Elliptically polarized light & partially-polarized light

[peace]

How to distinguish elliptical polarized light from partially-polarized light?

 

[sophiecentaur]

How would you define partially polarised light? Do you mean a mixture of polarised and unpolarised ? If so, which polarisation - linear or circular polarised?
I suspect that a single, linear polariser would show the same result as you rotate it but I think you would need a combination of circular polariser and linear polariser to identify elliptical polarisation for certain.

This polarisation business is far easier to get to understand in the context of RF antennae and waves because you have much easier control of the source. I guess there are Optics Guys who might disagree, though. (But wadda they know?)

 

[peace]

How would you define partially polarised light? Do you mean a mixture of polarised and unpolarised ? If so, which polarisation - linear or circular polarised?
I suspect that a single, linear polariser would show the same result as you rotate it but I think you would need a combination of circular polariser and linear polariser to identify elliptical polarisation for certain.

This polarisation business is far easier to get to understand in the context of RF antennae and waves because you have much easier control of the source. I guess there are Optics Guys who might disagree, though. (But wadda they know?)

Yes of course. I mean is a mixture of polarised and unpolarised.
Suppose it is a mixture of linear polarized light with unpolarised light.
Now the question is how to detect this from the elliptic polarized light.

 

[sophiecentaur]

Yes of course. I mean is a mixture of polarised and unpolarised.

If you mean plane polarised then the word is plane polarised and not "of course" polarised. I had to ask and you needed to specify your question properly.
You need to test for both forms of polarisation and there's a logic to the process which I think makes sense. The answer to the following would be reliable, I think.
Presumably a linear polariser is showing different throughput at different angles(?).
Elliptical polarisation can be looked on as a mix of circular and linear polarisation with the two components co-phased (of course).
The circular polariser would (when the right way round) eliminate any circular component. If it doesn't change the level then there is no circular component. So that would, I guess, tell you that it's not elliptical, which could be enough for you.

But for completeness:-
If there is some reduction then that means there's a circular component. What is left could either be canceled when it's orthogonal to the plane of the elliptical axis or not canceled at any angle if it's unpolarised. You could still get incomplete cancellation, in which case there could be some additional unpolarised light present with your elliptical.

 

[Andy Resnick]

How to distinguish elliptical polarized light from partially-polarized light?

A Mueller polarimeter (or Mueller ellipsometer) would work well.

 

[sophiecentaur]

I had one of them but the wheels fell off.
I was thinking more kitchen table stuff. Any keen photographer would have basic polarizers in his bag.

 

[peace]

If you mean plane polarised then the word is plane polarised and not "of course" polarised. I had to ask and you needed to specify your question properly.
You need to test for both forms of polarisation and there's a logic to the process which I think makes sense. The answer to the following would be reliable, I think.
Presumably a linear polariser is showing different throughput at different angles(?).
Elliptical polarisation can be looked on as a mix of circular and linear polarisation with the two components co-phased (of course).
The circular polariser would (when the right way round) eliminate any circular component. If it doesn't change the level then there is no circular component. So that would, I guess, tell you that it's not elliptical, which could be enough for you.

But for completeness:-
If there is some reduction then that means there's a circular component. What is left could either be canceled when it's orthogonal to the plane of the elliptical axis or not canceled at any angle if it's unpolarised. You could still get incomplete cancellation, in which case there could be some additional unpolarised light present with your elliptical.

Thanks for the necessary explanation

 

[Andy Resnick]

I had one of them but the wheels fell off.
I was thinking more kitchen table stuff. Any keen photographer would have basic polarizers in his bag.

More than just polarizers are needed- a retarder/compensator (for example, a quarter-wave plate) is also required. Rotation of the compensator allows determination of the orientation and eccentricity of the polarization ellipse (alternatively, determination of the randomly polarized component).

https://www.jawoollam.com/resources/ellipsometry-tutorial
Is it still kitchen-table stuff? Unlikely, for a variety of reasons. Sometimes you really need lab equipment.

 

[sophiecentaur]

Is it still kitchen-table stuff?

