No..Its only mirror. Hopefully it will change the polarization

Well, a periscope is an instrument for observation over, around or through an object, obstacle or condition that prevents direct line-of-sight observation from an observer's current position. In its simplest form, it consists of an outer case with mirrors at each end set parallel to each other at a 45-degree angle. 

So, it's only mirror and has nothing to do with the polarization. 

I think laser polarization does not change after passing through periscope.

Santanu and Ramy,

So what is happening is the beam is not only taken up using periscope but it is changing direction. Say the main beam hitting lower mirror upward is P polarized with respect to the table (as a reference), then it is coming out through the upper mirror which is 90 degrees turned toward the left. I think in this case its polarization will change into S due to oscillation of the electric field changing direction.

Hello Rad --

The short answer to your question is YES, the polarization WILL change in the periscope you described. See the attached figure that I found online; hopefully this gives you a clear visual reason of why the polarization changes. In the case presented in this illustration, the picture is valid for any type of reflective surface.

The long answer (i.e. to the general case) is that the resulting polarization depends on 1) the input polarization, 2) the angles of incidence on the two periscope mirrors, and 3) the type of mirrors used (i.e. silver, aluminum, dielectric, etc.) 

Consider the situation like the one in the picture: a horizontally-polarized (or p-polarized) laser beam is propagating parallel to the surface of a table. When the beam is incident on a mirror oriented face-up at 45 degrees (such that the beam will be directed straight up at 90 degrees), what is the polarization? For the purposes of the mirror, the beam is actually s-polarized (not p-polarized!) because polarization is defined in relation to the incident surface, so if you change the angle of your interface then the polarization of the beam "changes" for the purposes of calculating the interaction. In this case,while the polarization is parallel to the table, the polarization is perpendicular to the mirror interface, so the calculation of the reflection must be treated using the boundary conditions/Fresnel coefficients corresponding to s-polarization, not p-polarization. Does that make sense?

This is why the angle of the top mirror is so critical: if the top mirror is angled such that the beam is sent back directly above the input beam (like a roof mirror), then the polarization is the same because the plane of incidence is the same for both mirrors. If the top mirror is angled such that the beam exits perpendicularly from the periscope (like in the picture), the plane of incidence rotates by 90 degrees, so the polarization "rotates". If the top mirror is angled such that the beam exits at some other angle between 0 degrees and 90 degrees, then the incident beam will appear to have two polarization components in the coordinate system defined by the interface. Because the reflected waves of the two different polarization components will in general have different magnitudes and phases (according again to the Fresnel coefficients, which are dependent on the optical properties of the mirror), the resulting polarization will be elliptical rather than linear. This is a very common lab mistake!

Did this explanation make sense or address your specific concern? I'd be happy to try to clarify.

Best,

~Eric (Institute for the Frontier of Attosecond Science and Technology, UCF)

Hi Eric,

Thanks for your high caliberd answer.

Cheers,

Rad

Eric`s answer is absolutely true and complete.

The answer to your question also depends on the type of periscope arrangement you are considering, i.e., whether it's a conventional or twisted periscope design and whether the incident light is elliptically polarized. As it turns out, a general elliptical polarization state is quite well preserved when passed through a 90-degree "twisted periscope" arrangement (such as the periscope depicted in the figure above). An initially linear polarization state is indeed rotated upon exiting, as shown, but an elliptically polarized incident beam will exit the twisted periscope with its polarization state virtually unaltered. This result is in stark contrast to changes in the elliptical state of polarization, which necessarily result in the case of a conventional periscope (whereby light rays exit in same propagation direction as the incident rays). SEE: Ashish Arora and Sandip Ghosh, "A twisted periscope arrangement for transporting elliptically polarized light without change in its polarization state", REVIEW OF SCIENTIFIC INSTRUMENTS 81, 123102 (2010) ABSTRACT: "The authors describe a dual-mirror 90◦ twisted periscope arrangement for transporting polarized light in air, wherein a general elliptic polarization state of the light is preserved at all wavelengths. It is experimentally demonstrated that in the wavelength range 400–1000 nm, this arrangement preserves the polarization state fairly well when using either metallic or dielectric mirrors."~~~

Is there any way other than the twisted periscope configuration that can be used to preserve the circular polarization? We are presently using a normal periscope in our setup and would like to modify around that in order to reduce reflection induced variations in the circular polarization of the light.

This answer is a bit late seeing that you posted your question on Feb 24, and so you may have resolved this issue by now. Depending upon your budget and the diameter of your laser beam, you might consider implementing something like a Berek Variable Waveplate (see: https://www.newport.com/f/bereks-wave-plates) positioned after your periscope to correct for the changes induced by reflections. There are alternative variable retarders available, such as Soleil-Babinet Compensators (see: https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=871) These items are admittedly a bit pricey and have a limited clear aperture of ~10mm.