A little bit before Christmas, I spent an afternoon swapping mirrors out of one line of the apparatus. I was losing too much of the laser light before it went into the chamber, and replacing the mirrors increased the power entering the apparatus by a factor of two or so.
Here’s a picture of the two types of mirrors, side-by-side:
“Well, of course you had to replace them,” you say. “The one on the left is a perfectly nice mirror, but the one on the right is utter crap. You dolt.”
The thing is, the mirror on the right is the type I was putting in. The one on the left is the type that wasn’t any good.
The key thing here is that the laser wavelengths I’m working with are in the near infrared. They’re close enough to visible that you can just barely see the beam from the high-power laser hitting a piece of paper, but 819 nm is outside the normal range we can see. Which means that a mirror for 819 nm light doesn’t necessarily look like a very good mirror when if comes to reflecting visible light. In fact, you can see right through it, to the “E03″printed on the back to tell you what kind of coating it has.
Which is not to say that the mirror on the left is all that bad. The reflectivity at 819 nm is probably somewhere between 80 and 90 percent, which isn’t all that bad in the grand scheme of things. But if you bounce your laser off two mirrors that each reflect 80% of the light, you’re down to 64% of the initial intensity after the second bounce. A third takes you down around 50%, and I replaced three mirrors that day.
The mirror on the right is supposed to reflect better than 99% of the light at 819 nm. Which is about the same as the mirror on the left at 550 nm, in the middle of the visible range. That’s a good mirror.
Why did I have visible light mirrors in the line in the first place? Because we had them lying around in one of the teaching labs, and I was low on cash. Mirrors of this quality aren’t cheap– either of the two in the picture will set you back $79 from ThorLabs, or for $109 they’ll throw in a kinematic mirror mount. Compared to the $7,000 UV lamp we have, that doesn’t seem like much, but I’ve got something like 60 of these mirrors and mounts tied up in the apparatus already, and I’m always running low (I’m going to need to put in another big ThorLabs order in the next week or so…).
In a pinch, you can get away with using a visible mirror or two, and there are always a few rattling around the lab. It’s not a good idea to use too many, though, hence the afternoon of mirror-swapping.
Of course, those aren’t the worst really good mirrors I have in the lab. The guy who had the lab before me worked with a YAG laser, which has a wavelength of 1064 nm, well outside the visible range. There are still a few mirrors from his apparatus floating around in the lab– I use them as beamsplitters, because they reflect maybe two-thirds of the light, and let most of the other third pass on through…