Excellent question. You frequency-double 532nm and make a typo. I meant 266nm.
There are crystals that have nonlinear responses to high electric fields, and if you hit them with enough laser light, some of it comes out at half the wavelength. A lot of it also comes out at the original wavelength. Most 532nm lasers work like this, but other input wavelengths are possible, like starting at 532, doubling again, and getting 266nm.
This was a long time ago, and it wasn’t my project, so it’s possible it was slightly different, but I definitely remember the 532nm stage. And 266nm sounds credible for what the group was trying to do with the laser.
There are some recent papers on doing it, and they even seem to have gotten decent efficiency.
The lab I was in was doing this in 2000, and I suspect they got their frequency doubler from some other lab. It worked, but it certainly didn’t work well. The 1064nm laser was decently large (a couple J per pulse, from vague memory), and the expected UV energy was quite low. The laser was being used for some form of imaging at short range (fluorescence or absorption in burning gasses? Maybe Raman spectroscopy if everything got lucky?).
There are crystals that have nonlinear responses to high electric fields, and if you hit them with enough laser light, some of it comes out at half the wavelength. A lot of it also comes out at the original wavelength. Most 532nm lasers work like this, but other input wavelengths are possible, like starting at 532, doubling again, and getting 266nm.
This was a long time ago, and it wasn’t my project, so it’s possible it was slightly different, but I definitely remember the 532nm stage. And 266nm sounds credible for what the group was trying to do with the laser.