# Help to find a simple setup for optical surface topography

1. Nov 25, 2015

### FrodCube

Hi everyone.

I'm looking for some help from someone expert in optics to create an inteferometry based setup to map the surface of a small object. This is for a small lab project so I'm not looking for a really complex setup, I just need something simple that can be built in a short amount of time.

I'd like to find resources that explains all stages of the process like the actual building, noise reduction and data analysis.

Keep in mind that me and my lab group have basically no experience in this field at all. If you need more details about what I want to do please ask and I'll try to explain it,

2. Nov 25, 2015

### Staff: Mentor

Welcome to the PF.

I've sent a link to this thread to a friend of mine who may be able to help.

3. Nov 25, 2015

### Staff: Mentor

BTW, have you looked for such a resource near you? Are you at a university or near one?

4. Nov 25, 2015

### FrodCube

Thank you.

Yes I am at a university. This is for a lab course and my professor is actually in the Virgo collaboration, so he knows quite a lot about interferometry. Of course we asked him but he said that he doesn't know very much about this kind of application, he has no simple resources to give us and also he thinks that this would probably be too difficult for us. That's why I'm asking this: I want to know if this is possible to do in an easy way because this was our initial idea for our project.

Browsing through the web I found papers on the subject but most of them are from a theoretical point of view and they only explain the basic principles behind this without actually discussing the practical implementation and some of them are way too advanced for us.

5. Nov 25, 2015

### Staff: Mentor

How small is the object? Can you just use stereo video analysis to map it instead of interferometry?

6. Nov 25, 2015

### Andy Resnick

What size is the object, and how would you characterize the surface roughness? For example, the surface of a small ball bearing is very smooth but not flat; the surface of sandpaper is flat but not smooth. Is the surface reflective or absorptive? How accurately and precisely do you need to quantify the surface?

Reflection Differential Interference Contrast imaging (https://www.microscopyu.com/articles/dic/reflecteddic.html) is by far my favorite method to image surfaces- the local slope is encoded in color, and the images are quite striking. However, the surface needs to be reflective and nearly flat.

7. Nov 26, 2015

### FrodCube

Our first idea was to see the surface of a cd-ROM to get something like this:

The main point is that since we are doing this for a lab course, we should design and assemble most of the setup ourself.

We were thinking about interferometry since we've already built one and we can sort of make it work.

8. Nov 26, 2015

### nasu

What you see here is not the surface of a CD but of the inner aluminum layer. It may be an AFM image.
You need to peel one of the plastic halves to get this by AFM.
You may be able to get a similar image by optical means without destroying the CD. Just be aware that it is not on the surface and you may need to deal with the multiple reflections in the plastic layer.

9. Nov 26, 2015

### FrodCube

Yeah we considered that. We thougth of someway removing the plastic coating and perform the measurement only on the metal layer that has the bumps

10. Nov 28, 2015

### Andy Resnick

I'll try a few different reflection imaging methods when I get back to my lab on Monday and post what I get. The plastic is probably going to cause a few problems.... we'll see.

11. Nov 30, 2015

### Andy Resnick

As I suspected, this is not trivial. The plastic prevents use of DIC, and the thickness of the plastic largely prevents use of high-NA lenses needed to resolve the features. Here's the best of what I could get, using a CD-ROM, because it has the largest feature sizes. I located the border between recorded and blank tracks for these images:

This is using a 63x dipping objective, broadband source:

This is using a 16x objective, reflected DIC, with the prism shoved all the way to one side and the aperture stop nearly closed to enhance contrast:

And lastly, a 16x darkfield objective:

None of these qualify as "simple and cheap"-mostly because the required mechanical stability and precision is very high (submicron). And the results are not quantitative.