I am working on a business proposal that has as its heart the creation of an electron beam lithography laboratory. I have been trying to find some sources of information on exactly what sorts of equipment are needed, what sorts of chemicals, etc. I know that I can get a used SEM and some (relatively) inexpensive software for exposing the masks and have found information on preparing and developing the masks, but the information I have on etching, doping and layering all seem targeted toward large-scale production (i.e., Intel sized), not research scale where I will probably be working with cm square sized chips or smaller. Can any one suggest some references to get me started? I am a biochemist by formal education, so please don't assume too much.

Thanks!

I think this is a hard "business" to get into.

While a SEM is a good and common tool to use, the other process control part is still very difficult.

For example, you will need a cleanroom with controlled humidity and temperature and contamination, to get respectable qualification. Otherwise your research results will not be trustworthy.

Are your substrates 6" or 8" wafers? There might be some used equipment to help set up something. Otherwise, I think it will be hard to even get the right platform to fit other substrates.

Thanks for the input guiding_light!

I don't have a clean 'room' but I have a clean working area (approx. cubic meter) that should be equivalent to class 100.

For prototype development I do not plan on using full sized wafers, rather centimeter (perhaps millimeter) square sized.

In that case, you cannot do your own coating or etching or any materials handling.

You at best can only operate the SEM and you will not be able to use chemically amplified resists.

The customer would have to do the rest of the processing. But it may occur to them to use their own SEM, especially if it there is already one in their cleanroom.

SEM resolution varies from time to time, few nm or so, especially depending on the way you measure. For features > 100 nm, this might be tolerable (few %) but for smaller features, don't see how this can be tolerable.

How can this be controlled? Measure 30 nm today, 25 nm tomorrow, 27 nm the day after? Is my CD-SEM reference trusted by my customer? Is resolution ultimately operator-defined?

I came across a recent paper on SEM impact on device damage, I also posted it on other threads. You need to reduce the landing voltage in order to prevent the beam electrons from penetrating to the device.