Hacker News Reader

Comment by sevensor | Hacker News Reader

I worked at a semiconductor factory in the US for five years. I'm glad it wasn't in Photolithography -- that end of the fab reeked of acetone. My end had arsenic, but it was pretty well contained unless you went into the bead blast room. We also had elemental phosphorous, which has a tendency to explode if you look at it funny. Wets has hydrofluoric acid. CVD uses tons of silane, which is not only toxic but explosive. You really can't do without these chemicals if you want to have semiconductors. At best, you can stay away from the really carcinogenic resist chemistries. But just try to make an NMOS transistor without arsenic for the source and drain!

Edit: This is not to say that it's OK to destroy people's health to make semiconductors. Toxic chemicals are unavoidable in semiconductor manufacturing, and we need to handle them properly even if it causes a rise in prices.

View on HN

Eh, acetone smells gross but I can understand why they weren't too concerned about it. It's one of the strongest solvents out there that's actually fairly safe biologically. It's really volatile and humans can smell it down to a couple dozon PPM. Not a pleasant olfactory combination. However, your body produces it in small amounts so you can readily detoxify it.

When given a choice between various mostly dangerous solvents, I Choose Acetone™

I too spent some time working in an industry cleanroom in the past. (Specifics are probably still under NDA XD...)

All I feel comfortable saying is that I would have been a LOT happier if I were either tele-operating a robot to do the work or if the whole process were fully automated by a robot running a set of moves (and halting for remote operator if things fail checks).

>CVD uses tons of silane, which is not only toxic but explosive

Silane is even more delightful than that- a toxic irritant that turns to sand on contact with water. So silane gas that touches your eyes or gets into your lungs will deposit on your mucous membranes as fine glass. Blindness is a likely outcome of sufficient exposure, and if there is a really huge amount you can suffocate on blood and glass shards.

Another fun Silane fact for you - The auto-ignition temperature is right around room temperature / a bit lower, and pure silane burns with "invisible" colorless flames!

I worked at a semiconductor equipment manufacturer, and the silane gas delivery systems had to have separate flame detectors (from the normal one's we used) in the gas boxes for cutting off the gas supply because of this risk. We also had special strobe lights in the building with a sign below them indicating "Silane Pad on Fire if flashing", to let us know the gas bottle out behind the building (gas was piped in) was likely burning, since you wouldn't know from looking out the door / window unless you got close enough to feel the heat!

Luckily, I never experienced a silane fire, but for a point, I did design and implement the safety system for an experimental ion implant system, which used silane along with other nasty process gases containing arsenic, boron and other dopants. Cool job, but the stress of "make sure you get this right" was a lot higher than other software jobs I've had, haha!

This is easily the most terrifying thread I've ever read on HN.

Your comment reads like something from "Things I Won't Work With" by Derek Lowe.

Do you think the company you worked for wasn't handling them properly? Do you think the industry in general doesn't handle them correctly? If so, what changes do you think need to be made?

The company I worked for did a good enough job so that if you poisoned yourself they could plausibly say you were in violation of SOP. To be honest, I don't know what you could do to fix a lot of the dangers in semiconductor manufacturing. Many of them are just inherent to the process. Avoiding super carcinogenic resists is the low hanging fruit. Maybe there are a few other places where you could use safer, more expensive chemicals. But you have to have arsenic. You have to have HF. There's no known safe alternative.

I think the biggest thing you could do is remind the technicians that their safety is paramount, even when we're line down waiting for a critical tool to come back up. The pressure to restore a multi-billion dollar fab to production is intense, and it can cause people to cut corners on safety.

In your experience, did you find there was a independent, empowered worker-safety advocate in the decision line?

As you note, I seem most of the dangers not coming from dangerous chemical, but from management pressure to sidestep SOP in order to make quota.

>This is not to say that it's OK to destroy people's health to make semiconductors. Toxic chemicals are unavoidable in semiconductor manufacturing, and we need to handle them properly even if it causes a rise in prices.

