Sciencemadness Discussion Board

I finally won a cheap difussion pump

wg48temp9 - 7-9-2020 at 06:46

I have not been able to find the manual for the pump yet.
Can anyone tell me what the control knob at the bottom of the pump does?

mydiffpump-1.jpg - 52kB

I want to connect it to one of these:

20lreactor.jpg - 5kB

I will have to find some cheap diffusion pump oil next.

[Edited on 9/7/2020 by wg48temp9]

1KEE - 7-9-2020 at 20:42

>> Can anyone tell me what the control knob at the bottom of the pump does?

Temperature regulator for the heating element that boils the oil would be my guess...

You will need a “roughing pump” downstream of the diffusion pump.

Twospoons - 7-9-2020 at 22:10

I keep looking longingly at all the diff pumps on ebay - but the shipping to my country is always horrendous.

wg48temp9 - 8-9-2020 at 02:40

Quote: Originally posted by 1KEE  
>> Can anyone tell me what the control knob at the bottom of the pump does?

Temperature regulator for the heating element that boils the oil would be my guess...

You will need a “roughing pump” downstream of the diffusion pump.


Yes that was my guess specifically that it raises or lowers the heating element.

I have a two stage rotary vane pump for backing that I have resisted doing any vacuum distillations with. I will use the same pump for roughing with the ballast open.

Still no high vacuum gauge so I will have to make do with a discharge tube.
Connecting the 130mm dia port of the diff pump to any thing will be difficult.

A worrying statement in paper I read, suggested my pump is used with mercury which of cause I will not be doing that. Ideally I want a none combustible oil but that could cost more than the pump.

Twospoons@ Yes for something this size and weight the delivery charges could be huge. I was lucky its coming from Northern Ireland to the uk for just £20.

PS: ebay has silicone diff oil £40 for 250ml. I need the manual for the pump I will have to try calling the manufacture they can be helpful sometimes.

[Edited on 9/8/2020 by wg48temp9]

yobbo II - 8-9-2020 at 05:29


I wonder would brake fluid work instead of the (crazy) expensive oil

Yob

Sulaiman - 8-9-2020 at 06:49

Looks good, nice.
I'm guessing that there will initially be many re-assembly sessions, and mastering sealing, cleaning etc. so a silicone based oil is probably a good start.
YT vlog?

wg48temp9 - 8-9-2020 at 06:54

This is all I could find on my diff pump. Apparently it was designed for mercury and has three stages one of which is the ejector. I hope it can operate reasonable well with a suitable non Hg fluid.

Edwards-EM2.JPG - 115kB

macckone - 8-9-2020 at 09:23

Diffusion pump oil operates at 100-200C less than mercury.
Depending on the minimum temperature with the pump it may be ok.
If the heater is putting in too much energy it can of course cause the oil to decompose.

I would get mercury, it only requires 75ml (about 1kg).
Worst case, get some cinnabar and distill your own.

My guess is that in your jurisdiction mercury is going to cost more than the pump and high end diffusion pump oil is probably pricey as well.

Dr.Bob - 8-9-2020 at 16:43

I remember in ungraduate research work in a profs lab having a drying system with a huge Welch pump, then an oil diff pump, then a Hg one. I think it could go down to a level that was hard to find a gauge for (all Hg of course.) It was a nightmare to use and I was not allowed to tough it without help. By graduate school, we had way less Hg around, but more stills inthe labs than I had ever seen in life. They had a big fire a few years after I left, and then went to Alumina drying systems. How we avoided any big ones is a mystery as we had a lot of close calls.

But if you can addjust the system, it may work. If not, where are you, mercury is not that hard to find.

1KEE - 8-9-2020 at 17:45

BTW I’ve always wanted to find a glass Hg diffusion pump.... would be fun to watch the action inside :-)

wg48temp9 - 9-9-2020 at 02:25

Quote: Originally posted by macckone  
Diffusion pump oil operates at 100-200C less than mercury.
Depending on the minimum temperature with the pump it may be ok.
If the heater is putting in too much energy it can of course cause the oil to decompose.

I would get mercury, it only requires 75ml (about 1kg).
Worst case, get some cinnabar and distill your own.

My guess is that in your jurisdiction mercury is going to cost more than the pump and high end diffusion pump oil is probably pricey as well.


I can reduce the temperature relatively easily with a variac,dimmer or possibly a built in system via the knurled knob at the bottom of the pump.

I am in the uk where possession of mercury and its compounds without a license is illegal unless it a vintage apparatus. Curiously as the pump is about 50 years old or the design is, it could have been legally supplied with mercury. That is only my interpretation of the law.

Suliaman: Yes lots of careful checking, and leaking hunting. Previously I tried with a an old vacuum oven and a rotary vane pump. Never got much past a diffuse discharge never a dark discharge. I did manage to deposit a blue tungsten oxide from a tig rod heated with one turn on a toridal tansformer. I may produce a vlog (certainly not a mindless one with no explanation) or pics of interesting stuff.

