A Few Questions About The Way A Linear Polarizer Filters Light

Question

I'm trying to wrap my head around how polarizers work out of curiosity and I have a few questions.

1. Let's say I have a polarizer whose bars are vertical. I see some sources which claim that it lets only vertical waves to pass through because vertical waves get through the small slits. Some sources contradict this though, as claim they're actually the ones getting absorbed by the bars and don't make it through. Which theory is correct?

2. Given an unpolarized wave composed out of many polarized electric fields. I kind of "can" decompose them to parallel and perpendicular to the bars, and I "can" claim the components parallel get through (that is assuming the first option in 1 is correct!). Is this wrong simply because the decomposition isn't a valid move here? After all the waves with a perpendicular component physically shouldn't get through.

3. Why does nobody talk about the magnetic fields in relation to polarization? Is this because only the electric field is related to vision? Even so, does the polarizer affect the magnetic fields?

4. Is there a simple reason that can help one can understand why reflective surfaces (like water or ice) polarize light? and how do you determine the direction of propagation of the electric field after polarization?

Thanks in advance.

Answer 1

$1$. For a wire-grid polarizer, the light polarized perpendicular to the wires will pass through. This counterintuitive result is because any electric field polarized parallel to the wires would excite currents in the wires, thereby reflecting/absorbing the light. But for electric fields perpendicular to the wires, the confinement of the electrons pushes the plasma oscillation frequency too high, effectively suppressing the AC conductivity of the wires at the light frequency. Thus, perpendicular polarized light basically doesn't see the metal, and it is transmitted.

$4$. The reason why smooth dielectric interfaces (such as water or glass) polarize light is due to the fact that Maxwell's equations impose different boundary conditions for in-plane magnetic and electric fields. Thus s-polarized light at a boundary (in-plane electric field) reflects differently than p-polarized light (in-plane magnetic fields). So if you start with unpolarized light incident at an angle, you'll get at least partially polarized light in reflection/transmission.

Answer 2

Let me answer you a few of them in short:

2 - It is a valid move. Light can be decomposed using polarizers, in effect, into vertical and horizontal bases.

3 - Because electric and magnetic fields are related and, therefore, talking about one is the same as talking about the other one.