Dear Eugenio! To my opinion and experience the best solution of  the problem  is application to reflection grating made of cholesteric liquid crystal with homogenic orientation of  director on glass substrates. Similar grating will separate the even usual unpolarised light  on two beams: backward and forward directed  with RCP and LCP. Rotation direction of  E-vector of backward reflected beam is determined by  twisting direction of  cholesteric liquid crystal. Unfortunately I  do not know  if commercial manufacturing of similar grating exist. 

Polarization grating can do that. It can diffract LCP and RCP into two different directions. I don't know exactly if there is any other commercial device.

If you have a device which separates orthogonal linearly polarized lights, then just keep a quarter wave plate in front of it. For a suitable orientation, the device can separates RCP and LCP, however the out coming light will be linearly polarized. If you want the nature of circular polarization to be preserved then use some combination of wave plates at the output port.

Dear Eugenio! To my opinion and experience the best solution of  the problem  is application to reflection grating made of cholesteric liquid crystal with homogenic orientation of  director on glass substrates. Similar grating will separate the even usual unpolarised light  on two beams: backward and forward directed  with RCP and LCP. Rotation direction of  E-vector of backward reflected beam is determined by  twisting direction of  cholesteric liquid crystal. Unfortunately I  do not know  if commercial manufacturing of similar grating exist. 

A Fresnel rhomb can be used to convert linearly polarized light into circular by orienting the TIR surfaces wrt the incoming polarization, so it may be that it can convert incoming LCP and RCP into two orthogonal linear polarizations which you can then resolve conventionally? The nice thing is that compared to wave plates etc the rhomb is far less sensitive to wavelength.

The solutions for separation are suggested above. At the same time, there is a rigorous analogue of a birefringent crystal - any gyrotropic media acts as an anisotropic with normal waves being LCP and RCP. 

Thank you all for your quick and nice answers! 

You can also use a quartz cube, place it in the path of the linearly polarized laser light beam and exert some pressure on top of the cube by some mechanism. At some value of the pressure the light will be converted in to circularly polarized light which can be confirmed by placing a polarized or you may use a lambda by four plate.

have fun 

You can convert the linear polarized light to LCP and RCP using QWP or Fresnel Rhomb . Polarization grating may be useful to diffract RCP and LCP due the angular dispersion.

Hi Eugenio: Just had another thought. More complicated than the first idea, but if you want to make a measurement with a single detector, you can use a photoelastic modulator http://www.hindsinstruments.com/products/photoelastic-modulators/ to modulate the polarization of the beam in real time from linear to RCP to linear to LCP and back. This device is quasi-achromatic in that you set the amplitude of the driver to get the correct retardance at any given wavelength. The crystal modulates the polarization at tens of kHz, so if you have a beam containing some mixture of CP, putting a polarizer after the PEM will alternately transmit LCP and RCP on each half-cycle of the PEM. You may get enough info to get the full Poincare sphere (the two linear polarizations) as well if that's helpful? John

Imagine Optix is a commercial source for gratings that diffract circular polarized light.

What about  trying to use an incident linear polarized light beam upon a non isotropic cristal?

Since a polarized beam is equvalent to a superposition of  clockwise  and an anti clockwise  beam.  Hence...

I am not sure of commercial availability, but there has been excellent work in this area at the University of Central Florida (CREOL) on developing wide band circular polarizers using gold etched metamaterials.  Basically, they have extended RF meanderline polarizers down to dimensions usable in the infrared band (8-12um) at least. 

Dearest Eugenio,

As someone pointed out, polarization Gratings based on Liquid Crystal (nematic or ferroelectric) could be solution to your question. Please take a look to:

Polarization beam splitters using polarization diffraction gratings

http://dx.doi.org/10.1364/OL.26.000587

Regards ;-)