Yobbo II: Most brake fluid contains corrosion inhibitors and other components to control viscosity and water absorption. I guess it could be triple distilled with long column but you don't need much high vapour pressure impurity to reduce the vacuum.

DrBob: I have tried to find some info on replacing mercury but not found anything specific. I did find a paper describing the properties of various diffusion pump fluids but nothing like advice on a direct replacement.

1KEE: I have never seen a working mercury diffusion pump but have seen a working mercury three phase rectifier, spectacular and how the heck did they make its huge glass envelope???

There are a few glass diffusion pumps on ebay uk at the moment.

I will probably by some DC 704 silicone oil or a relatively cheap equivalent. I don't have a butterfly valve to isolate the diffusion pump so I need a tough oxidation resistant oil.


wg48temp9 - 9-9-2020 at 16:32

1KEE@ You will probably like the following YT video from Applied Science showing his glass diffusion pump (sorry only silicone oil)

https://www.youtube.com/watch?v=SrNVLCHrJtY

glass=diff-pump.JPG - 53kB

Applied Science has some great toys, an electron microscope and even X-ray fluorescence gun to mention only two.

1KEE - 9-9-2020 at 20:20

>> 1KEE@ You will probably like the following YT video from Applied Science showing his glass diffusion pump

I watched that video and really enjoyed it! Other than the fact it’s not filled with boiling Hg :-) Applied Science videos are awesome, I’m rather fascinated with the kind of stuff that guy plays around with.

>> I have never seen a working mercury diffusion pump but have seen a working mercury three phase rectifier,

Wow, that’s quite a privilege! I’ve collected weird electron tubes for a long time and have yet to see a proper mercury arc rectifier, let alone a running one. I do have some 10EE Monarch lathes that use Xe thyratrons, and the older ones have small Hg/Ar thyratrons on the side. Those are cool. And I have a deuterium thyratron, a water cooled ignitron with like a pound of Hg, plenty of short-arc Hg lamps... but have never seen the mythical 6 (or sometimes 12) legged Hg filled “brain tube” (I call the big polyphase mercury arc rectifiers “brain tubes” cause it looks like an alien with its mercury brain boiling hahah sorry for me being goofy)!

>> and how the heck did they make its huge glass envelope???

I’ve thought about hiring a glass blower to make one.... and I’ve thought about using a large 3-neck round bottom flask, upside down, to prototype a simple mercury arc rectifier....

[Edited on 9-10-2020 by 1KEE]

[Edited on 9-10-2020 by 1KEE]

1KEE - 9-9-2020 at 20:27

BTW this is a good book to have handy if you don’t already... “A User's Guide to Vacuum Technology” by John F. O'Hanlon

http://www.dmf.unicatt.it/~gavioli/corsi/MSFM/ref/Vuoto+Came...

[Edited on 9-10-2020 by 1KEE]

wg48temp9 - 10-9-2020 at 07:48

Thanks for the link to the guide on Vacuum Technology.

Applied Science's active solder and ultrasonically assisted soldering iron was interesting. Apparently capable of soldering glass and other impossible/difficult to solder materials. It could be usefully for making glass to metal seals. Though I suspect a reel of it would cost more than my diff pump (£100 including delivery).

1KEE - 10-9-2020 at 13:14

Watched the video on ultrasonic soldering.... quite fascinating! Never heard of the process before.

Not sure if it would be applicable to glass-to-metal seals for high vacuum tho. For one, I’m not sure if the joint would be reliably vacuum-tight. For another, I’m having a hard time imagining a practical geometry for the glass-solder-metal connection: yes, you can coat the glass with the solder, and also coat the metal with the solder, but how do you connect the two? If you butt them up after coating with solder, and place in an oven, will the solder bond together without de-bonding from the glass?

Probably easier to do traditional glass to metal with Kovar, or knife-edge glass to copper (“housekeeper” seal if I remember correct). Tungsten to uranium glass is a fun one too.

Oh, another silly idea... look for old He mass spectrometer leak detectors. They usually have two roughing pumps, a diffusion pump, a cold trap, lots of plumbing and valves, a Pirani gauge etc. Or just use the whole unit as a portable high vacuum workbench. Veeco and Heraeus are common brands. I’ve obtained two for nearly free over the years, just gotta be in the right place at the right time.

wg48temp9 - 10-9-2020 at 14:59

Quote: Originally posted by 1KEE  
Watched the video on ultrasonic soldering.... quite fascinating! Never heard of the process before.

Not sure if it would be applicable to glass-to-metal seals for high vacuum tho. For one, I’m not sure if the joint would be reliably vacuum-tight. For another, I’m having a hard time imagining a practical geometry for the glass-solder-metal connection: yes, you can coat the glass with the solder, and also coat the metal with the solder, but how do you connect the two? If you butt them up after coating with solder, and place in an oven, will the solder bond together without de-bonding from the glass?

