@Sand

I didn't know you were a necromancer. What's that like?

In any case, yes, it was a lot of fun. Going in again on Thursday to clean some wafers.

to clean wafers? the producer don't deliver them clean? I mean: they seal them in a clean room as well, or at least I hope so, otherwise my whole understanding of wafer production is wrong.
Anyway, good luck with your research! Watcha trying to prove wih that IC? Or is that classified?

...I don't get it?

What are you doing?

I like vanilla wafers. Don't clean them, though, you'll rub off the sugar.

Mmmm, vanilla wafers.

In other news, that is pretty cool, abge. Congrats.

@semck, hehehe, nice! I like those wafers sugared as well, powder-sugared to be exact.
Or did you make those wafers yourself? Because apart from using billion-worth of machinery, that would be very awesome.

Glad to see you're finally cleaning up your room, getting your life in order.

Sheesh.

I don't even know what a necromancer is. haha

Argh, not another wafer jockey. Enjoy your life of artificial lighting and booties.

I remember when I was in school having the opportunity to do that...
and I turned it down to specialize in software and artificial intelligence.

I was an idiot...well, I'm better in software than hardware, but guess how much AI I'm doing in the real world, where the opportunity actually presented itself several times.... that's right... none :(

@sandgoose

Because you brought this thread back from the dead...


@carp

Yes, they are mostly clean (they go through an enormous cleaning process at the fab), but they still get a little oxide on them between there and here, so it's best to do one final cleaning before you start growing components.

This is just an intro course, so I'll be growing a few devices on the same wafer (MOSFET, NAND gate, OpAmp) and doing verification on them once they're done.

@obi

http://en.wikipedia.org/wiki/Semiconductor_device_fabrication#Wafers

@FK

This course hasn't been offered in a few years. I've been bugging the dept to bring it back. I can't tell you how happy I'm able to take it before I graduate.

@abg, growing components, that sounds like artificial biology. It's a lab course, does that mean that you're going to make those units yourself? (transistors, amplifiers &c, frankly I'm not very familiar with these devices, took me a 15 min recap to have a clue how they might operate, but I guess you want to measure the ideality?) Well, I can see why you would be thrilled to do the course and/or job, I'm very interested in what other people find in the labs (scientifically speaking) than do something similar myself.
@FK I pity you, as long as you have work at the level you studied for, it should be okay.

abge,

I saw that nobody replied from your excitement so, I needed to congratulate you :)

@carp

It's called growing because crystals grow a lot like plants. You start with a Si substrate and then you grow an oxide layer on top. After that, you grow various metal and oxide layers to make your device.

Yes, I'll be growing the MOSFET and other components. It's pretty much the most basic step for making an electronic device, which is why I think it's so cool. I'll then be doing verification to make sure what I made actually works as it should.

Here's an SEM picture of more-or-less what I'd be making:

http://www.silvaco.com/tech_lib_TCAD/simulationstandard/2003/may/a1/may03_a1_fig8.gif

@sand

It was a joke. Thank you; I'm glad for the opportunity to brag about what I'm doing : )

Neat! My professionality is not in this field, but I got the sense that without any of such components it is quite impossible for me to reach you at this moment. In my opinion the process of stacking layers onto eachother could be performed faster by using other methods of layer application (but it could compromise the integrity your chip). I was quite familiar with selectively remove (parts of) doped material or apply that dope material on a substrate, it is clear you're doing something totally different. This can also be seen from your SEM picture (for the layman: those rounded curves you see could be straighter by using the right etching technique), why exactly do you use this crystallization technique? Is the integrity/reliability of the component so much more important that a faster technique? Or is such a doping technique in fact slower, because the crystallization is performed at room temperature and the dryer is impractical in the cleanroom?
More questions than I anticipated :s... Anyway, this is quite interesting to me and I wish you good luck in the cleanroom!

Yes, you are correct, etching is an important part of chip making.

Oxidation is typically grown on the silicon because it makes very, very pure electronic grade oxide. It's possible to make oxide other ways, such as bombardment, but this is not suitable for electronic grade devices.

Etching usually comes later, after the substrate has been grown. For instance, the contacts are made by growing a layer on the substrate and then etching away the parts you don't want.

Those wells (or curves) you see in the SEM I linked were most likely done with Impact Ionization. I suppose you could etch them and then fill them back up, but I'm sure there's a reason people don't do that.

The oxide layer is not made at room temperature (that would take forever). The oxide layer I'm growing is going in a furnace at 1100 degC for about an hour. We're using a dry oxidation, which takes longer than wet oxidation, but the result is much better.

The short answer is: Yes, with modern devices being as small as they are, reliability is much, much more important than how quickly you can manufacture them.

Oh, and what do you do, if I may ask?

Well, I'm not doin anything fancy as you do right now, I'm at university, and need to do another year for my MSc degree in Chemical Engineering. Right now I'm on the verge of starting my internship.
After attaining my BSc degree I changed universities, primarily because I saw the level of my education at the university was falling and there was no fast and/or easy way out according to the program director. The mentor of my BSc thesis was actually doing quite some interesting stuff with the techniques you mentioned. Imagine that for half a year you get a 2-hour lecture about state-of-the-art research every friday morning from a PhD on his research. Techniques were applied to different situations, explained (briefly) why, how and what technique was chosen, what the results were and why it is interesting. It was (and still is) quite appealing to me.

My cousin is a ChemE. Don't understand at all what he does, but he seems to make a lot of money.

Just finished cleaning and oxidizing the wafer. Cleaned it in a Hydrogen Peroxide and Sulfuric Acid mix and then in HF. It's baking in the oven now for the oxide layer.