Looking “inside” a silkworm using Frustrated Total Internal Reflection

[purkinje]

Hello Ramon, Holger and Jürgen,

Interestingly, the question of these color-shift phenomena had already been intensively discussed at the beginning of the 20th century. Here, too, the issue concerned their misinterpretation as fluorescence phenomena. Organic structures stained with certain dyes such as methylene blue, fuchsin, or eosin appear orange, blue-green, and green under dark-field illumination. Natural pigments as well, such as those of Stentor coeruleus, appear blue-green in brightfield microscopy and pink in darkfield microscopy: Stentor coeruleus ein Porträt
This phenomenon, referred to as selective diffraction, or later as anomalous dispersion, was described in detail by Berek already in 1921. It was of particular importance because researchers sought to understand the color phenomena observed in weakly stained bacteria under darkfield illumination.

Best Regards Stefan

Anomalous Dispersion in Ocean Optics

M. Berek, Über selektive Beugung im Dunkelfeld und farbige Dunkelfeldbeleuchtung, Z wiss Mik 38, S 237ff.

[Spectrum]

Hello Stefan and Jürgen,
I agree,  the colour shift pheenomenon is definetely a thing, worth considering, in this case.
But what brought me to flourescence, as (maybe only one of several) exlanations, of what we see here, is:
1. Why do we see the "green stuff" only at some, but not all, reddish areas?
2. Why do we find these odd colour prominently at exactly the parts, where you can observe the stongest fluorescence of fern specimen, under "conventional" Epi fluorescence?

...and the green resembles well with what we see with fern flou...

Lest to forget about the blueish background in this picture.🤔

I think a comparison picture, taken in "classic" blue light Epi Fluo could be very helpful.

Sounds complicated, and unfortunately these things are complicated, but hopefully it helps at least a little bit for your understanding.
Jurgen

 

I agree!
But very interesting topic, isn't it.

Greetings Holger

[jcs]

Hi Stefan,
very interesting paper, thanks for digging that out! Nowadays, physicists use the complex refractive index to describe various optical phenomena. From this complex refractive index, absorption and reflection can be derived. That's why I had to struggle with Berek's language, and I am not sure if I understodd his key points correctly.

Reading the key message on page 245 "... sondern die Farbwirkungen resultieren lediglich nach Maßgabe der Brechungsindizes und Absorptionsindizes an den Grenzflächen der beugenden Elemente gegen die Umgebung" he uses refractive index and absorption coefficient as material constants. When instead using reflective and absorption coefficient as material constants, you can come to similar (or even the same) conclusions.

But the important conclusion from this paper is, that you need neither frustrated total internal reflection nor fluorescence to explain these color changes that can be observed in darkfield microscopy. At least that's how I understand it at the moment.

Jurgen

[purkinje]

Hello Holger,
i agree with your concerns, too. So eating the pudding will be the proof of the concept: but if the reddish stain is eosine, than classical Fluorescence will show us bright green everywhere, cause unfortunately eosine shows fluorescence as well as anomalous dispersion. And dispersion of the stains is depending on the biological stained matrices as well, too.
A conventional darkfield would be of interest, may be it looks different to Ramons technique, may be not.
After that an unstained fern preparation in Ramons technique, classical darkfield and of course fluorescence epi will show the real situation.

Hello Jürgen,
fortunately Berek was a physicist and not a medical doctor of these days, so his language was not so pretentious and cloudy   

best regards Stefan

[Ramon]

Hello Jürgen and Holger

This resembles mostly some kind of darkfield illumination.
Most of the light raus are hitting the specimen from a flat angle from all around the sides.

It is indeed a kind of Darkfield but in this FTIR device we have index-matching (immersion oil) so that grazing rays that would be mostly reflected by the slide in standard Darkfield in FTIR, since it is index-matched, we have no reflection and all the light will hit the specimen.

But why does the fern show green colour?
Hm, the sporangium of ferns show strong fluorescence...

Hello Holger,

Yes, the fluorescence hypotheses was proposed by some people at the Photomacrography.net community. We were lucky that one of the moderators has an epi fluorescence microscope and lives not far from my place. I sent him the fern leaf sample and he took the images with the epi fluorescence microscope. You can find the images at:

https://www.photomacrography.net/forum/viewtopic.php?f=14&t=49734

His conclusion is that we have a lot of fluorescence in this sample.

There is apart of the darkfield also some fluorescence happening. That, so my guess, is what causes the "green stuff".

I think you are right. As you read, we have proof that fluorescence is happening here.

Perhaps (no bet on that!) the green fluorescence is also amplified by some effect also used in a technique called "TIRf".
In TIRF illumination you use total reflection in order to illuminate only a narrow layer of the probe with the exiting wavelengh for fluorescence.

I don't know about that. In TIRf light does not get through, only the evanescent wave.

The frustrated total internainternal reflection is an effekt, that similar with TIRF but with white light and with no fluorescence effects. It also only occurs very close to the layer facing the illuminated surface. And I mean very close.
It is used for fingerprint sensors or touchscreens.

