Needle Valves

1. Comments on commercially available needle valves.
2. Details of a needle valve you can make yourself very cheaply.
3. Original enquiry to Molecular Dynamics list.

1 Commercially available needle valves

• The only one we've used is an edwards one (the top of the range one)....... there is also a Hoke one , but we've found it pretty flakey [ Megitt aviation made it for Edwards, but they have stopped making it now and I do not think Edwards have another source of these valves from the last time I enquired ]
• I've been looking at two possibilities recently: Parker Hanifin http://www.parker.com/ead/cm1.asp?cmid=325 and Hoke http://www.hoke.com/ both do finer control flows than swagelock, but they are cheaper than the UHV all-metal leak valve types. The Hoke valves are call micromites, and are in the range ?150-?200 ea. The Parker ones are their HR series metering valves, and the finest ones are H0 - similar pricing, but they look nicer cos they also incorporate a shut off feature.
• We use a Nupro/swagelok bellows sealed metering valve (p/n SS-4BMW-TSW) to adminisiter argon to a molecular beam source. I'm not sure what its flow rates actually are but it allows us to control the pressure in our source chamber right down to the limit of the diffusion pump (low 10^-6 mbar) before the valve is set at zero. In case this is sufficiently small flow and you've not seen them already, below is a link to the catalogue for this series, http://www.swagelok.com/downloads/webcatalogs/MS-01-23.pdf
• MDC sells variable leak valves within the price range you specified. If you check http://www.mdcvacuum.com/urd/uniface.urd/ec0070w.display?2.6.1 you may be able to configure the fittings and get an updated price online.
• It's not exactly a needle valve, but Granville-Phillips sells quite a nice variable leak valve. We've used themfor a long time.
• Parker Hannifin Corp. (a US company that is kind of a competitor for Swage) makes a nice fine metering valve - an example part number is 2A-HOL-V-SS-TC that is capable of controlling very low flows. (I didn't check our specs to see if the flow was in your range though.) The only caveat I would provide is that we used it for a mass spec interface and got quite a bit of background ion signal from the lubricants they use in the valve. If you're not using mass spec, not heating the valve, or are somehow very detection specific, you might find that they work well for you.
• I had the same problem several months ago. We wanted to bring small amounts of xenon and other gases into a vacuum chamber. I found also this Leybold valve and heard from several people, that this valve is of very high quality but is also very expensive(in Germany I think around 1600 Euro). So I make some tests with the valves we had from HOKE. My first experience was, that with a good feeling for turning knobs its no problem to adjust the pressure in the chamber from 2*10^-7 mbar to 1*10^-5 mbar with a good precision. I calculated the flow and it was a little bit bigger than the region you mentioned, but I compensated it with sensitiveness. This was the Valve "milli-mite" from HOKE with 6mm metric Gyrolok connection(in Germany around 120 Euro). There is also a "micro-mite" called version from HOKE I ordered a few weeks ago. This one is around 200 Euro and should be exactly in your region of flow. Unfortunately I can't test it until now, but we have the same one at another vacuumchamber and there it works very good. The Disadvantages: -the "micro-mite" and "milli-mite" valves can't close the line completely - you can easily damage the package where the needle goes in if you use too much force when screwing - so you should use a second standard valve for closing the line. -both valves have several materials in vacuum who can gas out. We connect them to a vacuumchamber and take them under vacuum for a week or so, then they work ok for us. Because of the big price difference it was the right choice to use this valves from HOKE. The Leybold valve is mainly used at UHV chambers and there it should be the right one. Let me know what kind of valve you decided to buy.
• try the Balzer valve UDV235 series. you can either operate it manually or automatically w/ a PID circuit.
• There multiple different options you could consider. There is a company based near Pittsburgh, Pennsylvania in the U.S. called Lesker. They sell numerous different valves. The website is www.lesker.com and if you look under "Gas and Fluid Management", then "Fluid Management", then "Gas" and finally "Needle Valves", several different options will appear. You can search the website in order to find some other alternatives. The prices of most are under $1000 and will be sufficient to control you flow rate. Two specific needle valves of interest will be Model LV10K, Part # ED-C37 102000 and Model FCV10K, Part # ED-C37 101000.

2 A needle valve you can make yourself very cheaply

• The best needle valve I know is made by squeezing a hardened copper capillary ( 1 mm id and 3 mm od ) between a brass plate and a brass wedge. The capillary needs to be well cleaned before use. This valve controls the flow down to very low values and can be changed in setting many times before the capillary gets "tired". It is the best valve I know vis a vis contamination and dead volume but it cannot, of course, be closed completely and needs a secondary valve set up to empty the upstream volume separately. It is, I believe , described in a book on vacuum thechniques edited by John Yates. G.S.

• What we use to control very small fluxes is a "pinch valve": a piece of copper tubing (1/8 inch diameter) squeezed between two brass wedges. Two screws hold the wedges together, and the amount of tension on these screws, through the elasticity of the copper tubing, determines the flow. (notice that the elasticity of brand new copper is quite low; it needs to be bent a few times to work-harden).
  Very small flows can be achieved thus, with surprising stability. One of the main advantages is that there is no contamination at all. Also, it is practically for free.
  We use this to control the flow of various gases into pickup cells for helium nanodroplet isolation spectroscopy. Roman

• Greetings: We make our own "pinch values" to produce such low flows. They are not "reproducible" but very stable and cost almost nothing. They involve a controlled crushing of a small Copper tube. Kevin

• The procedure to clean the tube is as follows ( if I remeber correctly of course! ): First you flow nitric acid through the tube and then wash, first with water and then with alcohol to make it dry. The hardening of the tube ( also necessary to give it some elasticity ) is done by heating it up and then quenching it, dumping it, while still hot, in clean cold water. Sincerely I do not remember if the quenching was done before or after the cleaning. Probably it can be done either way.
  When I used first in the sixties we called it the KISTEMAKER valve, from the name of Jaap Kistemaker who was ( for many years ) the director of the FOM lab in Amsterdam whom we belived to be the inventor. I was recently told that Prof. Kistemaker , now approx 90 years old, still goes regularly to the laboratory, so , if you do not feel that this thing is being blown a bit out of proportions , you could try writing to him for a proper quotation of the real source. By the way, one of the main reasons of existance of this valve is that does not allow for fractionation ( isotopic or otherwise ) to arise because of the ( molecular ) flow out of the narrow orifice like a normal needle valve would instead allow for. This is because the flow in the capillary just before the orifice does not allow the back diffusion of any concentration gradient that may form because of the fractionation. Giacinto.

2.1 pictures of this needle valve

3 Original enquiry to Molecular Dynamics list