Material/vacuum compatibility; H2/D2

How to get to vacuum, what the classes are, and what is needed for what job.

Material/vacuum compatibility; H2/D2

Postby chrismb » Tue Dec 28, 2010 11:14 am

Similar to Doug's last, not quite vacuum tech but can someone please lay down the 'law' on what materials should be allowed to come into contact with hydrogen and deuterium. I am a little concerned over embrittlement and a bit more so over other chemical goings-on at the surfaces that we do not want.

I have done my searching and all sources suggest that 'copper and its alloys' are all OK. Then we get on to a plethora of stainless alloys, some good some bad.

Then there are also plated products, for which I can find no indications. I have bought some fittings for gas flow (which are actually designated for oil) one of which appears to be steel with a zinc plate and [I am assuming may have] silver passivation. I am not too sure about this part and am thinking that maybe I should get all brass connections; that [presumed] zinc plate part [between bottle tap and regulator] will remain under continuous D2 pressure for ever after until the bottle is empty and I do not know what will be going on in there, chemically or otherwise...

All comments most welcome...
chrismb
 
Posts: 620
Joined: Thu Aug 05, 2010 6:32 pm

Re: Material/vacuum compatibility; H2/D2

Postby Doug Coulter » Tue Dec 28, 2010 2:49 pm

Oh, this is vacuum tech to be sure. Maybe one of the more relevant parts of it.

Keep zinc out of your system, not for reactivity, but for vapor pressure -- it tends to get around and make a mess in bad places. Ditto any of the easy to evap metals. Copper has some slight issues, mostly with its oxides outgassing, but it does become very clean in a vacuum or D atmosphere eventually. Back in the day, a lot of brass was used as it's easy to machine, these days they say don't do it.

I have successfully used many thing that supposedly don't like hydrogen. It seems embrittlement is the main issue, and that seems to only happen under heat and/or pressure. So yes, just about anything one makes a fusor grid out of will get brittle. But it doesn't seem to matter to it working, either. Ti has worked real well, as has tungsten, as has graphite (the really pure stuff). Though I do see some hydrocarbons after a run with graphite grid parts. Not much, enough to notice though, on a good mass spectrometer, and they gradually get eroded, so that's where the carbon must be coming from. Haven't tried tantalum yet, but I will soon. It will embrittle, but -- where I'd use it, that simply will not matter and either it or Ti act like great getters for other gases in the bargain.

I've had decent luck with aluminum (6061), and found that if I hard-anodize it (no dye) it comes with fairly decent insulation from that, and doesn't outgass too badly to use. Melts easy, though.
In my case, I've used it for its thermal conductivity (lots cheaper and more machineable than copper) to carry heat away from hot stuff, and since I'm cooling the other end, no problems with that yet. I'm using a fat section now for one of my main HV feedthroughs, for example -- you can have something orange hot right near one end, but cooling one end keeps the hot end cool enough to survive fine.

An issue that has plagued me here is insulators becoming not-insulators via either sputtered metal onto them, or chemical reduction by hot D.
So an important thing to consider is how easy things are to either sputter or evaporate in these conditions.

So far, quartz and pyrex, as well as the commercial HV feedthroughs (a mix of alumina and other junk that makes them cheaper and easier to make) have all failed pretty rapidly, more a budget issue than anything else. I have had great luck with pure alumina from Accuratus when I make my own feedthroughs. I use a layered approach, with quartz inside as the real insulator, where it's protected from the hot ions hitting it at full speed (temperature) and that seems to be working out very well -- alumina isn't that great an insulator, but it doesn't reduce to metal under bombardment at least.

Just about every piece of plastic I've had in there has melted unless extreme measures were taken to keep it out of beams of anything and cover it up. Mica has also not fared very well, usually because it picks up enough charge on one side to "punch through" to the other side (I only get it as thick as I can get it), which then leads to other problems since I was using it as an insulator. It then burns up and gives off lots of gas doing it. Pretty in a sickening way...

For D supply lines I've used a mix of copper tubing, pyrex, and epoxy -- all are fine it seems, no issues there. Ds really leaky stuff so you have to pay attention, though. Takes better than water plumbing quality to have no leaks, and the stuff permeates through all the plastics I've tried (silicone, pvc, viton, buna, etc). You just can't use plastic pipe with it at all, even thick wall stuff -- it gets away.

I don't think the SS alloys that theoretically embrittle are going to see much effect from low pressure D at all -- unless you're going to get them to welding temperatures.

Here some ancient but nevertheless accurate info on sputtering and evaporation. As always, click the pic to get a bigger version so you can read it.

Sputter1.gif
Sputtering info

Sputter2.gif


Note, to our advantage, that most things don't sputter as well in the lighter gases. But beware of any heavy gas you have in there to do glow discharge cleaning or just contamination.

Evap.gif
Evap info


Here's a list of how hot in Kelivn (some C numbers penciled in) things have to be to really evaporate fast -- as in pressures around e-2 mbar for the metals.

So, you can see what to avoid, and what works well.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3023
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA


Return to Vacuum Technique

Who is online

Users browsing this forum: No registered users and 1 guest

cron