Getting the water out

[Doug Coulter]

Water is the enemy of good vacuum, as anyone with a residual gas analyzer knows, especially if you have Viton seals anywhere, which allow a little through.
Well, I do have an RGA, so I know what works and what doesn't. Here is some info on what works, and what doesn't do much.

Starting off, here's an old table of drying agents used "back in the day". The physics hasn't changed since, so it's still good and matches what I measure here.

Vapor pressure of water, John Strong

Note the odd units -- a liter is less than a mole of gas in the ratio 1/22.4, and mg is a mg, so you have to do a little dance with units to get to percentages. One mg of water
is 0.000055556 mole of it (assuming 18 mol weight). One liter of a gas is 0.044642857 mole of it at STP. So one mg/liter works out to .001244454 or about 1.2%.

As you can see, even after some pretty good drying agents are used, there's still a pretty big partial pressure of water in the cleaned gas. But look at liquid air temperatures!
Now, that's going places. This has led a lot of people to assume that dry ice would be nearly as good, but sadly, it's not so great. At the temperature of dry ice-acetone mix,
which is about -78c the vapor pressure of ice is still e-3 millimeter of mercury (a little more in millbars, 1000 millibars == 760 millimeters of Hg) or so. So some of the inexpensive
chemicals are actually better than a dry ice trap.

So if you want a low partial pressure in your tank, for say D made in an electolyzer, you'd better dry it while the gas is at high pressure and have the fact that the proportions
stay the same as it becomes rarefied in the tank stay the same work for you. For that special case, we've seen noticeable (and bad, depending on what you are doing) effects
from as little as .3% water in the introduced gas.

There is a further trap with cold traps, often called virtual leaks. This happens if the level of coolant is allowed to change during use. As some of the trapped ice gets above the coolant,
the temperature there rises and the stuff re-evaporates into your vacuum. Most people work around this by starting at room temperature, pumping for awhile, then slowly
adding the coolant, a little at a time, so most of the junk is in the bottom of the trap, not right up at the top, but it is something to watch out for. If you could arrange for a slowly
rising level of coolant you would -- but you hit the top of the trap at some point....Cryo kinds of things bring their own problems to the table in general.

For getting out water that's already in the tank, full speed pumping plus heating is the usual way. You can bake to a few hundred degrees C in a fully metal gasket setup, but have to stay down around 100 C if you have Viton in the system, or it starts to decompose. Here we have found that bright short UV works better yet, and nothing
really works as well as the combination of UV, ions , and electrons from a big glow discharge, even if you have to let in something like Argon to get up to the glow pressure,
then pump it all back out. Of course if you are a purist and can afford it, you'd use the real process gas for this part too so as to avoid putting in anything that wasn't
what you were going to want in there later on. Water is a sticky polar molecule and is hard to remove, but deadly when it reacts with things in the tank, like tungsten
for example -- it erodes tungsten in a cycle that doesn't wind up eliminating the water. What the short UV does is simply break up the water into H's and O's and those are far easier to pump
out than the water was, and they will be carried out in the glow gas when that is pumped away. Ions and high speed electrons, ditto -- they break up the water so you are
only pumping some plain gasses, which aren't as sticky as water is.

If you can arrange for that glow discharge to evaporate or sputter some titanium onto the tank walls, so much the better -- Ti is a fantastic "getter" under those conditions and will tie up the water as Ti compounds. The only problem there is preventing metallic Ti from winding up on things that are supposed to be insulators, so you need to pay attention to the mechanical design so as to "shadow" those from the Ti atoms flying about, in more or less straight lines, so it's not impossible, just something you have to pay attention to. Barium and Calcium work well too, but are a little weird to handle for the average experimenter. They were use extensively in vacuum tubes which could be pre baked a lot hotter than any vacuum tank can be, and of course are a one-shot deal.

[lutzhoffman]

Hello:

I realize that this will only work with air, but its dirt cheap and easy, its the bone dry air from a scuba tank. I learned this from a local car buff who was looking for a solution for eliminating H2O from the air used to spray the enamel and clear coat onto his cars. The humid air here wreaked havoc with his finish, and refrigerated systems are just to expensive for him. Then he tried using the air from his scuba tank for the finish coat, and bingo, his best paint job ever.

Here you can get 80 cubic feet of bone dry air for about 3 bucks, it also works great for flushing your system before pump down, to lower the mechanical pump water vapor load, etc. Also if you ever need to mess with any hygroscopic materials, like NaI(Tl) the dry air from the scuba tank, can be passed though another chemical dryer to eliminate the last trace amount of H2O left, with no fear of overwhelming the drying agent, to feed an improvised glove box made from a 10-20 gal aquarium, some varnished plywood with 2 long rubber gloves, and foam weather stripping. At a flow rate of 5+ liters per min, one fill will last a long time. Just make sure all of the tools that you will need are inside. For me this means making a "dry run", and a written list, to avoid having to build an air lock pass box. If you do not own a tank, you can rent a full one for 10 bucks.

I know that there are many fancy ways to make dry air, but this one is so dam simple that it really can come in handy, with little investment. In addition you can get a free first stage from an old scuba regulator from any dive shop which services regulators. This will drop the tank pressure from 3000 PSI, down to a very manageable 125PSI.

