torcwa_AuNanobar2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "mount_file_id": "1bEAujdNhz6jSglhh7cj2X_BgIpW-CvQN",
      "authorship_tag": "ABX9TyNuQQf7oExntDBoqGt+i691",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/MahmutRuzi/photonics-sim-examples/blob/main/torcwa_AuNanobar2.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Reflectance/transmitance spectrum of gold nanobar\n",
        "\n",
        "In this example, we simulate the reflectance/transmitance spectrum of arrays of gold nanobars periodically (square lattice) placed on a silicon substrate. For simulations, we use the rigorosuly coupled wave analysis approach (RCWA), as implemented in [torcwa](https://github.com/kch3782/torcwa). The aim is to reproduce Fig.3 (red) from a published work by H.Altug et al. [Optical Express **2011**, 19 (12) 11202](https://doi.org/10.1364/OE.19.011202). The whole simulation takes around 16 s in free version of colab. \n",
        "\n",
        "\n",
        "\n",
        "Author: Mahmut Ruzi\n",
        "\n",
        "Contact: mruzi17@gmail.com"
      ],
      "metadata": {
        "id": "UZjnz4eWc_97"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "We first check if GPU acceleration using cuda is enabled. "
      ],
      "metadata": {
        "id": "zvafdMGOixKS"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "torch.cuda.is_available()"
      ],
      "metadata": {
        "id": "AIMwd74o3cj1",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "66c95629-0c41-4e2c-faba-f43c462a0dbf"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {},
          "execution_count": 1
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's install the RCWA package [torcwa](https://github.com/kch3782/torcwa)."
      ],
      "metadata": {
        "id": "pQA97XsgjZss"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!pip3 install torcwa"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7uxkJ1h8zwtR",
        "outputId": "320c6457-fcd3-4bd8-9c9d-0482f2ba85e1"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting torcwa\n",
            "  Downloading torcwa-0.1.3-py3-none-any.whl (12 kB)\n",
            "Requirement already satisfied: torch>=1.10.1 in /usr/local/lib/python3.8/dist-packages (from torcwa) (1.13.0+cu116)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.8/dist-packages (from torch>=1.10.1->torcwa) (4.4.0)\n",
            "Installing collected packages: torcwa\n",
            "Successfully installed torcwa-0.1.3\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Check if it is correctly installed and which version."
      ],
      "metadata": {
        "id": "vq0smOjEkxcy"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import torcwa\n",
        "print(torcwa.__version__)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "THOROLM_yJnC",
        "outputId": "f1037aff-891a-42e0-cce2-14061215041b"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.1.3\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "For this simulation, we need to import the refractive indices of gold. We used the data for evaporated gold measured by [Olmen et al](https://doi.org/10.1103/PhysRevB.86.235147). The data is conveneintly collected in [refractiveindex.info](https://refractiveindex.info/?shelf=main&book=Au&page=Olmon-ev). The file is available in this github depository, called 'Au_Olmen2012.csv'. It needs to be uploaded to your gdrive. Use the 'Files' menue (top left) in colab and upload it. Then, we use numpy to read it."
      ],
      "metadata": {
        "id": "u8GYPfeJlACn"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np"
      ],
      "metadata": {
        "id": "1rOy3DqOyTvy"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's see the contents of current working folder."
      ],
      "metadata": {
        "id": "ZP8MSyVAobpR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!ls"
      ],
      "metadata": {
        "id": "dstI4QGQyrr1",
        "outputId": "a41375a7-6be2-4a76-ed8f-7081c2716378",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Au_Olmen2012.csv  sample_data\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Once we're sure that file 'Au_Olmen2012.csv' is visible in the folder, we can read it and calculate the complex dielectric constants from it."
      ],
      "metadata": {
        "id": "0FMO939copGs"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "refrat_cSi = np.genfromtxt(\"Au_Olmen2012.csv\", delimiter=',',skip_header=1, unpack=True)"
      ],
      "metadata": {
        "id": "nqjqju7Jz2P7"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def eps_cSi(lamba0):\n",
        "    # lamba0 should be in um\n",
        "    n_interp = np.interp(lamba0, refrat_cSi[0,::], refrat_cSi[1,:])\n",
        "    k_interp = np.interp(lamba0, refrat_cSi[0,::], refrat_cSi[2,:])\n",
        "    epsr = n_interp**2-k_interp**2\n",
        "    epsi = 2*n_interp*k_interp\n",
        "    eps_cSi = epsr+1.0j*epsi\n",
        "    return eps_cSi\n",
        "\n",
        "#%%"
      ],
      "metadata": {
        "id": "-8Js-qdAz71d"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "We simulate the spectrum for a total of 27 discrete wavelengths. "
      ],
      "metadata": {
        "id": "qxTCIxbSpGDT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# lamba is in um units here\n",
        "lamba=10000./np.array([1250,1300,1350,1400,1450,1500,1550,1580,1600,1610,1620,\n",
        "                       1630,1640,1645,1650,1655,1660,1665,1670,1675,1680,1690,\n",
        "                       1700,1720,1750,1800,1850])"
      ],
      "metadata": {
        "id": "AQZ5Imrl0D5P"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's set the data and devie type."
