Passing the Gas

I know.  Terrible title.  I couldn’t help it.  Sometimes these things just come out.  See.  There.  I did it again.  If a bad joke can be made here, I just have to.  But I digress…  This is actually going to be about a gas transfer system for a neon processing manifold.

Traditionally, rare gases for neon and other scientific apparatus were supplied in glass flasks in volumes ranging from 1/4 to 2 liters.  They could be had in soft or hard glass (ie:  soda lime/leaded or borosilicate “Pyrex”) types.  In practice, one would insert a small iron weight in the connection tube, CAREFULLY turn the unit upright, then seal it onto the glass manifold system.  After fully evacuating the manifold and connection tube, all one needed to do is close the valves and lift the weight with a magnet.  Upon removing the magnet, the weight drops, breaking the seal and allowing the gas to be withdrawn from the flask as needed.  (note:  Some people install these horizontally.  Don’t.  Gravity is your friend.  Let it help drop that weight to open the little seal tube!  …and, gravity will also help keep the bits of broken seal from entering your valves with the initial rush of gas flow into the evacuated section.)

The glass flasks have a few inherent drawbacks.  Fragility is obvious but it goes beyond this.  First and foremost, they are at or near atmospheric pressure when full and so you only have the volume you see there.  (often 1 liter, for example)  Each time you withdraw some of the gas for filling a tube or running a process, the remaining pressure in the flask is reduced.  This results in a partial vacuum.  And….as we know…nature abhors a vacuum.  This increases the likelihood that air will seep past a valve or other  non-permanent connection point and contaminate the remaining gas inside.  In addition, if you are producing enough units or are using some processes that require higher fill pressures, then these flasks get depleted more frequently and must therefore be changed often.  It is also difficult to tell how much is left inside each time prior to using them so it is possible to come up short and have to swap one out to complete a job and this results in some waste.

The other option is to purchase gases in high pressure cylinders.  This eliminates most of the shortcomings, and in spite of the high initial cost, the actual cost per liter is greatly reduced and there is the added benefit of less contaminant risk and less frequent swaps.  There are various systems available that accommodate tanks of different capacities that can be adapted to existing manifolds.  This is not a new development, of course, but is one that didn’t really seem to hit the mainstream neon market in the US until the 1990’s.  Even today, after more than 100 years, the glass flasks are still sold and used by many…even me.

During the 1990’s, the shop I was working at decided to go with a cylinder system and purchased a unit from Eurocom.  This took some nifty little 12 liter disposable canisters that simply screwed onto a well crafted stainless valve body and manifold.  Since this was sold as a retrofit to the traditional glass manifolds in use, it was supplied with a thin stainless tube that terminated in a KF16 type flange and a mating flange of glass was supplied to weld onto the manifold.  This setup worked pretty well and while no system is perfect, it greatly increased our capabilities.

In more recent years I have wanted to do some plasma projects.  Such projects often require gases other than Neon or Argon or the Ne/Ar mixes.  Xenon and Krypton become common ingredients.  Sometimes with a halogen added.  But, I did not want to purchase flasks of Xe and Kr and install them in place of my Ne and Ar flasks as my main use is the neon work and inevitably I would have to remove the flasks of Krypton or Xenon to switch back.  This would be wasteful.  And… VERY expensive!  Xenon is especially costly.  So costly that some experimenters have taken old camera flash tubes and placed them in an evacuated chamber with a mechanism to break them open in order to use their contained Xenon for other projects!  It works.  But is not very elegant, is prone to contamination…and unless you have a large supply of old strobes to destroy, it too, is not cost effective.