Thank you for having a reasonable opinion. I'm reminded that they're rare every time the weld puddle heat soaks the non-asbestos glove on my left hand

I'm also reminded that they're rare whenever I go to buy paint for a surface that kids will never lick.

Paint, when it gets old, flakes off the out-of-reach surface onto the floor. The lead paint flakes are sweet like candy. As such, it makes sense to just not have leaded paint, period.

Also, the societal cost of more expensive and less durable paint seems a decent trade-off for people not developing impulse control issues later in life.

Acetone isn't very toxic. And it's quite biodegradable. I've had girlfriends who had acetone breath when dieting. But hey, I rather like the smell :)

Different strokes for different folks --- I know acetone isn't that bad for you, but I can't stand the smell. I have to be in another room when nail polish is being applied.

It's genetics, I suspect.

I notice this in sick kids. A decent fever and no eating and they smell this way quite quickly.

https://en.m.wikipedia.org/wiki/Ketonuria

Bead blast? Is that a type of abrasive blasting? Why? Some kind of annealing?

To sandblast the oxide layer off the GaS ingots.

Whoa there, cowboy. GaAs doesn't really make an oxide, and the last thing you want to do to an ingot is bead-blast it. You want to be in a room full of GaAs dust?

It's not for the wafers, it's for the equipment. Parts get coated up with various substances, and then have to get stripped back as part of regular maintenance. One of those substances is arsenic.

wouldn't robots solve this issue?

Also, a question for you: if chip factories were black boxes with robots in them that you put stuff into and got stuff out of, is the "output" also toxic? (Are the actual boards and such also toxic)? Or could the output be non-toxic as long as waste was handled properly, so that it's only toxic inside but neither in its waste output or product output?

How do you bootstrap a robot factory if you need the factory's output to make the robots?

maybe very slowly with the just a few workers wearing super expensive hazmat space-suits with their own self-contained breathing systems. that probably wouldn't scale to production volumes but could make sense if you're building a one-off robotic factory.

but nobody answered the question - are the normal outputs of a factory also toxic, or just what goes on inside? would a robotic factory even "solve" the issue?

You can't grind up a circuit board or chip and eat it (or drink it) - which down the line leads to problems:

https://en.m.wikipedia.org/wiki/Electronic_waste_in_Guiyu

That is to say, one would need proper recycling in order to avoid toxicity for the entire life cycle of products.

You'd have a similar situation as that of a nuclear power plant.

Well the "non-waste" output of a nuclear plant (what it actually produces) is just electricity, but are the chips these plants produce themselves toxic following production, or is only the production process toxic? I don't know about toxicity of these things.

Radiation is hard on electronics. Presumably some chemicals are too but it would seem easier to block a chemical contacting a part than it is to keep radiation out.

Robots help. The "lights-out" fab has been an industry goal for thirty years at least. (Lights-out = no human operators. I.e. you could turn out the lights and nobody would mind.) We've moved a long way in that direction, but we're not going to remove humans in the forseeable future. The reason is something your question, like many discussions about automation, completely overlooks.

Robots break. All the time.

Robots need recovery when they break. They also need preventative maintenance. Even a fully automated fab has an army of technicians in there turning wrenches 24/7. In fact, the more automated the fab, the more people you need turning wrenches. So you can't have a factory that's a toxic swamp inside so long as it's clean on both ends, because there are people in there. You really don't want to ask them to wear SCBA gear all day long either. Nobody likes that. It's hot, it's uncomfortable, and it doesn't work well in tight spaces.

And what's worse is those wrench-turning techs have a different mindset from the test-tube shuffling techs they replace. So when they come in, they're more likely to get exposed to something nasty.

I get the point of the phrase ("light-out fabrication") but I think it's important to note that robots smart/capable enough to replace humans would probably use visible light in a manner quite similar to the way it is used by humans.

Also, if you're unhindered by things like preventing tumors by isolating chemicals from your squishy meat employees, imagine the kind of Superfund sites you'd get when these places were finally decommissioned.

Reply on news.ycombinator.com