interference microscopy

Images made through a microscope. All subject types.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

mtuell
Posts: 289
Joined: Wed May 11, 2016 12:42 pm
Location: Tucson, AZ
Contact:

interference microscopy

Post by mtuell »

I finally got back around to trying my Nikon M Plan DI 10x objective, which is a Mirau interferometer objective. It takes some doing, but I finally got some pictures of a bare microscope slide with some gunk on it. The fringes are a half wavelength apart, so for the green version (with a GIF filter), about 275 nm (edited typo) in vertical surface height apart.

Image

Image

As you go away from perfect focus (the central fringe with black fringes around it), the contrast reduces quickly due to the low coherence length. A laser source would have uniform contrast fringes. The colors change in white-light because the different wavelengths interfere differently and it gets smeared out.

Generally, this technique is used for very flat objects such as polished glass or silicon wafers, etc, to measure height variation.

Mike
Last edited by mtuell on Sat Mar 25, 2017 11:40 pm, edited 1 time in total.

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Haven't heard of that. What's special in the objective itself?
Chris R

mtuell
Posts: 289
Joined: Wed May 11, 2016 12:42 pm
Location: Tucson, AZ
Contact:

Post by mtuell »

It has a reference mirror and beam splitter to give interference.

Image

Image

The diagram is not mine, but the objective photo is. There is some info on my site, with some good links.

Mike

mtuell
Posts: 289
Joined: Wed May 11, 2016 12:42 pm
Location: Tucson, AZ
Contact:

Post by mtuell »

For a little more concrete example, here is a computer generated hologram (CGH) with the 10x DI objective, but not aligned for fringes.

Image

So, we have basically vertical bands that define the binary diffraction pattern.

Bringing in the interference, we see a jump of the colors at each transition. The distance (image) between the red and black lines is approximately 25% of the distance between fringes (green lines). Since each fringe is lambda/2, we can approximate that each vertical bar is lambda/8, or about 70 nm high. With a 10x objective, we are getting vastly more vertical resolution than lateral, showing the power of the interference technique for certain types of specimens.

Image

Using a green filter (GIF), the coherence length is increased, giving better contrast and less confusion in many regards.

Image

All images are 100% crops of 10 MP USB camera images.

With phase-shifting and computer analysis, accurate surface height maps can be derived, able to measure on the order of 1 nm rms roughness.

Another point to make is that the CGH is chrome-on-glass, whic explains why some bands are brighter than the adjacent ones. Also, 70 nm is a typical chrome thickness, so our measurement makes some sense!

Mike

Smokedaddy
Posts: 1951
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

Very interesting but over my head. Is this another means to measure the surface roughness of a telescope mirror?

mtuell
Posts: 289
Joined: Wed May 11, 2016 12:42 pm
Location: Tucson, AZ
Contact:

Post by mtuell »

Yes, James, surface roughness is one of the main applications, but it can be used for other things. Without some difficulty, it is limited to surfaces with small slopes, so wouldn't be too useful for many everyday subjects.

In the future, I'll put up some more examples of what it is good for and what it isn't.

Mike

Smokedaddy
Posts: 1951
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »

When I was doing QC stuff in our high purity fab-shop I had to measure the surface roughness of the inside of the electropolished tubing to see if it was within spec., but I did that with a stylist. I think it was called RMS, dunno, it's been a long time since I've done that. Interesting to me but not to most that I worked with. <g>

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Interesting Mike, thanks.
Good for Si wafers?
Chris R

mtuell
Posts: 289
Joined: Wed May 11, 2016 12:42 pm
Location: Tucson, AZ
Contact:

Post by mtuell »

Yes, Chris, silicon wafers and just about anything smooth, flat and reflective would make good subjects. To make measurements, the slope can't be too high, or you will lose track of how many fringes it has jumped.

siliconGary
Posts: 73
Joined: Tue Oct 18, 2016 2:40 pm
Location: Canada
Contact:

Post by siliconGary »

I can recall using this technique (back in the 80's) to measure Silicon junction depth. We polished a shallow bevel in the silicon, and measured electrical junction with a 2/3 point probe that auto-stepped down the bevel, and then calibrate the depth using the interference microscope to accurately measure the bevel angle. I recall it being a very sensitive set-up with a small knob on the objective - it was difficult to get set-up correctly.

These days any fine step height measurement is done using Atomic Force Microscopy (Which uses a diamond tip stylus and sophisticated electronics).

mtuell
Posts: 289
Joined: Wed May 11, 2016 12:42 pm
Location: Tucson, AZ
Contact:

Post by mtuell »

Steps can also be measured by this technique if you have very fine control / knowledge of the vertical position of the objective. This is called vertical-scanning white-light interference microscopy. Various manufacturers, including Nikon, make such instruments. But, yes, AFM is also used.

As for James' comment, contact stylus probes can be used to some level, but likely not to the sub-nm rms level. One brand is Talyrond, but there are likely others.

Mike

carlos.uruguay
Posts: 5358
Joined: Thu Feb 23, 2012 8:05 pm
Location: Uruguay - Montevideo - America del Sur
Contact:

Post by carlos.uruguay »

Nice and Interesting

Smokedaddy
Posts: 1951
Joined: Sat Oct 07, 2006 10:16 am
Location: Bigfork, Montana
Contact:

Post by Smokedaddy »


Post Reply Previous topicNext topic