Macro: I am glad that I persisted.

[lukaszgryglicki]

6 hours ago, colinbm said:

I have these filters on the 255nm LED lights.
https://www.asahi-spectra.com/opticalfilters/detail.asp?key=ZUV0325

I saw them, considered once, do you have OD chart for them?

Are they complete black when you look through them?

How about looking at sharp angle?

 

[colinbm]

The OD chart is there if you select it in the " View Settings " on the right of the page.
If you find anything better let me know please.

[lukaszgryglicki]

I only have dichroic from Edmund, it seems to be good only at 90 degrees angle, otherwise I can see visible light through it using my own eye (without any 255 light).

 

[Lou Jost]

9 hours ago, colinbm said:

Thanks Lou
I have often wondered about doubling & halving wavelengths.
A common green laser are actually an IR laser, emitting natively 1064nm light, but cut in half.

 The issue I just mentioned is another consequence of wave behavior....If two parallel reflective surfaces are one wavelength apart or ten or thirty wavelengths apart, the interference between the two reflections is the same.

 

So instead of saying that the interference-producing structures must be about the same size as the light they are manipulating, we should have said that the PRECISION of those structures has to be at the level of the wavelength of the light they are manipulating.

[colinbm]

Thanks Lou

[Lou Jost]

 

So here is an oblique view of the ribs of a scale from a Sunset Moth. You can see that each rib is made up of many thin layers piled on top of one another. These are the structures that produce the colors. So it can be done, Colin!  However, it is important to note that the objective I used for this has an NA of 0.90. I think you can manage with a lower NA if you use UV light, but still, you should use the highest NA objective that your set-up will permit. Remember, the magnification does not matter much, it is the NA (and wavelength) that determines the resolution of the aerial image. You can adjust the size of the image on the sensor by choosing a longer tube lens or adding a teleconverter.

[colinbm]

Thank you very much Lou, this looks good.
I have just setup a 20x objective on a stand what I thought was pretty stable.
The camera is fixed & I move the stage, I can guess about 26 steps in half a mm.
For a 60x objective I will need to build a substantially more stable setup.
But I have next year to do that.
I will post some pictures tomorrow.

PS, in one of my recent posts I have shown a butterfly scale that has bowel shaped structures that give it a green iridescence.
I am guessing that the Green Jewel Beetle has something similar ?

PSS, I have this coming, hopefully before Christmas.


Thanks again for joining us in the forum.

[OlDoinyo]

5 hours ago, lukaszgryglicki said:

I only have dichroic from Edmund, it seems to be good only at 90 degrees angle, otherwise I can see visible light through it using my own eye (without any 255 light).

 

That is typical dichroic behavior. Once the light path is tilted, the interlayer spacing becomes greater than specification and performance degrades.

[OlDoinyo]

3 hours ago, Lou Jost said:

 The issue I just mentioned is another consequence of wave behavior....If two parallel reflective surfaces are one wavelength apart or ten or thirty wavelengths apart, the interference between the two reflections is the same.

 

So instead of saying that the interference-producing structures must be about the same size as the light they are manipulating, we should have said that the PRECISION of those structures has to be at the level of the wavelength of the light they are manipulating.

Technically true. But once we are at 10x or greater relative spacing, the alignment issues become hugely more daunting in practice. In a lab you could no doubt pull it off, but I am skeptical that any naturally evolved structure will turn out to work that way. Heck, even 3x would be a surprise to me, and that would be 1.5 micrometers with green light, which will be pretty challenging optically. Supposedly you can see something down to 500 nm at that wavelength, but it may be heavily distorted by diffraction.

[Lou Jost]

Colin, that 60x is a finite objective, I guess I had forgotten that you are using finite optics. So my comment about changing magnification by changing tube length is not applicable to that objective. It makes sense to use a finite optic for UV imaging. Tube lenses add a lot of glass. But I think you should consider a finite Fluor objective instead of (or in addition to) that one. Those are designed to transmit UV.

 

[Lou Jost]

1 hour ago, OlDoinyo said:

Technically true. But once we are at 10x or greater relative spacing, the alignment issues become hugely more daunting in practice.

Yes, we agree, my example was just to illustrate the theory. But even if the separation is only a few wavelengths, the structures could be visible with light microscopy. There is hope!

[Andy Perrin]

18 hours ago, Lou Jost said:

I have been thinking about why my photos  (see link above) seem to capture the iridescence-producing layers even without using UV light. We all have made a mistake in our assumption that the structures that produce the color must have the same scale as the wavelength of light. The physics of interference does not require the layer spacing to be a set distance. The physical requirement for interference is for the distance to be an integer multiple of the special wavelength. The structures could be much bigger than the wavelength of the light that they reflect.

Whatever the merits of the theory here, it's never too late to inject some experimental facts into the discussion... this is from a 2013 paper, "Structural coloration in nature" by Jiyu Sun, Bharat Bhushan, and Jin Tonga. 

 

 

The first picture in row A shows a green Japanese jewel beetle and the smallest structures are in the final image on the right of row A, which I have upsized here (with Gigapixel in standard mode):

 

The scale is 0.4um, so 400nm, and the structure repeats about half that. 

[Andrea B.]

I have been reading this topic with great interest.

 

(I cannot say that I know even a few beans about the theory so I should buckle down some day and learn a bit!)

 

Andy, great link. We see a lot of UV iridescence from flower petals and have discussed in the past the conical cells, scattering, etc.

[ulf]

Here is SEM results from a butterfly scale:

https://www.photomacrography.net/forum/viewtopic.php?t=24683&highlight=scale

14000x magnification for the last image!!!

 

Nature is really amazing.

[colinbm]

Wow, thank you for these amazing pictures & information.
It's another world down there.

[Lou Jost]

That's very useful. Jonathan has been resolving this sized structures using UV:

 

https://www.photomacrography.net/forum/viewtopic.php?f=14&t=44731

 

[ulf]

If anyone wonders how to get a SEM equipment and what it costs, there is one on eBay now.

It is the same type as used for the images in my link above.

https://www.ebay.com/itm/275476946354

US $85,000.00!

This is not the only thing needed for SEM imaging. You'll also need some small sputter, but I think that only cost maybe 1/5 of the sputter itself.

I have no up-to date information in that field though.

 

Most advanced scientific equipment also need a lot of training to get useful results.

 

I was not aware of that you could get a SEM that small and handy.

[dabateman]

There is a kid whom makes his own computer chips in his garage,  who bought a used scope for less.

If your really handy,  you can build one with a carefully removed electron gun from an old Crt monitor, some magnets and a nice stage with excellent cooling. There where some schematic published from University of Toronto. 

I maybe don't reget getting the notes from Dr. Ottensmeyer, when he retired from my graduate department.  He build a couple really fascinating electron microscopes. One design was sold to Zeiss. 

[OlDoinyo]

Sample preparation for electron microscopy is also a craft in itself, I understand. Sometimes the samples need to be coated with metal nanofilms in vacuo prior to use.