Talk:Polarizer

From Wikipedia, the free encyclopedia

WikiProject Physics This article is within the scope of WikiProject Physics, which collaborates on articles related to physics.
Portal:Physics
 Physics Portal
B This article has been rated as B-Class on the assessment scale.
Mid This article is on a subject of Mid importance within physics.

This article has been rated but has no comments. If appropriate, please review the article and leave comments here to identify the strengths and weaknesses of the article and what work it will need.

I'd like to store the underlying structure of the continuous text on this page, so that any discussion can point precisely.

Contents

[edit] Structure

[edit] Absorptive polarizers

Is the image (Wire-grid-polarizer.svg) correct? The wire grid seems to be 90 degrees out-of-phase with the emergent wave.—The preceding unsigned comment was added by Schneeman (talk • contribs) .

  • As it says in the article: "Note that the polarization direction is perpendicular to the wires; the naive concept of a wave "slipping through" the gaps between the wires is incorrect." --Bob Mellish 17:25, 28 June 2006 (UTC)
    • I just want to comment on this section. It is way to close to word for word from the book Optics by Eugene Hecht. Does anyone feel like rephrasing some of this?—The preceding unsigned comment was added by 67.189.113.62 (talk • contribs) .
  • blush - thanks!—The preceding unsigned comment was added by Schneeman (talk • contribs) .
  • It's easy to understand if you remember that a wire grid polarizer works because when the electric field of the wave is aligned with the wires, it makes the electrons in the wires move along their length, leading to absorption of the wave. If the e-field is perpendicular to the wires, the electrons can't move very far and there is much less absorption.--Srleffler 21:29, 28 June 2006 (UTC)

[edit] Beam-splitting polarizers

Beam splitters can sorted according to their shape (cube or plate) or according to the phyisical principle:

[edit] Use of birefringence

Materials

[edit] Principle

Snell's law and refraction The rays experience differing refractive indices in the crystal, depending on their (linear) polarization. Snell's law holds, but now produced the ordinary or o-ray and the extraordinary or e-ray.

total internal reflection The critical angle is different for o-rays. than for e-rays.

[edit] Types

Stop! The factual accuracy of this section is disputed.

A Nicol prism that consists of a crystal of calcite which has been split and rejoined with Canada balsam.

  • Refraction takes place at the surface.
  • Total internal reflection only of the o-ray occurs at the balsam interface, because the refractive index of refraction of Canada balsam lies between e and o.
  • Advantage: Nicol prisms produce a very high purity of polarized light.

A Glan-Thompson prism consists of a crystal of calcite which has been split and rejoined with Canada balsam.

  • no Refraction takes place at the surface, because the crystal axis lies in the surface and in the interface of the cut.
  • Total internal reflection only of the o-ray occurs at the balsam interface, because the refractive index of refraction of Canada balsam lies between e and o.
  • Advantage: produce a very high purity of polarized light.

A Glan-Foucault prism consists of a crystal of calcite which has been split and rejoined with Canada balsam.an air gap between the two halves. [[[User:Srleffler|Srleffler]] 21:07, 8 January 2006 (UTC)]

  • nolittle refraction takes place at the surface, because the crystal axis is in the surface but not in the interface of the cut.the surface is typically perpendicular to the incident light.
  • As the specific article states as least in the image, is that refraction takes only place upon the o-ray leaving the crystal--Arnero 20:45, 9 January 2006 (UTC)
    • This is true only if the light happens to be incident exactly perpendicular to the input surface. Polarizing prisms have some range of acceptance angle, however. They can accept an input beam that is not exactly perpendicular to the input face, and then of course there will be some refraction at the input surface. This is actually an important issue for this treatment, since one of the key practical differences between different types of polarizing prism is how large their acceptance angle is.--Srleffler 23:14, 9 January 2006 (UTC)
  • Total internal reflection only of the o-ray occurs at the balsam interface, because the refractive index of refraction of Canada balsam lies between e and o.
  • Advantage: produce a very high purity of polarized light.

A Glan-Taylor prism consists of a crystal of calcite which has been split and rejoined with an air gap between the two halves [ed. Srleffler 21:07, 8 January 2006 (UTC)].

  • noLittle Refraction takes place at the surface, because the crystal axis is in the surface and in the interface of the cut.the surface is typically perpendicular to the incident light.
  • As the image in the article shows that refraction only takes place upon the o-ray leaving the crystal
    • See above for why this is not exactly correct, in general--Srleffler 23:14, 9 January 2006 (UTC)
  • Total internal reflection only of the o-ray occurs at the balsam interface, because the refractive index of refraction of Canada balsam lies between e and o.
  • Advantage: Can be used at high laser powers
  • Disadvantage: More losses at the air gap for the transmitted beam, may need AR-coating [?]

A Wollaston prism consists of a two crystals of calcite which have been split and re joined with Canada balsam or without nothing in between, with their optical axes perpendicular.

  • no Refraction takes place at the surface, because the crystal axis lies in the surface and in the interface of the cut. But it takes place at the interface between the crystals.
  • As one can see on the specific arcticle refraction takes place only at the rear surfaces. --Arnero 20:45, 9 January 2006 (UTC)
    • See above for why this is not exactly correct in general.--Srleffler 23:14, 9 January 2006 (UTC)
  • no Total internal reflection takes place
  • divergence angle: 15°-45°
  • Advantage
    • low losses at cut
    • Polarization of both rays after separation is conserved

A Sénarmont prism consists of a two crystals of calcite which have been split and re joined with Canada balsam or without anything in between, with their optical axes perpendicular.--Srleffler 16:06, 10 January 2006 (UTC)

  • no Refraction takes place at the surface, because the crystal axis lies in the surface but not in the interface of the cut. But it takes place at the interface between the crystals.
    • This is not true in general. See above.--Srleffler 23:14, 9 January 2006 (UTC)
  • no Total internal reflection takes place
  • Advantage
    • low losses at cut
    • Polarization of both rays after separation is conserved

A Rochon prism consists of a two crystals of calcite which have been split and re joined with Canada balsam or without anything in between, with their optical axes perpendicular.--Srleffler 16:06, 10 January 2006 (UTC)

  • no Refraction takes place at the surface, because the crystal axis lies in the surface and in the interface of the cut. But it takes place at the interface between the crystals. The axis of the second crystal lies in the direction of the rays.
    • This is not true in general. See above.--Srleffler 23:14, 9 January 2006 (UTC)
  • no Total internal reflection takes place
  • Advantage: low losses at cut

A Foster prism is like a Glan-Thompson prism but adds another Total internal reflection to achieve:

  • divergence angle: 90°
  • Source

Some Glan-Thompson prism adds another Refraction to achieve:

  • divergence angle: 90°
  • disadvantage: dispersion

Some new design at a single surface use Total internal reflection to reflect both rays and then Refraction to split then.

  • divergence angle: 90°

This section is so badly flawed I'm not sure it is worth saving. It is riddled with factual errors. I've tried to mark them where I could, but there are more and I don't feel comfortable direct-editing it too much since this is a talk page. Beyond, that, though I'm not sure that this would be a good scheme for reorganizing the article. Right now, the article organizes the types of polarizers more or less chronologically by development. It would be useful to add a comparison of the merits of different types of polarizers, perhaps as a separate section, but I can't see replacing the nice explanation currently in the article with this ugly list of redundant bulleted points. --Srleffler 21:07, 8 January 2006 (UTC)

>redundant bulleted points

This is happening with the individual arcticles on the prisms. I wanted to prepare a merger.
I do not think you should attempt to merge these articles. The prism articles are very well written. The stuff above is a nightmare.
NB: Talk pages are not a good place to "hash out" text for a page, as you can see by what is happening. The conventions about never deleting other peoples' comments, and about signing every contribution on talk pages quickly reduce the proposed text to an unreadable mess. In the rare cases where it is necessary to hash out text "offline" from the real page, I have seen people create a "subpage" for the proposed text. I've never done this myself, though. --Srleffler 23:14, 9 January 2006 (UTC)

>chronologically

Absorptive polarizers is not chronologically nor is the separation into absorption and splitting nor is it standard. The history subsection may be better suited.
Nicol was first, followed by Brewseter's angle, says the 2nd edition of:


>factual errors

Sorry, I should have carefully looked up the factual (as I did not write them) articels on the individual prisms, thank you that you helped me. And the article does not support memorizing. --Arnero 20:45, 9 January 2006 (UTC)

>direct-editing

Maybe only two turns.

[edit] History

Nicol was first

[edit] Application

and were extensively used in microscopy

[edit] Use of Fresnel equations

[edit] Principle

Beam overlap leads to Interference. This can be used as an advantage to increase the intensity of the s-beam or reducing the number of plates or films.

Some math collected from the linked wiki articles:

  • Air. at 589.3 nm n= 1.0002926
  • Magnesium Fluoride: Refractive index at 500 nm: no = 1.37397, ne = 1.3916.
  • glass (typical) at 589.3 nm nn 1.5 to 1.9. BK7: 1.5164
  • Critical angle. \theta_c = \arcsin \left( \frac{n_2}{n_1} \right) n1 is the refractive index of the denser medium. For Air-MgF this is 46°.
  • Brewster's angle \theta_B = \arctan \left( \frac{n_2}{n_1} \right) n1 is the refractive index of the medium, were the angle is measured. For MgF-BK7 this is 47°. So it is just in the region of total internal reflection. For Air this angle is 36° inside the MgF.

[edit] Types

When choosing an angle of 40° inside MgF, reflections from the Air-MgF interface and the MgF-BK7 interface will interfere destructivly for p-polarization and constructivly for s-polarization. --Arnero 18:34, 8 January 2006 (UTC)

NB: Not all thin films are MgF.--Srleffler 21:13, 8 January 2006 (UTC)

But MgF was given as a typical example to me, especially for AR-coating. And as MgF has a low index of refraction it seems also suitable for thin film polarizers. The trouble is, I do not know the second material to build the stack with, so I used BK7, which is more typlically a substrate. --Arnero 20:49, 9 January 2006 (UTC)

Yes, MgF is very commonly used for cheap single-layer AR coatings, because of its low index. It is typically not used for multilayer coatings. A much more common structure for multilayer stacks is alternating layers of SiO2 and a higher index material. There are many choices for the high-index material. The coatings I order usually use tantala (Ta2O5) or hafnia (HfO2) but other materials are used for other applications. Coating design is quite complicated. There is a lot of knowledge that goes into the choice of the materials, the deposition methods, etc. --Srleffler 23:32, 9 January 2006 (UTC)

[edit] History of the discussion

[edit] Comments moved out of article:

  • Many more prisms have been invented, and it would be nice, if this arcticle would help the user to choose the right one for him : Fock Polarizer. User:Arnero 04:45, January 6, 2006
  • The big question is: What is the difference between a thin-film polarizer and a beam splitting dielectric mirror? Are the thin films birefringent? Then maybe a single film is sufficient, kind of inverser Nicol prism. Anyone who used a beam splitting dielectric mirror has observed, that it works like a polarizer. A .5 beam splitter in 45° may transmitt .7 p-polarization. What is the prove, that the internal Brewster's angle does not play a role? With the right dieelectric materials, this angle becomes 45°. Polarizers who are optimized for having a small stack may have to be used with anoter angle. User:Arnero 04:45, January 6, 2006
  • BIREFRINGENT THIN FILMS AND POLARIZING ELEMENTS The second part focuses on an emerging planar technology in which anisotropic microstructures are formed by oblique deposition in vacuum. User:Arnero 04:45, January 6, 2006

[edit] Discussion copied from Arnero's talk page

I think the comment about Brewster's angle in the context of thin film polarizers is at best confusing, and probably outright wrong. The response is mostly due to the variation in the interference in the film with angle of incidence rather than Brewster's angle. When you make a thin film polarizer, you aren't constrained to keep the angle at or even near Brewster's angle. Unlike Brewster's angle, the polarization can change dramatically with wavelength. Perhaps there are thin-film polarizers that do use an internal Brewster's angle reflection, but this certainly need not always be the case.--Srleffler 14:15, 30 December 2005 (UTC)

>variation in the interference in the film with angle of incidence
This does not distinguish between directions of polarization, only wavevector directions

>Perhaps there are thin-film polarizers that do use an internal Brewster's angle reflection, but this certainly need not always be the case
So why are you so certain?

