Positive Phase Contrast Objective Phase Plates

[microb]

For positive phase contrast, the objective has a two layer ring, Figure 3a, that dims and phase shifts.

The phase shift of a quarter wave confuses me. A ¼ waveplate needs an orientation with polarized light, but phase contrast is not using polarized light. Typically it uses a single frequency, like a green wave length.

So what exactly are the two layers used for the phase plate?

https://micro.magnet.fsu.edu/primer/tec ... phase.html

[Greg Howald]

Sorry. I don't know the answer to your question. I hope someone can answer it. The article you referred to didn't seem to give a very definitive answer either. I just got my phase turret. I'm enjoying it very much, but am also in the midst of experimentation with it. What happens if you simultaneously use cross polarization or singular polarization? What happens if you add color to the center stop to also add a reinburg effect?
Phase contrast offers a vast variety of imaging possibilities including mixing of rings and objective sizes for different affects, all of which is both interesting and scientific.

[hans]

I believe 1/4 wave is referring to different things in the two contexts. For the phase plates, 1/4 wavelength of optical path difference between direct and diffracted independent of polarization. For example a 250 nm difference in glass thickness of the phase ring would give 125 nm optical path difference (250 nm * (1.5 - 1)) which is 1/4 wave at 500 nm. In the context of birefringent waveplates it is the OPD for ordinary vs. extraordinary polarizations but the OPD for O and E separately vs. air can be much larger. For example weakly-birefringent polycarbonate as discussed recently in Louise's Sanderson prism thread could require several mm to get to 1/4 wave retardation O vs. E but at that thickness OPD vs. air for either O or E is huge, thousands of wavelengths at 500 nm.

[BramHuntingNematodes]

The phase difference is used to create destructive or constructive interference with light travelling through or around the sample. The sample is also assumed to rrtard light by about a quarter wave.

Keep in mind the only light that will not pass through the phase ring is that which is scattered off the sample.

[BramHuntingNematodes]

I should dig up the article but early in there were experiments with many degrees of phase plates, with certain degrees of retardation useful for different kinds of samples.

[hans]

This book has a lot of interesting stuff in it.

The sample is also assumed to rrtard light by about a quarter wave.

I haven't yet managed to understand the arguments in Bennett's book for why this is true and the "Molecular Expressions" primer and related manufacturer-specific sites don't really explain it in detail. Would be nice to find something in between.

[microb]

Any pointers to example materials used for the ring in the phase plate would be appreciated.

[microb]

If one puts two orthogonal quarter waveplates together, would assuming a linear additive effect be valid in making a 1/4 wave retarder regardless of the polarized state of the light? I haven't figured out what the Jones matrices would be. Need to look that up.

[microb]

Thorlabs doesn't say what the material is for the ring:

https://www.thorlabs.com/newgrouppage9. ... p_id=11083

[hans]

If one puts two orthogonal quarter waveplates together, would assuming a linear additive effect be valid in making a 1/4 wave retarder regardless of the polarized state of the light?

I think no... the 90 degree rotation swaps the fast and slow axes which is equivalent to inverting the sign of the birefringence, so the effect of the two plates cancel out no matter what the initial polarization state is?

[patta]

Agree with Hans, the "phase plates" of polarization and Zernike are different things.

How a Zernike plate is made?
It is a flat plate with some areas with 250 nm or like higher relief. Many ways to manufacture it, all related to photolithography:
1) carved or etched from a flat glass; so the plate is 100% glass, but with those tiny thickness differences.
2) The relief can be painted, by applying a thin layer of transparent photoresist where you want the step.
3) The relief is made on plastic by molding, the mold from 1).
The method 1 may be somehow feasible at home.
I couldn't find which method is most commonly used for microscope objectives. Would guess 1) in the 1960s and 3) today.

Another link on phase:
https://www.photometrics.com/learn/ ... roscopy

Edit: ancient DIY phase method; the "step" and the light attenuation annulus is made by depositing candle sooth over glass. The phase material is the sooth.
https://www.micromagus.net/microscop ... scn.html

[BramHuntingNematodes]

If you are very careful you can cut a ring from plastic phase material and sandwich it in glass.

[hans]

If one puts two orthogonal quarter waveplates together, would assuming a linear additive effect be valid in making a 1/4 wave retarder regardless of the polarized state of the light? I haven't figured out what the Jones matrices would be. Need to look that up.

Have never used Jones calculus and wouldn't claim to understand the derivation but it looks like Wikipedia has the matrices for this case and the product is the identity matrix:

[patta]

Yes two quartz plate placed perpendicular, do no effect to polarized light.
The first retard one polarization direction by 1/4 (w.r.t the other direction); the other retards by 1/4 in the other direction (w.r.t the first); in total the two direction are retarded the same, or, relative to each other, they are the same, so no effect.
Note that this is RELATIVE retard, relative between the two polarization directions. This is due to the birefrengence of quartz.

Instead the plate of the phase objectives, the annulus retards by 1/4 with respect to the light passing outside the annulus. This is due to the refractive index of glass, that is a bit thicker at the annulus, against the refractive index of air. no birifrengence.

[microb]

If you are very careful you can cut a ring from plastic phase material and sandwich it in glass.

But that's what led to my question.

If I put a ring of film-1/4-waveplate into the Fourier conjugate plane of an objective, then that just retards the light in that polarized plane. All the other light would basically not be effected. Well I guess there is a non-linear fall-off as the angle goes from 0 to 90. and 1 degree to 89 degrees is semi-effected.

[microb]

Instead the plate of the phase objectives, the annulus retards by 1/4 with respect to the light passing outside the annulus. This is due to the refractive index of glass, that is a bit thicker at the annulus, against the refractive index of air. no birifrengence.

So correct me if I'm wrong or mixing things up: traditional phase contrast is positive contrast which is Figure3a shown above. So the 1/4 retardation is that 50% is the ring. And that ring is some type of smokey layer that Zernike mentioned in his development paper someone posted recently but I can't find right now.

[hans]

If I put a ring of film-1/4-waveplate into the Fourier conjugate plane of an objective, then that just retards the light in that polarized plane. All the other light would basically not be effected.

In this case the ring will retard direct-ordinary relative to direct-extraordinary by 1/4 wave. But it will also retard both direct-O and direct-E by a large amount (maybe hundreds or thousands of wavelengths) relative to diffracted. I think this is not what you want for phase contrast -- the retardance direct-O vs. direct-E is not desirable and the retardances direct-O vs. diffracted and direct-E vs. diffracted are way too large.

[microb]

If I put a ring of film-1/4-waveplate into the Fourier conjugate plane of an objective, then that just retards the light in that polarized plane. All the other light would basically not be effected.

In this case the ring will retard direct-ordinary relative to direct-extraordinary by 1/4 wave. But it will also retard both direct-O and direct-E by a large amount (maybe hundreds or thousands of wavelengths) relative to diffracted. I think this is not what you want for phase contrast -- the retardance direct-O vs. direct-E is not desirable and the retardances direct-O vs. diffracted and direct-E vs. diffracted are way too large.

From what I understand, a PH objective has a ring inside it that is basically at that objective's back focal plane. That back focal plane is a Fourier conjugate plane from what I understand. That ring is documented as being 50% transmission and a 1/4 wave retarder. That "1/4 wave retarder" part confuses me, since I can't figure out a material for that which is a thin smokey-grey film typically glued into these objectives.

So I'm wondering what is that dark 50% transmission ring made of that adds a 1/4 retardation to light coming from a sample that has supposedly retarded up to a max of 1/4 wave already, leading to according to Dr. Z. from his invention an accumulated max 1/2 wave retardation that causes interference leading to image contrast that an observer sees?

Is it just a thin film with 50% transmission and a larger index of refraction that the light is just slower travelling through? But the problem with that answer (and maybe I just don't have the right marketing label to search on) I don't see Edmund optics, Newport Optics, or Thorlabs selling a 1/4 wave retarder film that isn't a 1/4 waveplate with slow and fast axis to complicate things. I would assume that it has to be a very particular thickness tied to the material's index of refraction to make it this general 1/4 wave retardation film for unpolarized light. And that seems like a good product for someone to be selling. But I haven't seen it for sale anywhere. Maybe they do sell it, and I just haven't found the right name to google yet.

[BramHuntingNematodes]

The amplitude diminishing part of the plate is described by Zernike as "a thin metallic deposit" or by others as "dielectric film." This is treated separately from the "phase strip" which is a grooved taken out of the glass of the plate.

I have been experimenting with 1/4 wave film to see if it can produce the effect, however my attempts have been stymied by internal reflections creating a general loss of contrast paired with no satisfactory dielectric film to use as the amplitude dampening material. I should get back to it, but it requires a bit of patience.

[BramHuntingNematodes]

I see what you mean though, hans. I would also need a way to to get the relative phase shift to translate into amplitude differences. Might have to work this out again.