On a related note a close friend recently got a 45 cal. scuba tank charged airgun from Korea. Very nice, hand rubbed walnut stock, nice blueing etc. You get 3 shots per charge before the POI starts to drop at 50-100 meters. The smoked pork that I had for dinner tonight, came courtesy of this amazing thing, which is why its on my mind : ) Maybe if I am real good, Santa will bring me one from Korea this Christmas, with the welded on silencer option. (Remember its not a firearm thus 100% legal)

[schneyolo]

Hello,

Is there any rules how should use this technics.(UV, ion bombardment, HV discharge)

I mean what order at wich pressure wicht technich is the best?

Best Regards

Attila Schné

www.ebeamguns.com

[Doug Coulter]

That's one of those questions for which the only answer is "it depends" (which stinks, but there it is). Heat is good, but not if it will ruin something else in the tank, or gaskets, for example. Same idea with a glow discharge used to blast water off the tank walls -- it may affect something else in the tank in a bad way. Patience always works if the system is leak tight and has no viton in it, just pump till you can't stand it anymore, but that might take a week. So without more information on what particular thing you're trying to do, no better answer is available than "try some things and watch for what works the best for your system and process".

For example, it seems you are doing mirror coating. So, some things, like heat, might make the mirror warp unless applied very slowly and evenly, and a big glow discharge might help condition the surface via sputter-etching, or that may roughen it too much if carried on for too long. Whereas, a nice loud UV source without heat might be good -- but I don't know of one. A cold trap on a side arm (so the cold doesn't make temperature gradients on your optic) will help, but not much more than a good pump does anyway. But I don't know if you have a good pump!

The idea of any of this is to cut the waiting time down so you can run more things through your system in less time. One trick that really helps is if you let the tank up to atmosphere to take something out, do it with some dry gas, and don't have the door open long while you make the switch. Even if some normal, wet air gets in, it takes time for the water to get stuck into things well, so a quick shot to atmosphere and back is far quicker than letting a system sit at atmosphere, then pump back down. With a week open to air, it takes a lot longer to get rid of the contamination than with a minute or a few seconds exposure.

So more details are needed or we can't really answer the question very well for your situation.

[schneyolo]

Hello,

Well my system is a Balzers BA500 equiped with its original oil diff pump and rotary wane duo 25.

The machine was originally builded with resistant evaporators. I started to rebulid it, it is going to be installed one of my new 4 pocket 6Kw e-beam evaporation source and a end hall hollow chatode ion source 30mm diam. opening.Also I want to install a 18W UVC lamp, the HV discharge soure is originally installed.Also a planetary fixture with 3 domes will be installed.And qaurtz qristall thickness deposition monitor.

I want to use the machine to coat mirrors with enhanched alu coating, and later broadband AR coating. The orings are viton rings.

I read in other articles the UV is wery useful to eliminate the water vapour, but from 10x10-4 torr pressure instead.

I planned that during the pumpdown about 0,1 torr I let the HV discharge to run for 15 minutes.

After this at lower pressure about 10x10-3 or little less let run the Ion source with Argon plasma for another 15 minutes.

And from 10x10-4 torr I turn on the UV for 20-30minutes.

After this let run the system until I reach 1x10-5 torr or so. And than start the aluminizing process.

I hope this helps for better understanding my system and my first question.

Regards

Attila Schné
www.ebeamguns.com

[Doug Coulter]

Sounds like a pretty nice system. Good oil in the diff pump is important of course, as is some kind of vapor trap on it for most uses, oil is a pretty nasty thing to have on your substrates in terms of coating adhesion.

I've only done R type evap, and some sputtering here, so I don't know how badly your E gun will respond to gasses in the tank when fired up, just haven't studied that, but a couple other people here might have useful things to say about it. I did make an E gun once, but only ran it in either real good vacuum or a bit of inert gas (so I could see the beam for experiments). I know JohnF and JerryB do this kind of thing, and Jerry is building a system for optical coatings. Joe Jarski is also doing some work with sputtering, which he hasn't said "what for" about yet, but his system is looking really nice.

One thing I've noticed here when doing evap is that Al makes a real good getter. I would watch the gages (ion and pirani, a Pfeiffer PKR-251) jump up on pre-heat, then go way down fast when reaching evap temperatures. To avoid messing up the thing I was coating, I put a shutter over the source. Doing a little evap inside the box did wonders for helping the vacuum, at least the easy to get gases. It might have been better if the Al vapor could hit a larger area of the tank, while still protecting the substrate until I was ready to coat it, dunno. Ti also works well for this, if you wrap some Ti wire around a tungsten heater and evaporate it into some large area sidearm on the tank, while shielding other things from the resulting deposition. Of course, as I found out the hard way, you have to be really careful to shadow all your insulating feed throughs from this kind of stuff. Since I have some windows on my tanks, I use a piece of sheet mica or just another piece of glass clipped in there on the inside side, so if that sacrificial stuff gets coated, it's not a big deal, and it can be either discarded or easily cleaned in acid and reused. No so with the thousand dollar door/window! So I protect that as well as I can.