Sunday, November 15, 2020

Etching silicon wafers to make colorful Rugate optical filters (porous silicon)

Passing an electrical current through a silicon wafer in a special acid etchant will create a porous layer with a variable index of refraction. I describe how this process works, and how the Fourier transform relates filter design to electrical etch waveform and resulting spectral response.

Notes: 1. The Rugate filters look especially good in polarized light because the reflection from the silicon wafer is reduced, but the reflection from the filter remains strong. I noticed this while holding the wafer in front of my computer monitor. Later, when showing it to friends, the color intensity was poor. Their monitor must have been horizontally polarized, so holding the wafer low, and tilting it upward didn't work! Some monitors are horizontally polarized, and some are vertical. 2. The magnet used to hold down the PTFE cup to the wafer may have a very slight impact on the etch process. The dramatic shift in filter performance at the periphery is due to O-ring restricting the conductive etchant to a sharp edge, creating an electrical field concentration. Outstanding visual Fourier series tutorial: LR-1 spectrometer: p-type wafers on eBay. (You have to hunt around and check the photos for info on the label attached to the box of wafers. Be sure to get wafers less than 0.01 ohm-cm) eg n-type wafers are photosensitive during the etch process. Online graphing calculator page from this video: Original Desmos page that I used: Code for Keithley 2450 control and waveform generation: Fourier transform to understand optical coatings: Porous silicon refs:

1 comment:

  1. Firstly I would just like to say it always makes my day when an Applied Science video comes out, keep up the amazing work!

    I actually have a question for you about your DIY waterjet project. I am a senior in Mechanical Engineering and my capstone design project is to create a CNC controlled waterjet cutting system. Your video was a great place to start to gain understanding about how waterjet systems work and we actually went with the same A2 cutting system that you used. We have our CNC system up and running but the waterjet portion is proving to be more difficult than we thought it would be. In the first video you mentioned that you had experienced clogging issues in the nozzle. We have run into the same issue (40 thou focusing tube and 22 thou orifice) and were wondering what you did to troubleshoot the issue. It seems that the garnet is getting wet right at the inlet to the mixing chamber and that the suction produced by the water is not enough to overcome the clogging. We have an 18 thou orifice on order to test with but aren't sure if it will solve the problem. See the link below for footage of a cut test that clogged.
    I can be reached at

    Link to cut test: