Switching polarizer and analyzer: the order matters -- why?

[PathEnthusiast]

TL;DR

I was playing around with the polarizer and analyzer on my microscope and discovered that the analyzer could block light from the polarizer--but not vice versa. I'm perplexed.

Hey all! I'm a pathologist and just got a polarizer/analyzer pair for my microscope. I decided to play around with them a little before I put them into the microscope to demonstrate the way they block light, and initially I found that no light was blocked no matter how I rotated them relative to each other. This was quite perplexing, and eventually, as I fiddled with them, I changed the order light was passing through and lo and behold light was blocked as expected when they were perpendicularly oriented. After some further experimentation, I have observed the following:

1) Light from an overhead fluorescent in my office is blocked when it passes through the polarizer and then the analyzer (when they're oriented perpendicularly).
2) Light from the overhead fluorescent is NOT blocked when light passes through the analyzer and then the polarizer (regardless of how they're oriented).
3) Light from my computer monitor is blocked by the polarizer alone (when appropriately rotated).
4) Light from my computer monitor is blocked by the analyzer alone (when appropriately rotated).

Based on these observations, I've concluded that the analyzer and polarizer are doing *something* different to light, but I'm perplexed as to what exactly.

My suspicion is that, when exposed to randomly polarized light, the polarizer blocks all directions of polarization but one, thus producing linearly polarized light. The polarizer will thus block all light polarized perpendicular to this direction.

On the other hand, I suspect that the analyzer is birefringent and is somehow producing circularly or elliptically polarized light. Such light would not be blocked by a linear polarizer. However, I'm at a loss to explain how a birefringent material is able to entirely block all light polarized in a particular direction. This seems like I've forgotten a really basic aspect of how circular polarizers work, but quick attempts at refreshing my memory reveal only that circular polarizers can cause destructive interference with light reflected off a surface that has the opposite direction of circular polarization. And I don't see how that helps explain the phenomena I'm observing.

It's been over 12 years since my last physics class, so I'm quite rusty and this is as far as I've been able to get on my own. Please help!

 

[DaveE]

Probably a birefringent film on one side of the polarizer. Play around with the combinations of order in the stack and flipping each polarizer. Any combination that has two crossed polarizers with nothing in between will block light. But if the polarization is changed in between them, they won't.
This is how circular polarizers are made.

 

[PathEnthusiast]

Probably a birefringent film on one side of the polarizer. Play around with the combinations of order in the stack and flipping each polarizer. Any combination that has two crossed polarizers with nothing in between will block light. But if the polarization is changed in between them, they won't.
This is how circular polarizers are made.

YES!!!

It hadn't occurred to me that they would coat just one side with birefringent material. I now have some additional observations:

1) Light from the computer monitor is not blocked by the analyzer when the "top" of the analyzer faces the computer screen.
2) Light from the overhead fluorescent is not blocked when light passes through the polarizer and then the analyzer, if the "top" of the analyzer faces the polarizer.
3) Light from the overhead fluorescent IS blocked when light passes through the analyzer when the "top" faces the fluorescent light and then passes through the polarizer.

So my explanation of the phenomenon would be:

The birefringent film on the analyzer converts linearly polarized light into "rotating" polarized light that cannot be entirely blocked by a linear polarizer. However, if linearly polarized light passes through the polarizing material of the analyzer *before* reaching the birefringent film, essentially all light has been eliminated so rotating the polarization doesn't make a difference. Similarly, if light passes through the film first, then the polarizing material of the analyzer, the polarizing material converts it to linearly polarized light which can then be blocked by the second polarizer.

 

[DaveE]

Sometimes you can peel the film off if you want.

 

[sophiecentaur]

TL;DR Summary: I was playing around with the polarizer and analyzer on my microscope and discovered that the analyzer could block light from the polarizer--but not vice versa. I'm perplexed.

My suspicion is that, when exposed to randomly polarized light, the polarizer blocks all directions of polarization but one, thus producing linearly polarized light.

This is not an accurate description of the process. what your description implies is that you would only get a tiny fraction of the incoming light through. In fact the polariser takes the component of all of the randomly polarised parts of the light, resulting in half the energy getting through.

When linearly polarised light hits a circular polariser, the light becomes circularly polarised (obvs) and half its power gets through a linear analyser with any orientation. When circularly polarised light passes through a linear polariser, only one particular component of the resulting linearly polarised light is selected for one orientation of the analyser. It's blocked when the polarisations are crossed. There's a logic to it.

It hadn't occurred to me that they would coat just one side with birefringent material.

A circular polariser has, by design , a thickness such that the path difference through the birefringent material is one quarter wave different for the two linear polarisations. That means it is a thin layer and has to be mounted on one side of the supporting linear polariser. The direction of rotation depends on the orientation of the linear polariser. There's a diagram in this link of what I mean.