[Prev][Next][Index][Thread]

manifold cleaning; repumping merc tubes

Tom,

My manifold is an NW/KF-type clamp-together stainless and aluminum system,
with a 25mm bore.  

To clean it I dismantle it (unclamp everything) and run plain old
hardware-store-grade acetone through it. I have a bottle of the expensive
stuff (reagent grade acetone) but it's such a pain to get here (small island
in a big ocean with stiff shipping charges) that I don't use it. I haven't
had a problem that's obvious to me, using the cheap stuff. Folks like Mark
Snyder insist that only reagent grade leaves no residue, but maybe that's
important only at higher vacuum levels. I do have to degas the manifold for
a few hours to a day to get up to full vacuum levels and speeds however.

But then, I don't clean my manifold but once every 12-18 months or so. I'd
say my volume of actual pumping is medium to low -- maybe empty a one-liter
rare-gas-flask every two to three months on average.

Usually I end up cleaning it when I'm making some change to the system and
have it all ripped apart anyway (like recently, gas transfer system, YES!
Had to talk my wife into this one for only about 5 years, and then it was
piecemeal -- first the valves, wait a year...)

Also I did a couple of design things that help keep debris from going in in
the first place:
        1) Wide-bore "gravity" double-trap up front. Picture the letter "M"
in 25mm glass, in one plane. Coming out of the plane towards you on the far
left stroke, about half-way up, is the 25mm connection to the manifold. On
the other end, the far right stroke, is a 25mm piece that heads off to the
tubes, splitting into two and tapering down on each leg to 10mm and then
6mm. The bottoms of the two left strokes are sealed off into a ball shape,
which becomes the resting home of most of the powder and mercury coming
through (once dubbed by a friend as "bakutis", a name I love).
        2) stainless-steel-screen-type O-ring fitting where the glass clamps
to the metal manifold. Occasionally I find mercury collected here, obviously
still vaporized 'till it hits the cold metal.
        3) The main body of the manifold sits off to one side, away from the
flow. From the top of the diffusion pump to the glass trap on the front of
the manifold it's about 6 inches, and the main body of the manifold is about
18 inches long, teed off to the side.

        (I did try recently  to make a "mercury condensor" -- 5mm coiled
glass that sits between the tube and manifold  in a dish of water -- but got
rid of it since I didn't discern it collecting a lot of mercury, as well as
being an obvious hazard and quite the amazing vacuum resistor.)

There's only one main stopcock on my system, and it's a bellows-type
stainless contraption that comes apart with hex screws. I've only had it
apart once. The diffusion pump goes to atmosphere every tube cycle, not
ideal (or so I've read) but again, I take the Harry Truman "show me" stance
-- the tubes are clean, I don't have a problem with staining. In Holland I
used a turbomolecular pump for a year, and aside from being a sort of
gee-whiz device I wasn't too impressed with it, as far as pumping speed and all.

I also keep the pump running day and night if I've got much production going
on, helps keep the system degassed. 

The reason I ever went metal manifold in the first place was that we live in
earthquake country, and I've got no skill with pyrex, and I HATE lead-glass
manifolds (being the dutiful employee meant dealing with them for years, and
pleading for better). When I set up my own shop I had *loan* money to spend,
which always feels unreal, and so dumped it on equipment that I figured
would cut down on repair and maintenance  (no regrets, but I'm still
repaying that loan six years later). At least I didn't buy a Townsend copper
manifold -- I almost did, then happened upon an HPS / MKS instruments
catalogue... now I've heard many horror stories about the problems with
copper (or bronze, or anything but stainless and aluminum, unless it's in
the gas-transfer section of the manifold).

-Ted Pirsig

ps
What process do folks use for mercury repumps on older tubes? I finally did
a whole batch of experiments recently, 'cause it seemed about
every-other-one I attempted always failed -- they'd go green and turn into
greenish-brown perma-stain while bombing... I tried and crossed out  propane
repumping (made it worse), vinegar soaking (stripped the powder), HOT water
washing (still plenty of merc in tube), cold water washing, and so on,
finally settling on a short soak with ammonia (a la Ruby Red anti-dimming
technique, from Henry Brown at Transco) followed by distilled cold water,
then about 8 cycles of bombing, starting very mild and each time going only
until the merc got heated a bit, then sucking it out. Kind of a whole lot of
work.

These tests were done on 10mm EXTREMELY stained-out three-year-old tubes
(not mine!) from a job I took down -- I figured if a technique would work on
these, it would work on anything. The tubes had 12mm EGL 'trodes on them,
and black wherever the bender's heat had touched them.

Wayne Strattman suggests oven-baking -- which I can see  would work great --
but that's not in this year's budget (!) and then there's venting and
collecting the mercury and all.

I was motivated to learn a reliable repump method because two years ago --
for a time perioud of about two months -- I switched electrodes (reason-why
being another story) and used 12mm EGL premiums on 10mm glass. WRONG.
Couldn't get the 'trodes hot enough without destroying the glass, and didn't
take the time to put proper 'trodes on (it was some mightly intricate
graphics, the kind where the 'trode is inaccessable by the time you're
done). Ended up pulling the sign down last month due to much blackening on
the ends, and repumping every single merc tube -- only six of them, but with
the above process it took forever. BUT, it worked, and they all look great now. 

I also bought Masonlite 'trodes around that same time, and found  that 12mm
'trodes on 10mm glass invariably led to black rings in various places in the
tubes, right at the end of pumping. This despite following their pumping
instructions, having all the guages etc etc etc. 

Since I'd bought 1,000 of the Mason 'trodes I still use them, but only for
red tubes of the same diameter. Can't afford the experimentation on that front.