Silvaco

A: Earlier version of ATLAS were able to handle refraction and reflection of rays during the optical simulation in Luminous. New features have been added to ATLAS version 4.4.25 or higher to include coherence and so allow simulation of the effects of anti-reflective coating.

The simulation of photodetector efficiency versus wavelength is done using the LAMBDA parameter of the SOLVE statement. This parameter sets the wavelength for each solution. A new ray trace is done at each solution point to recalculate the optical intensities. The syntax might be:

SOLVE B1=1 LAMBDA=0.6 INDEX.CHECK

The parameter B1 sets the intensity of the light source #1 in Watts/cm2. A useful parameter INDEX.CHECK is recommended for use here. It reports the real and imaginary parts of the refractive indices of each material involved in the ray trace calculation. This can be useful to check the default parameters for refractive index especially at very high or low wavelengths.

The INTERFACE statement is used to define the properties of the anti-reflective layer on the top of the silicon. Note that the layer does not need to be physically present in the structure defined for ATLAS.

INTERFACE AR.INDEX=2.05 AR.THICK=0.05

The parameters AR.THICK defines the ARC thickness in microns. Most commonly this is a quarter wavelength of the light taking into account the wavelength in the ARC material. AR.INDEX defines the real part of refractive index of the ARC. This is ideally the square root of the substrate material refractive index.

Figure 1 compares the spectral response of the cases with two thickness of ARC and without the ARC. Note the increase around 550nm wavelength for the 70nm ARC. The detection efficiency is defined as the ratio of two parameters. This quantity can be plotted using the functions in TonyPlot.

Figure 1. Comparison of detector efficiency for different anti-reflective coating thickness

source photocurrent is the amount of current generated by the light source. available photocurrent is the amount of current absorbed by the semiconductor. Differences between these two are due to reflection, transmission or absorption in non-semiconductor materials. The ratio of available photocurrent/ source photocurrent is often known as external quantum efficiency. An alternative approach using the ratio of actual terminal current from the detector to the source photocurrent is also possible.

In this example the ARC is planar and non-absorbing. The light is also normally incident. ATLAS also has features to model the ARC with non-normal incidence. More complex cases such as absorbing ARCs can be modeled by using user-defined equations in the C-Interpreter function for reflection, F.REFLECT.

Q: Can a similar detection efficiency be plotted against an oscillation frequency of the light?

A. Small signal analysis for light beams is available in Luminous in a similar manner to the AC voltage analysis for electrical simulations.

The syntax to do this is:

SOLVE B1=1 SS.PHOTO FREQ=1E6 FSTEP=5 MULT.F \ NFSTEPS=8 BEAM=1

The parameter SS.PHOTO enables the small signal light oscillations. The frequency parameters operate as in electrical simulations to control the initial frequency, frequency step size and number of steps to ramp the frequency. The BEAM parameter specifies the light beam to be ramped.

Figure 2 shows the results of small signal light response for an avalanche photodetector. The y-axis is detector efficiency defined as terminal current divided by the source photocurrent. Since the detector has carrier multiplication through impact ionization the efficiency at low frequencies is greater than unity.

Figure 2. Response of avalanche photodetector with respect to frequency of light oscillator