I took the idea for this device from backlight panels for screens. FTIR is also used in the rain sensors of cars. However, I found out that lots of FTIR research papers don't use any "connecting media" (I use immersion oil as connecting media) but instead they use mechanical pressure.

The first (left) picture looks like some common brightfield image.

It is indeed Brightfield. I use an old biological microscope with brightfield condenser (not a Kohler).

Hello Jürgen,

The same leaf will also appear green in reflection, although the reflection coefficient might have a very different spectrum compared to the absorption coefficient. The reason for that is the fact that the reflection of such an object under vertical illummination is rather weak. Instead, the light penetrates the leaf and part of it get absorbed by the chlorophyyl. Some of that light gets scattered back, leaves the object and is observed as green light in reflected light microscopy.

The AI (Claude) agrees with you. I prompted it with you initial comment: "in my view, the complementary colours you see in brightfield vs. your glazing incidence illumination are caused by the relation between absorption coefficient and reflection coefficient. In physics, these relations are dealt with by the so-called Kramers-Kronig relation: If you know the absorbance of a material over a broad wavelength range, you can calculate the reflection from that."
Claude added a consideration that I think is appropriate. The stain used in most commercial slides, Safranin. This dye has a strong absorbance at 500-510 nm. Below see absorbance graph for different concentrations of Safranin T.



According to the AI what is happening is as follows:

Safranin is highly selective when binding to different tissues. It binds specially to lignified and cutinized cell walls. Xylem, sclerenchyma, annulus cells, epidermis/cuticle (lignified or cutinized walls).

In Brightfield light passes through the specimen from below. Safranin has a strong absorption peak at approximately 510-530 nm (green wavelengths). The tissue absorbs green and some blue. What arrives at the objective is white light minus green, which appears pink/red. This is a pure transmission color - the color that was not absorbed by the stain.

For safranin-stained tissue: k (absorbance coefficient) is large at green wavelengths (510-530 nm). By Kramers-Kronig, n also changes anomalously near green wavelengths. The result is that the stained tissue behaves in a more metallic manner at green wavelengths - and metals reflect strongly. This is sometimes called anomalous surface reflection.

Light hits internal structures (annulus cell walls, xylem) from the side. Some is scattered upward into the objective. The scattered light carries the surface optical signature of those structures. Safranin strongly absorbs green (510-530 nm). By Kramers-Kronig, n changes anomalously at green wavelengths, causing preferential surface scattering of green (anomalous reflection). Green-enriched scattered light reaches the objective and forms the image.

I hope this helps.

You can view all the conversation with the AI at this link:
https://drive.google.com/file/d/1KICGh9jXW6ozjlgVIAi6bT5kQPfrVOVl/view?usp=sharing

Regarding FTIR, Holger pointed out a very important aspect: The "F" in FTIR is related to the evanescent waves leaving the surface of the light guide. The depth of these evanescent waves is very small, typically less than 300nm. This means that only features which are in extreme proximity to your light guiding plate can "frustrate" the light. In your fern sample, most features of the fern are probably much further away from the surface, and this needs to be considered when discussing the phenomena taking place in your device.

Aren't you referring here to TIRF rather than FTIR? In this device all the surface of the specimen is contacted to the light guide by a film of immersion oil. We have no evanescence wave at the contact point because we index-matched the interface.

Finally. With opaque inorganic specimens we also have color differences when using this FTIR little device. Below see an Integrated Circuit lithography imaged with Epi illumination, FTIR and "Wrap Around" Diffuser. The band that in FTIR is blue in the other techniques has a different color.



Is it that the surface of the lithography acts like a diffraction grating? Or is it the there is a thin film on top of the specimen? Or is it something else?

Pls., allow me to gather all the info on this issue and I will bring it to you. I hope you find it interesting.

Cya
Ramon

[jcs]

For safranin-stained tissue: k (absorbance coefficient) is large at green wavelengths (510-530 nm). By Kramers-Kronig, n also changes anomalously near green wavelengths. The result is that the stained tissue behaves in a more metallic manner at green wavelengths - and metals reflect strongly. This is sometimes called anomalous surface reflection.

Light hits internal structures (annulus cell walls, xylem) from the side. Some is scattered upward into the objective. The scattered light carries the surface optical signature of those structures. Safranin strongly absorbs green (510-530 nm). By Kramers-Kronig, n changes anomalously at green wavelengths, causing preferential surface scattering of green (anomalous reflection). Green-enriched scattered light reaches the objective and forms the image.

As mentioned, things are complex in this field. But in my understanding the Berek-paper, Claude and me are somehow saying the same thing, albeit in different language. Assuming that "normal dispersion" refers to the dispersion of glass (gradual decrease of n from blue to red), "anomalous dispersion" refers to materials with differing spectral dispersion. Stains exhibit such behaviour, with very pronounced absorption peaks at energies (wavelengths) where electronic transitions happen. According to Kramers-Kronig, these absorption peaks influence refractive index/reflection coefficient, which in consequence becomes visible under glazing illumination.

