optical MEMS stuff

They must be fishing for a huge OEM buyer. We're not worthy of a data sheet, much less a sample.

I was excited by the above APD's till I saw the spectral response. I don't know why it peaks at ~500 nm. I'd like something out to 800-900 nm.

George H.

It does, in both PIN and APD form, and you're exactly right. The vacuum APD is probably the state of the art single-photon detector.

It works even better if you couple the light into the faceplate with a prism, so that it has to rattle round in the glass and photocathode till it gets absorbed. You can easily double the QE of a photocathode that way.

Cheers

Phil Hobbs

How about an AR coating between the inside glass and the photocathode stuff?

If the light has a big enough transverse k vector that it's evanescent in air, it can't pass through the photocathode without being totally internally reflected at the PC/vacuum boundary. Thus it has to keep rattling round till it gets absorbed.

Cheers

Phil Hobbs

If the Lucas ignition system lasts long enough to get you there.

Are you saying that it's better to hit the photocathode at a low angle (in the TIR region) and let the mythical evanescent wave kick out the electrons?

That's interesting. A photon either gets totally reflected, to try again, or its evanescent component gets absorbed and makes an electron. So no photons are lost passing through the photocathode.

Something like that? Edge-launch PMT?

Same idea, just a different implementation. The real problem is that the photocathode has to be held at a bias potential, so it's a metal or semiconductor layer. That means very high dielectric constant, so high reflectivity. If the prism can be finagled to Brewster's angle, one gets a very nice coupling boost, as well.

The photocathode thickness must be tricky. Too thin, the resistance is high and the light passes through. Too thick, the electrons can't get out.

I visited the Hamamatsu PMT facility. A PMT is stuck into a light box with some fluorescent bulbs inside. There's an analog meter to indicate current. A long glass tube leads into the PMT. Inside that is a little steel boat with some photocathode material inside. That tube passes through a gas flame. The operator uses a magnet to slide the boat in and out of the hot zone, evaporating stuff onto the inside of the tube until the meter reads right; then they seal it off. Must be tradition.

I was told that the real hard part was getting all their processes documented for ISO9000.

Transparent photocathodes let a lot of the light pass through, which is a major reason for their comparative inefficiency. If you come in via a prism, the light can't escape from the photocathode into the vacuum, so it has N chances to be absorbed.

Cheers

Phil Hobbs

Right. With a negative electron-affinity (NEA) photocathode, any free electron that hits the PC/vacuum interface escapes. That makes an efficient but very slow photocathode. Normal bialkali photocathodes have a significant barrier, so they're less efficient but much faster.

Cheers

Phil Hobbs

As Bismarck famously said, "The law is like sausage. It's better not to watch it being made."

Cheers

Phil Hobbs

So coat an optical fiber with photocathode material. Must be good for something. Fast traveling-wave o/e converter?

A student-level scanning electron microscope doesn't need a lot of electrons, but the emitter needs to be small. I proposed a tapered optical fiber with some photocathode stuff on the flat end. Shine in light, get electrons. The numbers looked OK, but the people that I was working with went off an another tangent.