It depends what you want out of the experiment. I wouldn't reckon on doing measurements in addition to 'detection. 'Crossed' circular polarisers let nothing through so why won't one circular polariser eliminate CP of the other sense? If that is true then the result of passing EP through a CP filter should leave LP which can be eliminated with an LP filter.
That may be an over-simplification about the elimination of CP with a single CP filter; A photographic CP filter has a LP filter, followed by a quarter wave plate so a photographer would have all that's absolutely necessary (the cheapest polariser is a linear / plane polariser). I think, if you had only one CP filter, you might need a mirror to reverse the sense of the CP for cancellation. Why do you say it wouldn't work? Where would it be going wrong?

 

[Andy Resnick]

It depends what you want out of the experiment. I wouldn't reckon on doing measurements in addition to 'detection. 'Crossed' circular polarisers let nothing through so why won't one circular polariser eliminate CP of the other sense? If that is true then the result of passing EP through a CP filter should leave LP which can be eliminated with an LP filter.
That may be an over-simplification about the elimination of CP with a single CP filter; A photographic CP filter has a LP filter, followed by a quarter wave plate so a photographer would have all that's absolutely necessary (the cheapest polariser is a linear / plane polariser). I think, if you had only one CP filter, you might need a mirror to reverse the sense of the CP for cancellation. Why do you say it wouldn't work? Where would it be going wrong?

Using a circular polarizer is an interesting idea; one challenge is the fact that the relative orientation of the fast axis of the retarder and the pass axis of the linear polarizer is fixed; normally one must be rotated with respect to the other.

It's my understanding that in commercial circular polarizers, the fast axis is 45 degrees from the pass axis. If the retarder is first, the device cannot distinguish between orthogonal linear polarization states. If the polarizer is first, the device cannot distinguish between circular polarization and random polarization. Consequently, the device cannot measure the polarization ellipse. Adding a second polarizer after the circular polarizer won't do anything since the circular polarizer already deleted some information.

The bottom line: specification of the polarization ellipse requires 4 parameters (ellipticity, amplitude, azimuth, and phase) but a circular polarizer has only 2: the rotational orientation of the entire device and the order of elements.

 

[tech99]

Using a circular polarizer is an interesting idea; one challenge is the fact that the relative orientation of the fast axis of the retarder and the pass axis of the linear polarizer is fixed; normally one must be rotated with respect to the other.

It's my understanding that in commercial circular polarizers, the fast axis is 45 degrees from the pass axis. If the retarder is first, the device cannot distinguish between orthogonal linear polarization states. If the polarizer is first, the device cannot distinguish between circular polarization and random polarization. Consequently, the device cannot measure the polarization ellipse. Adding a second polarizer after the circular polarizer won't do anything since the circular polarizer already deleted some information.

The bottom line: specification of the polarization ellipse requires 4 parameters (ellipticity, amplitude, azimuth, and phase) but a circular polarizer has only 2: the rotational orientation of the entire device and the order of elements.

Reading the original question again, I think we have a slightly polarised beam, and we wish to know if it has a linear or elliptical element. This can be ascertained, I think, by using a circular polarised analyser to detect the presence and handedness of any rotational component.
Two further comments regarding detail only. A method for generating an unpolarised beam is to have two orthogonal sources which are driven by separate noise generators. We might then add some elliptical polarised energy to the beam. In this case, our efforts to detect the elliptically polarised energy will be limited due to noise, in particular as the minor axis of the EP may be below the noise.
In the radio engineering context, unknown polarisation can be determined by using a pair of cross polarised antennas and applying the outputs to the X and Y plates of a CRO. We then see an ellipse, whose eccentricity can be measured, together with the handedness.

hoganlalsources

 

[Andy Resnick]

Reading the original question again, I think we have a slightly polarised beam, and we wish to know if it has a linear or elliptical element. This can be ascertained, I think, by using a circular polarised analyser to detect the presence and handedness of any rotational component.

I had been assuming optical (UV-VIS-IR) rather than RF ellipsometry, so I'm not sure what RF polarization analyzer components are available. For example, since coherent detection is possible with RF, your radio engineering example works very well.