      ],
      "metadata": {
        "id": "ReRthh_ipiiT"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "sim_dtype = torch.complex64\n",
        "geo_dtype = torch.float32\n",
        "device = torch.device('cuda')"
      ],
      "metadata": {
        "id": "kQFB41R40KwQ"
      },
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now we plot the design and plot the initial structure. "
      ],
      "metadata": {
        "id": "WyiRVY2Lpw-U"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "# geometry\n",
        "L = [1750, 1750]            # period in the X and Y plane , in nm / nm\n",
        "torcwa.rcwa_geo.dtype = geo_dtype\n",
        "torcwa.rcwa_geo.device = device\n",
        "torcwa.rcwa_geo.Lx = L[0]\n",
        "torcwa.rcwa_geo.Ly = L[1]\n",
        "torcwa.rcwa_geo.nx = 1750\n",
        "torcwa.rcwa_geo.ny = 1750\n",
        "torcwa.rcwa_geo.grid()\n",
        "torcwa.rcwa_geo.edge_sharpness = 1000. \n",
        "\n",
        "x_axis = torcwa.rcwa_geo.x.cpu()\n",
        "y_axis = torcwa.rcwa_geo.y.cpu()\n",
        "\n",
        "# layers\n",
        "Lx = 870. # length along the X axis, in nm\n",
        "Ly = 230. # width along the y axis, in nm\n",
        "layer0_geometry = torcwa.rcwa_geo.rectangle(Wx=Lx,Wy=Ly,Cx=L[0]/2.,Cy=L[1]/2.)\n",
        "layer0_thickness = 70.  # in nm\n",
        "\n",
        "#%%\n",
        "# View layers\n",
        "plt.figure(2)\n",
        "plt.imshow(torch.transpose(layer0_geometry,-2,-1).cpu(),origin='lower',extent=[x_axis[0],x_axis[-1],y_axis[0],y_axis[-1]])\n",
        "plt.title('Layer 0')\n",
        "plt.xlim([0,L[0]])\n",
        "plt.xlabel('x (nm)')\n",
        "plt.ylim([0,L[1]])\n",
        "plt.ylabel('y (nm)')\n",
        "plt.colorbar()\n",
        "plt.grid()\n",
        "plt.tight_layout()\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "id": "Ie4Cn8tt0TOB",
        "outputId": "b188bd44-0187-4b6e-8b82-98365c6c81f5"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Generate and perform simulation\n",
        "import time\n",
        "start_time = time.time()\n",
        "order_Nx = 10 # Fourier truncation order\n",
        "order_Ny = 5\n",
        "order = [order_Nx,order_Ny]\n",
        "lamb0 = torch.tensor(1000*lamba,dtype=geo_dtype,device=device) #Wavelength in nm\n",
        "#%%\n",
        "# Simulation environment\n",
        "# light\n",
        "start_time = time.time()\n",
        "inc_ang = 0.*(np.pi/180)    # radian\n",
        "azi_ang = 0.*(np.pi/180)    # radian\n",
        "substrate_eps = 3.41**2\n",
        "input_eps = 1**2\n",
        "txx = []\n",
        "rxx = []\n",
        "for lamb0_ind in range(len(lamb0)):\n",
        "    lamb0_now = lamb0[lamb0_ind]\n",
        "    sim = torcwa.rcwa(freq=1/lamb0_now,order=order,L=L,dtype=sim_dtype,device=device)\n",
        "    #sim.add_input_layer(eps=input_eps)\n",
        "    sim.add_output_layer(eps=substrate_eps)\n",
        "    sim.set_incident_angle(inc_ang=inc_ang,azi_ang=azi_ang)\n",
        "    silicon_eps = eps_cSi(lamb0_now.cpu()/1000)  # note eps_cSi takes wavelength in microns\n",
        "    layer0_eps = layer0_geometry*silicon_eps + (1.-layer0_geometry)\n",
        "    sim.add_layer(thickness=layer0_thickness,eps=layer0_eps)\n",
        "    sim.solve_global_smatrix()\n",
        "    rxx.append(sim.S_parameters(orders=[0,0],direction='forward',port='reflection',polarization='xx',ref_order=[0,0]))\n",
        "    txx.append(sim.S_parameters(orders=[0,0],direction='forward',port='transmission',polarization='xx',ref_order=[0,0]))\n",
        "txx = torch.cat(txx)\n",
        "rxx = torch.cat(rxx)\n",
        "print(\"--- %s seconds ---\" % (time.time() - start_time))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lLMyNmtL0nje",
        "outputId": "059db598-5e97-4c13-e2ff-163ee98e915f"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--- 16.003005743026733 seconds ---\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Now let's plot the data !"
      ],
      "metadata": {
        "id": "UER4uTwBq0Mh"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(3)\n",
        "plt.plot(1e7/lamb0.cpu(),torch.abs(txx).cpu()**2,'bo',1e7/lamb0.cpu(),torch.abs(rxx).cpu()**2,'ro')\n",
        "#plt.plot(lamb0.cpu()/1000,torch.abs(txx).cpu()**2,'b')\n",
        "plt.title('Ref/trans (order: '+str(order_Nx)+')')\n",
        "plt.xlabel('Wavenumber (cm-^1)')\n",
        "plt.ylabel('Transmittance/Reflectance (a.u.)')\n",
        "plt.grid()\n",
        "plt.tight_layout()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        },
        "id": "D7VEsVkE7g7T",
        "outputId": "0ee44455-9607-4580-89d4-f508f69fb4a3"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}