lambda_min = 1.5e-6; lambda_max = 1.6e-6; freq_points = 101; sim_time = 6000e-15; Mesh_level = 2; mesh_override_dx = 40.5e-9; # needs to be an integer multiple of the period mesh_override_dy = 50e-9; mesh_override_dz = 20e-9;# FDTD addfdtd; set(’dimension’,’3D’); set(’simulation time’,sim_time); set(’x min’,-L_ex+1e-6); set(’x max’,L_gt+L_ex-1e-6); 158 Fundamental building blocks set(’y’, 0e-6); set(’y span’,2e-6); set(’z’,0); set(’z span’,1.8e-6); set(’mesh accuracy’,Mesh_level); set(’x min bc’,’PML’); set(’x max bc’,’PML’); set(’y min bc’,’PML’); set(’y max bc’,’PML’); set(’z min bc’,’PML’); set(’z max bc’,’PML’);#add symmetry planes to reduce the simulation time #set(’y min bc’,’Anti-Symmetric’); set(’force symmetric y mesh’, 1); # Mesh Override if (1){ addmesh; set(’x min’,0e-6); set(’x max’,L_gt); set(’y’,0); set(’y span’,width_ridge+Delta_W); set(’z’,0); set(’z span’,thick_Si+2*mesh_override_dz); set(’dx’,mesh_override_dx); set(’dy’,mesh_override_dy); set(’dz’,mesh_override_dz); }# MODE Source addmode; set(’injection axis’,’x-axis’); set(’mode selection’,’fundamental mode’); set(’x’,-2e-6); set(’y’,0); set(’y span’,2.5e-6); set(’z’,0); set(’z span’,2e-6); set(’wavelength start’,lambda_min); set(’wavelength stop’,lambda_max);# Time Monitors addtime; set(’name’,’tmonitor_r’); set(’monitor type’,’point’); set(’x’,-3e-6); set(’y’,0); set(’z’,0); addtime; set(’name’,’tmonitor_m’); set(’monitor type’,’point’); set(’x’,L_gt/2); set(’y’,0); set(’z’,0); addtime; set(’name’,’tmonitor_t’); set(’monitor type’,’point’); set(’x’,L_gt+3e-6); set(’y’,0); set(’z’,0);# Frequency Monitors addpower; set(’name’,’t’); set(’monitor type’,’2D X-normal’); set(’x’,L_gt+2.5e-6); set(’y’,0); set(’y span’,2.5e-6); set(’z’,0); set(’z span’,2e-6); set(’override global monitor settings’,1); set(’use source limits’,1); set(’use linear wavelength spacing’,1); set(’frequency points’,freq_points); addpower; set(’name’,’r’); set(’monitor type’,’2D X-normal’); set(’x’,-2.5e-6); set(’y’,0); set(’y span’,2.5e-6); set(’z’,0); set(’z span’,2e-6); set(’override global monitor settings’,1); set(’use source limits’,1); set(’use linear wavelength spacing’,1); set(’frequency points’,freq_points);#Top-view electric field profile if (0) {addprofile; References 159 set(’name’,’field’); set(’monitor type’,’2D Z-normal’); set(’x min’,-2e-6); set(’x max’,L_gt+2e-6); set(’y’, 0); set(’y span’,1.2e-6); set(’z’, 0); set(’override global monitor settings’,1); set(’use source limits’,1); set(’use linear wavelength spacing’,1); set(’frequency points’,21); }# SAVE AND RUN ############################################### save(’Bragg_FDTD’); run; # Analysis ############################################### transmission_sim=transmission(’t’); reflection_sim=transmission(’r’); wavelength_sim=3e8/getdata(’t’,’f’); plot(wavelength_sim*1e9, 10*log10(transmission_sim),10*log10(abs(reflection_sim)),’wavelength (nm)’, ’response’); legend(’T’,’R’); matlabsave(’Bragg_FDTD’);

总结

以上是生活随笔为你收集整理的利用传输矩阵法求解布拉格光栅的透射谱的全部内容，希望文章能够帮你解决所遇到的问题。

如果觉得生活随笔网站内容还不错，欢迎将生活随笔推荐给好友。