Aluminum in UHV: Pros, Cons and Experiences

[beion]

I recently had some discussions about the use of aluminum in a ultra-high vacuum (UHV) apparatus. There seem to be those who believe Al is good in UHV and those who believe it is bad. I was wondering if anyone has some authoritative source for information on this.

Those who believe it is bad seem to say the oxidization of machined Al creates pores that can trap junk that then spoils the vacuum. On the other hand, I always thought it was the tight, tenacious oxide that makes Al good in a vacuum. I've heard that anodized Al, due to its thicker oxide, is good for UHV. And I have heard that anodized Al, due to its thicker oxide, is bad for UHV. I tend to think anodization will lead to a more porous surface and thus larger surface area, which is bad, but I don't actually know that. I have also heard that you can do a special treatment on Al to make it good for UHV, but no one seems to know what that treatment is.

In my own personal experience, I have used regular machined Al, cleaned as per standard UHV cleaning techniques (Alconox soap, acetone, IPA, N2 blow dry), in a vacuum system that achieved 10^-11 Torr. So, it can't be too bad.

What do you think?

 

[Dr Lots-o'watts]

Perhaps you could do the experiment. See how good a vacuum you can get with your current system + a sample of say, 1-inch cube piece of aluminum in the chamber, then see how good a vacuum you get with a 1-inch cube piece of something else or nothing. Make sure you pump for the same amount of time in each case.

 

[beion]

Doing the experiment myself is not really feasible for a number of technical reasons. To obtain reasonable pressures would require a multi-day bakeout procedure, so it takes awhile. On top of that, you can't have a particularly accurate control experiment, since baking a system twice will almost certainly get you a better vacuum than baking it once, even if you do nothing more than let the system back to atmosphere in between. It is also difficult to ensure the same baking temperature for two bakes, which has a big impact on the final achieved pressures. Finally, vacuum gauges at or below 10^-11 are not trivial and tend to be finicky things (I would not trust my gauge to give me accurate results).

Basically, this would be a rather involved test. I'm sure it has been done in the context of some larger-scale research initiative (like at CERN). But it is not something I would spend my time on. I was hoping to find someone else who did the leg work here.

** I should say, the pressures at which this issue would even begin to become relevant are extremely low, well beyond typical roughing or turbo pump pressures.

 

[Evariste G]

We use aluminium in our UHV system to produce aluminium nitride films, as far as contaminations goes we haven't had any serious contamination issues yet BUT, we have had a rather big problem in another regard:
The heating crucible for our electron gun is made out of copper, as you know copper oxidises easily so thise crucible is constantly covered in a layer of copper oxide - this underwent a redox reaction with our aluminium - dissolving the crucible and creating thermite.

We now heat using a simple tungsten filament with boron nitride crucibles.

 

[Dan Kennedy]

I've researched this topic pretty extensively for some UHV projects I'd like to work on. I want to use Al it's much cheaper and easier to machine than stainless steel, I have access to tons of it, and I don't have tons of money to throw at buying stainless steel UHV components.

Aluminum is actually considered to be one of the best materials available for UHV applications if it is used properly. Reasons being that it's extremely easy to machine (cost), it has a much higher thermal conductivity than stainless steel (reducing bakeout time/temperature), low nuclear activation (no contamination issues in nuclear/particle applications), and can have lower outgassing rates if properly prepared. The troubles are that you need to do surface preparation for low outgassing, aluminum is not easy to weld to UHV standards, and it's not hard enough to compress standard Conflat UHV flanges.

There are ways around all of these troubles. Aluminum's typical high outgassing rate is due to a porous Al2O3 layer which forms on exposure to air during the casting or extrusion process. This can be etched away chemically, allowing a very dense oxide layer to form which doesn't create trapped volumes and decreases outgassing by orders of magnitude. I know that accelerators at Brookhaven National Lab used an etching protocol based on strong acids. However, I've also seen papers where people have used pretty common, cheap detergents to do the oxide layer removal. Can't remember the chemicals, but papers are out there if you look hard enough.

