4.  Selecting the wavelength [solution]
The first polarizer produces a polarized beam with the polarization angle at 45° to the optic axis. This means the light incident on the calcite has equal amplitude components polarized parallel to the optic axis and perpedicular to it. Both components are undeflected (polarization along principal axis), but they travel at different speeds and get out of phase with one another. When they leave the calcite the phase difference is then constant, so the thickness of the calcite plate will determine the polarization state of the emerging light. If the accumulated phase difference is d = p/2, the wave will be circularly polarized. If d = p, the polarization is reflected about the optic axis (which in this case will produce a polarization vector inclined at 135° to the optic axis) but the light that comes out is linearly polarized. For intermediate values of d, the polarization will be elliptical. The value of d produced by a plate of thickness d is given by

The problem statement tells us that for one of the two Na wavelengths (l1 = 588.9950 nm and l2 = 589.5924 nm), no light emerges from the second polarizer. This must mean that for the wavelength that is "eliminated," the light coming out of the calcite is linearly polarized perpendicular to the second polarizer (which is oriented the same way as the first). This requres that the total phase difference d be equal to an odd multiple of p for this wavelength. But we require not only that d for one wavelength be an odd multiple of p (so it is blocked by the second polarizer), but also that d for the other member of the doublet be equal (or very close) to an even multiple of p so that it passes through the second polarizer. Can we achieve both conditions for a single thickness d? We need to see how d changes with l, so take the derivative with respect to l of the equation above:

 We have all these quantities, so we just need to know what value of d gives us 

A little number-crunching yields d = 1.563 mm, a reasonably convenient thickness. What would happen if you rotated the second polarizer by 90°?
Back to HW2