My beamsplitter is of the polarized kind. If you don’t believe me just look at this picture of the delicious snacks Thorlabs sent me:
The reflection from the microscope objective is almost completely eliminated when the camera’s linear polarizer is rotated 90⁰.
If my laser had also been polarized the beamsplitter would transmit almost 100% of the light to the objective. Unfortunately it’s not, so approximately 75 mW of laser power is lost (not counting reflections), but what’s worse is that I end up with lots of light to dispose of safely, in a manner that will interfere minimally with the rest of the system. In other words; I need a beam dump.
A cheap way of constructing an efficient beam dump is to stack a bunch of razor blades.¹ The geometry of the razor blade edges traps the reflections of the light if the angle of incidence is not too large, making the stack of razors effectively act as a black body.
The following image taken from Den Hartog’s article shows the schematic of the beam dump they employed:
It’s very elegant. The first filter is oriented in Brewster’s angle reflecting very little of the light because of the horizontal polarization of the incoming laser beam (indicated by the arrows perpendicular to the beam).
The authors used a pulsed ruby-rod laser with a wavelength of 694.3 nm and a beam diameter of 16 mm. The intensity of the pulses were 10 J and the duration 40 ns. I think I can safely say my laser gets nowhere near that kind of output.
More lab snacks once I get all my razor blades!
¹A simple, high-performance Thomson-scattering diagnostic for high-temperature plasma research
D J Den Hartog and M Cekic 1994 Meas. Sci. Technol. 5 1115