Scattering of X-rays and polarization

[Gruxg]

I read once (I don't remember exactly where) that X-rays scattered perpendicularly to the direction of the incident X-rays are linearly polarized (even although the incident ones are not). I think the discussion was in the context of low energy X-rays, and the explanation used classical electromagnetism arguments: it was the same as the polarization by scattering mentioned here: http://hyperphysics.phy-astr.gsu.edu/hbase/phyopt/polar.html

Is it true only for very low energy X-rays (with Rayleigh or Thompson scattering, i.e, frequency of the scattering = frequency of incident radiation), or is it true also for higher energies where the scattering is mainly by Compton effect?

 

[M Quack]

This was discovered by Charles Barkla in 1905 (the paper is from 1906).

http://www.nobelprize.org/nobel_prizes/physics/laureates/1917/barkla-bio.html

The classical polarization dependence holds for Thomson scattering. There are other effects such as multiple scattering, magnetic scattering, and resonant scattering that have different polarization behavior.

http://link.springer.com/article/10.1140/epjst/e2012-01630-3


I am not sure about Compton scattering. There are polarization effects that have been used for experiments, in particular to study magnetism

http://geant4.cern.ch/G4UsersDocuments/UsersGuides/PhysicsReferenceManual/html/node56.html

 

[sophiecentaur]

Isn't the Brewster Angle involved in this? At large angles of incidence, you get reflection of X Rays at metal surface. It's how X Ray telescopes work.

 

[M Quack]

Total external reflection of x-rays is indeed used for x-ray optics (terrestrial and astro). However, it is not an efficient way to control or modify polarization.

The angles up to which you get total external reflection of x-rays is typically very small. e.g. for a Gold surface and 12.4 keV x-rays (1 Angstrom wave length), the cut-off angle is 0.351 deg.

http://www.x-ray-optics.de/index.php?option=com_content&view=article&id=52&Itemid=64&lang=en [Broken]

The large angles close to 90deg you get by using Bragg reflections, choosing a crystal and Bragg reflection that have the correct d-spacing (or choosing the wave length to match the crystal). With this you can get excellent polarization purity.

http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.110.254801

The obvious disadvantage of Bragg-based optics compared to mirrors is that Bragg optics only work for a very narrow energy band, whereas mirrors can work for a wide range of energies.