>Unlike Brewster's angle, the polarization can change dramatically with wavelength
Yes thats a difference, but I think the only one. One usually uses thin film polarizers for narrow bandwidth lasers and there the compactness is of greater importance. --Arnero 16:07, 30 December 2005 (UTC)

Please don't put discussion or questions in the article pages. That kind of material belongs on the article's talk page.--Srleffler 13:47, 6 January 2006 (UTC)

According to Macleod's book (see ref. below), a MacNeille prism works the way I think you have in mind, with a thin-film stack designed such that light passes through a series of thin-film interfaces at Brewster's angle. These polarizers have a very wide spectral range, but limited angular range for obvious reasons. Plate polarizers cannot be made based on this principle, because "the Brewster angle for normal thin-film materials...is found to be greater than 90° referred to air as the incident medium. In other words, it is beyond the critical angle for the materials. [In the MacNeille prism] this is solved by building the multilayer filter into a glass prism so that the light can be incident on the multilayer at an angle greater than critical."

Plate polarizers, on the other hand, depend on the fact that the width of the high-reflectance zone of a quarter-wave stack is different for s and p polarization. There thus exists a region at the edge of the reflection band where the reflection is high for s-polarized light and low for p-polarized light. Further layers are used to smooth out ripples in the transmission spectrum, etc. The spectral range for these polarizers is narrow, and shifts with angle of incidence like any other bandpass coating.

While cube polarizers can be made based on the Brewster principle, they are often made using the same principle as plate polarizers. The latter tend to have a broader spectral range than plate polarizers, because of the higher angle of incidence of the light at the coating layer interfaces.

  • Macleod, H Angus (2001). Thin-Film Optical Filters, 3rd ed.. Institute of Physics, 362-368. ISBN 0750306882.

--Srleffler 14:42, 6 January 2006 (UTC)

[edit] Beam splitters

The section on beam splitting polarizers may need to be broken up in two. I have just been reminded that the Glan-type polarizers are not truly polarizing beamsplitters, even though they are often called that. The transmitted beam is 100% polarized. The reflected beam is not. (qv. Talk:Glan-Foucault prism). The Wollaston-type prisms are true polarizing beamsplitters, as are the dielectric type (plate and cube).--Srleffler 23:17, 10 January 2006 (UTC)

[edit] Circular Polarisers

Circular Polarisers should be added too to the article. helohe (talk) 13:47, 20 March 2006 (UTC)

I agree, especially since circular polarizer redirects to this page. I can't find much information about them, though. I know what circular polarization is (see polarization), but I don't see how rotating a circular polarizer would make any difference. I've heard that you want to use a circular polarizer for any modern camera that does through-the-lense focus or metering. One site, [1], says that such metering uses semi-silvered mirrors to extract light from the optical path. Doesn't metallic reflection not change polarization? Don't digital non-SLR cameras do focus and metering by reading the CCD?
The thing that photographers call a "circular polarizer" seems to actually select a single linear polarization, and then make that circularly polarized. The effect you want for improving picture quality is linear polarization. The additional step of making the output of the filter circularly-polarized is just to make sure that the viewfinder and metering electronics in an SLR camera see the same thing as the film or CCD will. A semi-silvered mirror at an angle will reflect some polarizations more than others. With an SLR and a linear polarizer, this means that you (and the electronics) would see a brighter or dimmer image than the film will see. I wouldn't think that a non-SLR digital camera would be affected.
So:
  1. Rotating it makes a difference because it's a linear polarizer with circular output
  2. Metallic reflection does change polarization, because the (angled) mirror will reflect some polarizations more than others.
  3. You might be right about digital non-SLR cameras.
I agree that it should be added to the article, but that needs to be done by someone who knows how it works. It sounds like a linear filter followed by a quarter-wave plate, but that would produce horrible chromatic aberration. I can't imagine how one would make one that is suitable for colour photography.--Srleffler 04:59, 13 November 2006 (UTC)
Retrieved from "http://en.wikipedia.org../../../p/o/l/Talk%7EPolarizer_49f7.html"