Ordinary glass cube beam splitters are constructed with a dielectric or hybrid coating on the hypotenuse of a right angle prism which is then cemented to another right angle prism. These fail in high energy applications because the cement absorbs enough energy to cause problems.

For higher power applications, the two prisms are optically contacted so that there is no cement to fail.

I have read that for even higher power applications, the prisms are air spaced.

Are these air spaced prisms constructed by coating one prism, and bringing it close to, but not touching, the second prism? If so, why do these have a higher damage threshold than optically contacted prisms? Is the spacing chosen to be large enough to avoid frustrated total internal reflection?

Or, when people refer to air spaced prisms do they mean that the mechanism is frustrated total internal reflection, and there is no coating on either prism?


There's a really good answer in another thread. Basically its FTIR that's causing the one part of the beam, and evanescent waves that couple to the second prism for the second beam. That way there's no coating to burn off.

The air is not important, just the spacing of the gap, which controls how much of the beam goes where. Vacuum works just as well, but is obviously somewhat more difficult to arrange. I suspect vacuum gapped versions are common in NIF though.

  • $\begingroup$ Thanks. I saw that thread. What is not clear to me is if the term "air spaced prism" means exactly the same thing as "FTIR prism". $\endgroup$ – garyp Jan 10 '19 at 15:07
  • $\begingroup$ Hmmm, "air spaced prism" I can't say, but google suggests "air spaced beam splitter" does ultimately devolve to F-TIR, potentially through Glan. $\endgroup$ – Maury Markowitz Jan 10 '19 at 15:17
  • $\begingroup$ Argh! I'm constantly saying "prism" when I mean "beamsplitter". "Air spaced beam splitter" vs. "FTIR beam splitter". (I actually typed "prism" just now.) Sorry. And I'm definitely talking about glass here. $\endgroup$ – garyp Jan 10 '19 at 15:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.