To discriminate this phenomenon from fluorescence, you could illuminate your fern sample with green or yellow light. If the green color is still there, fluorescence can be ruled out: Green fluorescence can only be triggered by shorter wavelengths (Stokes shift), e.g. blue or UV.

Aren't you referring here to TIRF rather than FTIR? In this device all the surface of the specimen is contacted to the light guide by a film of immersion oil. We have no evanescence wave at the contact point because we index-matched the interface.

I understood that you are using immersion oil. But I guess your fern-sample is still covered with a cover glass?

Jurgen

[Ramon]

Hi Jurgen

As mentioned, things are complex in this field. But in my understanding the Berek-paper, Claude and me are somehow saying the same thing, albeit in different language. Assuming that "normal dispersion" refers to the dispersion of glass (gradual decrease of n from blue to red), "anomalous dispersion" refers to materials with differing spectral dispersion. Stains exhibit such behaviour, with very pronounced absorption peaks at energies (wavelengths) where electronic transitions happen. According to Kramers-Kronig, these absorption peaks influence refractive index/reflection coefficient, which in consequence becomes visible under glazing illumination.

There are many new terms for me here. I never went so deep in the understanding of light-matter interactions. Need time to study it.

To discriminate this phenomenon from fluorescence, you could illuminate your fern sample with green or yellow light. If the green color is still there, fluorescence can be ruled out: Green fluorescence can only be triggered by shorter wavelengths (Stokes shift), e.g. blue or UV.

That can be done easiy.

Have you seen the post I linked? https://www.photomacrography.net/forum/viewtopic.php?f=14&t=49734

I understood that you are using immersion oil. But I guess your fern-sample is still covered with a cover glass?

Yes.

Ramon

[Spectrum]

Hello Ramon,
Thank's for pointing me to this link:

https://www.photomacrography.net/forum/viewtopic.php?f=14&t=49734


At first glance that looks just exactly like the proof of the pudding Stefan was pointing at.
But the more I watch the fluorescence pictures from Pau, the more I'm getting nonplussed...

Yes, the emittance in his first picture shows light emission exactly at the regions, we see the odd "green stuff"
But something is off here...🤔

Ramon, do you have any spectral data of the LED's you use?

Best regards Holger

[Ramon]

Alo Holger

At first glance that looks just exactly like the proof of the pudding Stefan was pointing at.
But the more I watch the fluorescence pictures from Pau, the more I'm getting nonplussed...

Yes, the emittance in his first picture shows light emission exactly at the regions, we see the odd "green stuff"
But something is off here...🤔

Pls., share with us what do you think that does not fit.

Ramon, do you have any spectral data of the LED's you use?

I am using: Cree® XLamp® XP-E High-Efficiency White LEDs. Find the PDF at the following link:

https://drive.google.com/file/d/1qmXnVD3tPFycoHtnLBwbxbapkC0WwldK/view?usp=sharing


Ramon

[Spectrum]

Hallo Ramos,

Alo Holger

Yes, the emittance in his first picture shows light emission exactly at the regions, we see the odd "green stuff"
But something is off here...🤔

Pls., share with us what do you think that does not fit.

Hello,
The odd green is very bright in relation to the overall brightness of the picture. Much more than you would expect when you compare it with the Epi Fluo pictures from Pau.
And it's not only the "relative brightness" colour that puzzles me.
It's also the colour itself.
There is nearly no UV present in the light of your LED's.
So the exitation is mostly done by blue light, which is quite dominant  in "white light" emitters:


That' more or less the spektral content of such a LED. Your data sheet looks similar.
So most of your exitation is done by this blue wavelenghs.

The pictures where Pau uses blue light in classic epi fluorescence or Darkfield fluorescence with blue light just didn't blend with the "green stuff" we see in your picture with your device.

So fluorescence: yes, present. But....

The kind of fluorescence that would explain the bright green at it's given location, with no other green around: no

At least not with fluorescence as the only explanation.

And last but not least the other picture with some "odd green" that you posted, this time a "moss slide":

Here the brightfield equivalent:


That's definetely more than odd, to my eyes...
Don't know what's going on there...
Very strange!

How thick is the acrilic layer you are taking the photos through?
Greetings Holger

[Ramon]

That's definetely more than odd, to my eyes...
Don't know what's going on there...
Very strange!

I am quite loss here. I dont have the proper knowledge to interpret all this. I am trying to catch up on all the theory that you guys are proposing. Like Anomalous Dispersion that Stefan comented few posts up.

How thick is the acrilic layer you are taking the photos through?

The acrylic layer is 2.8mm thick in the center. BUT! In this photos there is not acrylic layer between the specimen and the objective lens. The light guide is mounted below the specimen slide. I attach a draw:




When the device is used with opaque samples it has to be mounted on top of the specimen. Between specimen and objective lens. When used with transparent samples you can use both configurations (top and bottom). But this photos were taken with the slide on top of the acrylic light guide.

It seems that misteries keep piling up.

Ramon