For me, the answer is to go with a set of high pressure cylinders to add on to the manifold.  The plan being to retain the glass flask connections but then add connections for high pressure tanks as well.  Much as I did at that shop many years ago but with the intent of using the flask points for the Xe and Kr and the tanks for Ne and Ar.   Luckily, prior to ordering one or the parts to build one, a new bender (Thanks, Lee…and congratulations on entering the world of neon!) who was getting set up in Michigan was parting out a couple of old Eurocom pump stations he had ended up with but opted not to use in his new shop and happened to have a unit nearly identical to the one I had prior experience with.  The 12l canisters are still available from other vendors and while the cost per liter is not as low as with the big tanks, it is still better than the glass flasks, and perhaps most importantly for me, they fit better within the confines of my small portable processing bench.  A bargain was struck and the assembly sent my way.  After inspection and making a plan, I located and ordered some fittings to mate this to my existing manifold.

My existing manifold is an old Townsend metal manifold.  These were the hot number back in the day.  Compared to the glass manifolds they were meant to replace they were, if not a great improvement, at least an incremental one.  Rugged and easily adaptable.  But…  if you do get a leak, you will have a hard time finding it.  They also must be kept free of Hg as cleaning becomes difficult since the copper construction will react with it if a failure resulted in mercury getting in them.  (one way I resolve this is I do not do repairs on mercury units, and really, given the risks vs rewards, nobody should… just remake damaged Hg tubes instead, lest you risk becoming mad as a hatter.  For an amusing and somewhat related post about mercury, see, “Keep It Shiny”  http://www.novioljourneys.com/?p=926) In addition, metal manifolds pose a greater flashback hazard and you should absolutely never touch a metal manifold while the bombarder is on.  Actually, you should not touch ANY manifold-metal or glass, while bombarding…..but definitely stay away from the metal!  As a result, although metal, particularly stainless, is the preference in many industrial and scientific settings, they are not ideal in a neon shop environment and have mostly been replaced with nice Pyrex types that are put together with flanged couplings and compression fittings.  The best of both worlds really.  But the Townsend is what I have, and it actually works very well.

Adding a gas transfer system to the Townsend is pretty easy really.  In my application, I opted to attach it onto the right side connection point.  I will be keeping both glass flasks, repurposing them for Kr and Xe, and adding the cylinders of Ne and Ar (actually an Ne/Ar cold weather mix…since it gets chilly here)   This increases my options and gives a bit of an update to my setup.

I have opted to dispense with the little thin stainless tube that came with the Eurocom setup and instead use a few larger diameter pipes, tubes, and fittings to make the connection.  This will improve one of the shortcomings of the original design:  evacuation speed.  A larger diameter tube flows better and will pump down much faster than a tiny one.  The flow of gas molecules in high vacuum ranges is greatly enhanced even by small increases of diameter so this change should be beneficial.

So this is the plan of action.  I might have even completed it today save for one minor snag.  Nothing is ever easy.  It turns out that the o-ringed fitting on my manifold will take 10mm glass tube perfectly.  This is how I connect tubing to be processed.  All well and good.  But.  The only Ultra-torr fitting I could readily get was for 3/8″ tube…which is closer to the 9mm glass tubing.  No big deal.  I’m a glass guy.  Easy.  Just join a section of 10mm at the manifold to a section of 9mm at the ultratorr at the gas transfer system.  No problem.  Well, it would be no problem…IF….If I had some 9mm glass.  It is a size I do not generally use.  In fact, most people do not.  And up here, where I am one of only 4 or maybe at most 5 neon people?….fat chance.  And no, ordering a 25 pound box full just for a 6 or 8 inch piece seems kinda silly.  So, I put the word out to an online neon group that I am a member of and will obtain some from someone that way.  Another option would be to use a glass lathe and turn out a glass adapter that steps down to 9mm while still being perfectly round.  I may yet try this, but will need to check with a friend who has a lathe to see if I can pop in and give it a try.  Cobbling things up is the nature of the beast, especially up here.  It is just the way it works.  In fact, when I order in bulk from a distributor, unless I am willing to pay air freight, the lead time is usually 30 to 45 days.  If nothing else, living up here teaches patience.  But that patience also offers opportunity.  It allows time to refine an idea before diving into it with power tools and that can be valuable.  So when this is all done and I’m ready to pass the gas, you can be assured it was planned and intentional.  (I know, my jokes smell.  Sorry.)