Frequency-dependent finite-difference time-domain method based on iterated Crank-Nicolson scheme

The finite-difference time-domain (FDTD) method based on the iterated Crank-Nicolson (ICN) scheme is extended to a frequency-dependent version. The Drude model is used to express a metal dispersion, which is incorporated into the iterated Crank-Nicolson formulation with the trapezoidal recursive convolution technique. The validity of the present finite-difference time-domain method with convolutional perfectly matched layers is discussed through the analysis of a metal-insulator-metal plasmonic waveguide. Numerical results obtained from a two-iteration technique are found to agree well with those from the traditional explicit finite-difference time-domain method.

Automated summary

This paper extends the finite-difference time-domain (FDTD) method based on the iterated Crank-Nicolson (ICN) scheme to a frequency-dependent version. The Drude model is used to describe the metal dispersion and is incorporated into the iterated Crank-Nicolson formulation with the trapezoidal recursive convolution technique. The effectiveness of the FDTD method with convolutional perfectly matched layers is examined through the analysis of a metal-insulator-metal plasmonic waveguide, and the numerical results obtained using a two-iteration technique show good agreement with those from the traditional explicit FDTD method.