h a l f b a k e r y"Look on my works, ye Mighty, and despair!"
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
In the electronics manufacturing industry, one of the most important things is the necessity to control dust. One speck can ruin a complex processor chip, for example. So, great pains are taken to filter it out of the air, as workers move into and out of the clean-room manufacturing area.
In the
future this technology needs to be improved, to keep pace with all the others. Here are my suggestions along those lines.
1. Paint EVERYTHING in the clean room with titanium dioxide, including the special suits people will wear.
2. Increase oxygen level in the work environment to 30% (20% is normal; back in the Carboniferous Era, Earth's atmosphere was as much as 35% oxygen).
3. Install lots and lots of ultraviolet lamps, to ensure that every painted surface is illuminated by UV.
4. Continue to use filters to catch the majority of dust.
5. Everyone needs to wear appropriate goggles to protect eyes from UV light.
What happens in this clean room is that the UV activates the titanium dioxide, such that any speck of dust that touches any painted surface will be destroyed. What the activated titanium dioxide does is catalyze a combustion reaction between oxygen in the air and whatever other substance is touching it, like dust. The enhanced oxygen level will increase the effectiveness of that reaction.
Since most dust is made of organic matter, rich in hydrogen and carbon, the result will be mostly water vapor and carbon dioxide, things that won't significantly affect the manufacturing processes (also see next paragraph).
Since most dust will be trapped by the ordinary filter system, only rather small quantities of dust will be "there" to be destroyed by this system. So, no huge fires and no significant air pollution from burned dust.
The manufacturing clean room will be super-clean, therefore.
[link]
|
| |
I don't know the details of modern semiconductor
production, but don't some of the steps rely on UV-
sensitive photoresists? |
|
| |
So instead of fine dust that the filters can catch, you will
have even finer ash. |
|
| |
Better hope there's not a fire. Even a 1% rise in
atmospheric oxygen levels, such as may be caused by a
leaky valve in a surgical theatre, can be a recipe for
disaster. A 10% increase, and we're talking about people
with fire in their lungs before automatic extinguishers can
be triggered. I've dealt with oxygen explosions before, but
never in an enclosed space. In the open air, your eyebrows
are gone before you hear the bang. In a sealed room... |
|
| |
//don't some of the steps rely on UV- sensitive photoresists?// |
|
| |
I believe so - but as the feature size shrinks, so too does the wavelength of light necessary to reproduce it on the wafer surface. Vernon is talking about future chip fabs, where we'll presumably be using X-rays, gamma rays, electron beams or such. |
|
| |
From where does the dust come? If it comes from
people, and from air getting in past airlocks, then
surely the solution is to have a people-free fab?
Operate the whole thing as a sealed unit. |
|
| |
Better yet, operate in a vacuum. Dust in a
vacuum must settle out almost instantly. |
|
| |
Even better yetter, design chips to cope with
failure. This could be done by bolstering their
egos and telling them it's OK. Alternatively, they
could be designed in modular way, with enough
redundancy that a one-time-only operation will
bypass any failed modules and pull in spare ones.
I'm sure this has been done or tried. |
|
| |
The Soviets tried; they gave the job to Mikhail Kalashnikov,
who came up with a motherboard that could withstand
punishing heat and Arctic freezes, worked just fine when
fully submerged in pig shit, was easy to clean and repair in
the field, but couldn't run DOS to save its godless commie
soul. |
|
| |
//couldn't run DOS// So, a complete success, then? |
|
| |
From the Soviets' standpoint, yes. |
|
| |
[Wrongfellow] is correct; they are already using "deep ultraviolet" in today's fabs. The UV from the lamps needed to activate titanium dioxide may already be too-long in wavelength to affect current photoresists. |
|
| |
I'm not sure how much "ash" would remain after dust is combusted. Most ash is associated with minerals, and minerals make up a pretty small part of most organisms (not counting bones). I once read something about an experiment done in which someone carefully measured some soil, grew a tree in it over the course of several years, and then carefully removed the tree from the soil and measured it again. It had been diminished by something like two ounces. For a whole tree. |
|
| |
If the uv /oxygen supply levels go out of control |
|
| |
Ash from, say, a wood fire is almost always less than
three percent. Other materials (hair, skin cells) are
probably similar, and of course silicates (dirt, wafer
particulate) wouldn't burn at all. |
|
| |
That being said, it's very friable, so even on the stuff
that will burn you go from one largish (~50micron)
particle to 10-20 1 micron particles. Still more than
big enough to cause problems, and harder for the
filters to stop. |
|
| |
No, didn't you read [MechE]'s annotation? //it's very
friable,// |
|
| |
// Dust in a vacuum must settle out almost instantly. // |
|
| |
Ahem. "Mean free path" ... |
|
| |
That being said, the vacuum idea is not without merit. |
|
| |
// X-rays, gamma rays, electron beams or such. // |
|
| |
MBE & EBE have been around for decades. The issue is more one of productivity. |
|
| |
Eventually your species will work out the method of making semiconductors that every other sophisticated spacefaring species uses, but until then, carry on. We're laughing with you, not at you. Honestly. |
|
| |