[microb]

Edit: ancient DIY phase method; the "step" and the light attenuation annulus is made by depositing candle sooth over glass. The phase material is the sooth.
https://www.micromagus.net/microscop ... scn.html

Yes, that's the smokey thing I remember reading. I was looking for the web page but couldn't find it. Thanks.

[microb]

Here's a page where someone tried the candle soot option:

" I tried using a circular cover slip, applying soot from a candle and and scraping it off to give a stripe, but found this difficult to do in a controlled fashion. I then tried masking the coverslip with adhesive tape before sooting, and this worked quite well, although one needs to select a tape that leaves minimal adhesive residue once removed. The substage stop was made by masking a transparent disc with PVC tape to leave a stripe that aligned just inside the phase plate stripe when the objective back-lens was viewed using a pinhole eyepiece."

https://www.micromagus.net/microscopes/diyphase.html

[hans]

That "1/4 wave retarder" part confuses me, since I can't figure out a material for that which is a thin smokey-grey film typically glued into these objectives.

I haven't done much searching on how phase objectives are made but seems likely it would need to be a deposition or etching process directly on the glass, as patta is saying, not a separately-manufactured film material. As mentioned earlier, if the difference in refractive index is around 0.5 the thickness difference to give 1/4 wave OPD at 550 nm is around 250 nm or around 40 times thinner than typical store-bought plastic cling wrap. Handling could be difficult. And it seems like the thickness variation would need to be kept within similar tolerances as are held on the rest of the optical surfaces in the objective in order to not interfere with imaging performance?

I would also need a way to to get the relative phase shift to translate into amplitude differences.

If you cut the ring but leave all the surrounding material in place and just rotate the ring 90 degrees, then use linearly-polarized light parallel to one of the optic axes, seems like that should give the desired 1/4 wave shift direct vs. diffracted. But maybe there would be problems related to how the polarized light interacts with the sample or preceding optics. Possibly this is related to how that AO Polanret system works?

[BramHuntingNematodes]

Hm and turning the polarizer would, well that would be a worthwhile experiment. Maybe I will cut a center strip first as that's easier to pull of. Have two layers of film at 90 degrees when cutting.

[patta]

Yes yes cut the ring from a 1/4 plate and rotate it, that will retard 1/4 for one polarization; rotating the polarizer, 0 retard at 45°; -1/4 for the perpendicular direction.
Sounds a very nice trick.

It will still need the attenuation; maybe we can sort it out again with bare polarization.

The "candle sooth" does both 1/4 retarding and 50% attenuation, but that is sheer luck, and probably doesn't work so well. Normally, as written above, two layers are needed, one glass step (1/4 retard) and a thin metal film (50% attenuation).

Similarly to the "rotated polarizer"
I was thinking to make this phase plate as a "mounted slide", with two mountants. The ring is made of a resin with slightly higher refractive index; maybe mixed with little black pigment to do the attenuation. The rest filled with transparent resin with slightly lower RI.
Since the difference in RI is small, we can make it thick, manageable.

[hans]

Hm and turning the polarizer would, well that would be a worthwhile experiment.

At 45 degrees to the optic axes that is the usual circular polarizer case so I think you end up with left/right-handed circular for direct/diffracted, swapping as you add 90 degree increments to the initial polarization direction. Not sure what implication that has for interference, maybe then a final analyzer would be necessary to actually get interference as in DIC?

[patta]

I've looked now at the AO Polanret; it really looks like a full implementation of the "rotated annulus"; with the phase contrast obtained (and adjustable!) by birifrengence plates & polarizers.
The "rotated annulus" is likely the "polazone plates" in the image.
At the link, he writes that was outrageously expensive
http://bettermicroscopy.blogspot.com/2 ... ml?m=1

[hans]

Two patents, overall arrangement doesn't look quite the same as Polanret, but maybe related:

2,516,905 - Microscope with variable polarizing and birefringent means for producing contrast in optical images
2,700,918 - Microscope with variable means for increasing the visibility of optical images

[microb]

2,516,905 - Microscope with variable polarizing and birefringent means for producing contrast in optical images
2,700,918 - Microscope with variable means for increasing the visibility of optical images

These are cool.

So instead of using masks, they used rings of different polarizations and a compensator to pick up the difference.

Why are there no modern systems using this? Is it dimmer than PH masks?

[hans]

Why are there no modern systems using this? Is it dimmer than PH masks?

From the blog patta linked, sounds like it was just too expensive for not enough practical benefit. I wonder if polymer polarizing filters and waveplates would have helped? I don't know what was available in the ~50s-70s. Then of course there was also competition from DIC.