Welding is also doable. However, the only way I've heard of to get a UHV tight weld is to use special protocols based on TIG welding. Anyone who has ever welded aluminum can tell you it's not easy. Getting a UHV weld is even tougher. But there are companies out there that routinely do this, so I have to imagine that you could eventually get it right with the proper amount of practice.

Final problem is creating a UHV flange from aluminum components. Typical UHV flange uses knife edges made out of stainless steel to bite into and compress a copper gasket. Aluminum is not hard enough to bite into copper, so it can't be used with Conflat flanges as is. Again, there are ways around it. Some people coat the aluminum knife edge with a hard material such as TiN (by vapor deposition) or nickel (by electroless plating). This hardens the knife edge enough to make it possible to bite into special aluminum gaskets which are unfortunately much more expensive than copper gaskets. The price difference is something like 10-20 times as expensive for aluminum gasket versus copper gasket. There are also compression fittings based on ISO flanges that can achieve a UHV seal with aluminum O-rings. Instead of the ISO wingnut fitting, they use chain clamps to provide lots of force evenly across a flange connection. Since they're not routinely used (except by Japanese semiconductor equipment manufacturers), the gaskets and chain clamps are not cheap. However, they allow UHV gaskets to be re-used, so they might be nice. Also, they can be used with ISO flanges which are much cheaper (or easier to make if you're doing it that way) than Conflat flanges. The final solution is to use an aluminum to stainless steel transition on the flange, so that the bulk of the flange is aluminum, but the knife edge made out of stainless steel is capable of forming a standard Conflat UHV seal. These flanges are commercially available. If you were looking into making things for yourself, you might be able to find plates of the transition metal available. It's probably not cheap, as the joints are made by so-called explosive bonding methods.

Here are some sites that you might want to check out.
Companies:
Atlas UHV - makes aluminum UHV chambers, Al/SS flanges, etc. http://www.atlasuhv.com"
Meyer Tool - makes aluminum chambers, etc. http://www.mtm-inc.com"
Evac - makes chain-clamped UHV seals based on ISO standards http://www.evacvacuum.com"
You could also check out any of the major vacuum suppliers (Kurt Lesker, LDS Vacuum, Duniway, NorCal, etc) to see if they carry these kinds of components. For welders, find one which can do UHV welding with SS and give them Al welding specifications. For coating, there should be plenty of places that do TiN plating.

References:
Atlas UHV and Meyer Tool have a fair amount of info available. Try:
Atlas references: http://atlasuhv.com/reference_info/aluminum/bib_al_uhv_chamber_material.php"
Meyer Tool oxide cleaning reference: http://www.mtm-inc.com/reduce_project_risk/cleaning_aluminum_for_vacuum_applications/"
Meyer Tool welding reference: http://www.mtm-inc.com/reduce_project_risk/hog_outs_versus_welding/"
Meyer Tool sealing reference: http://www.mtm-inc.com/reduce_project_risk/sealing_options_for_aluminum_vacuum_chambers/"

Sorry for the long post. Lots of info. If you need any more references, I could probably dig some up. Let me know.

 

[Dan Kennedy]

Just realized that you may not have been talking about using Al as a UHV construction material, just as something inside of a standard SS UHV vessel. If that's the case, sorry for all the unnecessary info.

 

[Questii]

Speaking of aluminum as UHV chamber material, the alumina layer if it is porus can be a problem as stated above, such as that of not-smooth or properly processed raw material aluminum from extrusion or casting.

But is the porus oxide layer a problem for UHV seal if you create the aluminum UHV chamber with all major sealing and internal vacuum surfaces with precision milling operations after say casting or even milling from a large solid block of aluminum? That is, without any chemical process involved in removing the more porus oxide.

In this case it would allow and expose the newly milled surfaces to air and allow for them to develop their own native passivation oxide layer, which hopefully is more compact?

