Consider the nanohole array illustrated here.
The structure is made up of a glass substrate with a thin film of gold on top.
A rectangular pattern of holes is etched into the gold layer.
Let’s say we are interested in obtaining the reflection and transmission spectrum of
plane wave light at normal incidence from the air above the structure over the visible
spectrum, 400-700 nm.
The thickness of the gold film and the radius of the holes etched in the gold layer are
both 100 nm which is sub wavelength.
For this structure, we will use the finite-difference time domain, or FDTD, a method to simulate
the device's performance.
But what is the motivation for running a simulation in the first place?
Running simulations is faster and cheaper than fabricating prototypes and then making
experimental measurements, especially if you want to get results for a range of different
Simulations are a great way to evaluate and optimize design parameters before manufacturing
Simulations can also be used to verify experimentally obtained results.
FDTD Solutions solves Maxwell’s equations for arbitrarily complex geometries, where
the device performance may not be intuitive, and where no analytical methods exist to predict
the device behavior.
The FDTD method is particularly good for stimulating devices which have feature sizes on the order
of a wavelength or smaller.
This is when geometric optics, such as Snell’s Law, break down which means ray tracing methods
will not give accurate results.
FDTD is also well suited to situations where broadband results are desired.
The time domain nature of FDTD means that broadband results can be obtained from a single
More details about the FDTD algorithm can be found in the FDTD Algorithm section of
In this section of the course, we will demonstrate how to set up, run, and collect the results
for the nanohole array structure that we just described.
The goal is to show the simulation workflow and introduce the some of the capabilities
of FDTD Solutions.