Frickin’ Lasers

I haven’t posted much about life in the lab lately, because even though I’m getting to spend a bit of time in the lab, I’ve been so fried from this past term that I haven’t had much energy for blogging. Things are finally settling into the summer routine, though, and I’ve gotten a little rest since handing in my grades, so I’ll try to post occasional updates on what’s going on in the lab.

Of course, life in the lab has its own frustrations, chief among them being equipment failures for stupid reasons. We’ve had another such event in my lab, so I’ll be spending a bit of time wrestling with my grating-locked diode laser. Which means, I might as well get some blog post about it, so there follows a brief explanation of lasers, laser diodes, and related topics, below the fold.

A laser, as this handy JavaScript tutorial will tell you consists of three basic elements: a pair of mirrors, a gain medium, and a way to excite the atoms in the gain medium. The key idea is that a photon of light passing by an excited atom can cause that atom to emit a photon that is identical to the original photon– same wavelength, same direction, same phase. This is called “stimulated emission,” in contrast to the process of “spontaneous emission,” wherein an excited atom emits a photon in a random direction with a random phase, at a random time.

The idea of a laser is to sandwich a bunch of excited atoms between two mirrors (the “laser cavity”). Any photons that end up between the mirrors (say, due to spontaneous emission by one of the excited atoms) will bounce back and forth between the mirrors, stimulating the emission of more and more excited photons. Each time a photon hits an excited atom, you create a second photon, so the number of photons between the mirrors increases exponentially, provided you have a means of continually re-exciting the atoms in the gain medium. Since no mirrors are ever perfect, a tiny fraction of this light will leak out through one mirror or the other, forming the laser beam that we see.

There are lots of different types of lasers, because there are lots of options as far as gain media go. The simplest sort of lasers are probably gas lasers, which use, well, a gas of atoms as the gain medium. The most common variety is the helium-neon laser, which uses a mix of helium and neon atoms excited in a plasma discharge, and produces red light at 633 nm. If you’ve played with a red laser as part of a science class, odds are it was a helium-neon laser.

Gas lasers are very limited in the wavelengths they can emit, because they depend on a single atomic transition, and there’s very little to perturb the energy levels of atoms in a gas. You can get a more flexible sort of laser by using a dye made up of complicated organic molecules for your gain medium. Molecules have lots of closely spaced energy levels, and sufficiently complicated molecules will emit over a broad, continuous band of wavelengths, and can work as a laser at any wavelength in that range. To make a dye laser work, you generally add some optical element in the laser cavity that selects a particular wavelength (say, by absorbing photons of other wavelengths, or reflecting undesirable wavelengths out of the cavity), to get a nice narrow line out, that you can tune by varying your wavelength selecting element. This lets you cover a wide range of frequencies with a single laser, which is incredibly useful for spectroscopy experiments.

Dye lasers work pretty well as far as they go, but they have a number of disadvantages. The molecules used are frequently fairly nasty stuff– if they’re not toxic, they probably cause cancer, or need to be dissolved in something that is toxic or causes cancer– and as the name “dye” suggests, they stain everything You can always tell a lab that contains or has contained a dye laser, because there will be little colored blotches on the floor, the walls, and in certain memorable cases, the ceiling.

Solid state lasers offer most of the nice features of dye lasers, without the mess. As the name implies, the gain medium in this case is a solid– generally, a small number of atoms of a particular element are embedded in a glass or crystal matrix. The most common example is probably the Titanium: sapphire laser (Ti:Sapph), which uses titanium embedded in sapphire (Al₂O₃, which is just a type of glass, for this purpose– the Ti:sapph crystals used for lasers aren’t especially pretty). In this case, the titanium atoms act as the gain medium, and the interactions between the titanium atoms and the aluminum and oxygen atoms of the crystal matrix broaden the atomic energy levels to allow laser operation over an extremely large range of wavelengths, spanning several hundred nanometers (roughly 650 nm (red) to 1100 nm (infrared))– a Ti:Sapph laser is tunable over a range of wavelengths that is almost twice as wide was the wavelength range of visible light.

Solid state lasers are a joy to work with, relative to most other types of lasers, but they’re big and expensive. Ti:sapph lasers also generally require another laser (either an argon ion laser (a type of gas laser), or a doubled YAG laser (a different type of solid-state laser) to pump them, which adds to the expense. I used a Ti:Sapph laser for my thesis, and I’d love to have another one now, but I just can’t afford it.

The budget option, for the scientist who wants a reasonably large tuning range (several nm) without a six-figure price tag is the semiconductor diode laser. As any pop starlet can tell you, these lasers make use of the fact that semiconductor materials have broad energy bands with a gap between them. The gap size for typical semiconductors corresponds to the energy of a photon in the visible or infrared range of the spectrum, so an electron flowing through a diode has a chance to drop down from the upper “conduction” band into the lower “valence” band, emitting a photon. This emission is the basis for a light-emitting diode (LED) like the little lights in the keys of your cell phone, and with a little work, you can turn that light source into a laser.

Diode lasers are relatively cheap (tens or hundreds of dollars, rather than a hundred thousand dollars), and very small– the actual laser chips are something like a hundred nanometers across, and they come in packages that are a few milimeters across. These are the lasers that you find in CD players, supermarket scanners, and laser pointers.

From a physicist’s standpoint, they’re almost ideal laser sources, but they have one major problem: an unmodified laser diode emits light over a fairly broad range of wavelengths. This isn’t a big deal if you just need a bright source of light, but it’s death if you’re trying to do precision spectroscopy or laser cooling. So we need to play some tricks to narrow the wavelength spread of the beam that comes out, and select the specific wavelength that we want. That process is the source of most of my current lab headaches, and deserves a post of its own.