 

[Dan Kennedy]

I'm pretty sure that aluminum oxide layer is going to be problematic if you mill. I can't give you a good reason why, but it stands to reason that, if you could get rid of the porous oxide layer just be milling, people wouldn't bother with all these wet etches. If you can burn through a bunch of aluminum with impunity, it might be worth a test. I don't have much hope for it, though.

 

[ZapperZ]

I recently had some discussions about the use of aluminum in a ultra-high vacuum (UHV) apparatus. There seem to be those who believe Al is good in UHV and those who believe it is bad. I was wondering if anyone has some authoritative source for information on this.

Those who believe it is bad seem to say the oxidization of machined Al creates pores that can trap junk that then spoils the vacuum. On the other hand, I always thought it was the tight, tenacious oxide that makes Al good in a vacuum. I've heard that anodized Al, due to its thicker oxide, is good for UHV. And I have heard that anodized Al, due to its thicker oxide, is bad for UHV. I tend to think anodization will lead to a more porous surface and thus larger surface area, which is bad, but I don't actually know that. I have also heard that you can do a special treatment on Al to make it good for UHV, but no one seems to know what that treatment is.

In my own personal experience, I have used regular machined Al, cleaned as per standard UHV cleaning techniques (Alconox soap, acetone, IPA, N2 blow dry), in a vacuum system that achieved 10^-11 Torr. So, it can't be too bad.

What do you think?

There's nothing inherently wrong with using Al in UHV. However, I would not use anodized aluminum. That tends to have a higher outgassing rate than expected. In the Varian vacuum training manual, it cites the outgassing rate after 24 hrs exposure to vacuum as on the order of 10^-8 torrliter/sec cm^2. Compare that to stainless steel at 10^-10, and you have 2 orders of magnitude difference in outgassing.

The other issue with Al is that if you have to weld it, then it requires a whole set of issues, and you have to make sure someone who is doing the welding knows how to do Al welding. This is not an issue with stainless steel or Cu.

I tend to not use any aluminum in my UHV system. It makes it easier to have a uniform set of material to deal with.

Zz.

 

[Questii]

I'm pretty sure that aluminum oxide layer is going to be problematic if you mill. I can't give you a good reason why, but it stands to reason that, if you could get rid of the porous oxide layer just be milling, people wouldn't bother with all these wet etches. If you can burn through a bunch of aluminum with impunity, it might be worth a test. I don't have much hope for it, though.

One of the links you provided described the cleaning a bit:

http://www.mtm-inc.com/reduce_project_risk/cleaning_aluminum_for_vacuum_applications/

But in it, it seems to refer specifically to extruded aluminum being a problem (although this does not absolve aluminum in general, even parts milled from aluminum blocks). But is the oxide a problem for aluminum chambers with freshly mill cut surfaces? Or maybe even a cast or extruded part but with the surfaces freshly cut to get rid of the 'dirty' oxide layer?

I guess the aluminum can also be slowly milled without coolant which helps reduce the surface contamination problem?

But I agree, one should use all means at one's disposal to make the fabricated part better quality, so etching should not be skipped. I was just wondering if even just cutting away the dirty oxide and let the native oxide reform, if it helps reduce the porus oxide issue (perhaps then milling needs to be performed at controlled conditions). Or maybe the cutting action of machine tools generates high temperatures and rough surfaces that leads to poor quality oxides for UHV. Of course, I also have no experience with Al UHV chambers, but I think cost wise and ease of manufacturing, they seem a better choice than stainless steel (weight too I guess).

 

[deepat]

I too think they seem a better choice than stainless steel due to weight, cost. However not sure about special treatment on Al

 

[thentangler]

Why can the gaskets be made out of Aluminum instead of copper? The stainless steel of the flange can easily bite into the Aluminum creating a good seal. Also Anodized Aluminum is bsetter than copper since copper oxidizes very readily when left outside. And I always have a problem with brand new copper gaskets left in the open for too long.

 

[ZapperZ]

Anodized aluminum has high out gassing rate and is never recommended in UHV. I mentioned this already. Scroll up!

Zz.

 

[thentangler]

What about the etched one which leaves a dense oxide layer as mentioned in one of the above posts?