Probably easier to do traditional glass to metal with Kovar, or knife-edge glass to copper (“housekeeper” seal if I remember correct). Tungsten to uranium glass is a fun one too.

Oh, another silly idea... look for old He mass spectrometer leak detectors. They usually have two roughing pumps, a diffusion pump, a cold trap, lots of plumbing and valves, a Pirani gauge etc. Or just use the whole unit as a portable high vacuum workbench. Veeco and Heraeus are common brands. I’ve obtained two for nearly free over the years, just gotta be in the right place at the right time.


If the active solder wets the glass it should be air tight but it also needs to be sufficiently strong with withstand the deferential expansion of the glass and the metal.

I would expect to be able to join the pretinned metal to the pretined (active solder) glass with tin lead solder.and a propane torch. The copper tube would overlap the glass tube so as the joint cooled it would put the joint in compression.

i have tried a housekepper seal a few times but it always cracked as it cooled. I did have success using the part of a reed switch.

I did see a He leak detector on ebay a while back but I could not collect at that time .

1KEE - 10-9-2020 at 18:47

>> If the active solder wets the glass it should be air tight

I wonder if the bonding is very consistent or not... Be fun to look thought the glass at the joint and see if there's bubbles/voids/tracks. I'd be worried about it being 95% attached and the 5% that's not attached leaking.

>> but it also needs to be sufficiently strong with withstand the deferential expansion of the glass and the metal.

The fact the solder is somewhat soft maybe very helpful, it may have enough "give" to compensate for unequal expansion. That's how the Housekeeper seal works, glass and Cu have totally different coefficients of thermal expansion but the Cu is soft enough to accommodate the difference as long as the geometry is correct. That's why a housekeeper seal uses thin copper with the glass bonded to the thin edge of the copper.

>> i have tried a housekepper seal a few times but it always cracked as it cooled. I did have success using the part of a reed switch.

I've had similar results... then again, I'm terrible at glass-blowing and my experiments were rather crude. I did have limited success with tungsten to borosilicate glass, later on I got some uranium doped borosilicate but have been too afraid to waste it goofing around. My attempts at housekeeper seal were pretty terrible....

All this is making me excited about giving this another try... I have a leak detector to scavenge parts from, and a turbopump that I bought with a dead power supply and managed to repair and get running. Too many fun projects, not enough time.. But yeah getting a proper high vacuum bench set up is on my to do list.

wg48temp9 - 11-9-2020 at 04:41

My diff pump arrived today and in one piece. The knurled knob at the bottom looks like its the cap for the oil drain.

I would love to run it but that's a while away. The flange is a little too small to fit my flanged flask with a side arm. So I will have to fabricate something.
Perhaps a flat aluminium plate with a hose connector to start with.

IKEE Wow you have a turbo pump. That will give you a very clean and high vacuum with out any liquid N2.

Yes I think you should get your vacuum bench together if only so I can drool over pics of your turbo pump LOL.

wg48temp9 - 15-9-2020 at 11:00

Viton rubber sheet is available on ebay uk. Very useful for making vacuum gaskets. £10 for a 8in x 10in 1mm sheet.

See https://www.ebay.co.uk/itm/VITON-RUBBER-SHEET-GRADE-A-10-X-8...

1,3,3,5-Tetramethyl-1,1,5,5-tetraphenyltrisiloxane

wg48temp9 - 17-10-2020 at 04:04

My 704 diffusion pump oil arrived today from Lithuania about one month since it was ordered, £45 for 250ml. I need to check it really is 1,3,3,5-Tetramethyl-1,1,5,5-tetraphenyltrisiloxane.

704-oil-.JPG - 18kB

So I will check its boiling point at AP and its density. I suspect the benzene rings could be nitrated or possibly the silicon benzene bond could be hydrolyzed though because of the steric hindrance it may require high temperature.

Can anyone think of a chemical test I could perform to help confirm the oil's identity?


wg48temp9 - 17-10-2020 at 05:01

Quote: Originally posted by 1KEE  
BTW I’ve always wanted to find a glass Hg diffusion pump.... would be fun to watch the action inside :-)


Yes that would be fun. I think there is a video clip of one on youtube.

PS: I could not find one. NileRed has a mercury distillation video.
See https://www.youtube.com/watch?v=d6SahlAFV-w

[Edited on 10/17/2020 by wg48temp9]

Dr.Bob - 17-10-2020 at 06:17

I would guess that the silicon oil has a pretty unique IR, so that would be one way, an NMR would be pretty unique also, but not sure where the phenyls would show up. The oil will not work in most modern MSs as nothing to ionize, unless you have a fancy MS. Density might be pretty good check as you stated.

Good luck.

wg48temp9 - 18-10-2020 at 01:02

Quote: Originally posted by Dr.Bob  
I would guess that the silicon oil has a pretty unique IR, so that would be one way, an NMR would be pretty unique also, but not sure where the phenyls would show up. The oil will not work in most modern MSs as nothing to ionize, unless you have a fancy MS. Density might be pretty good check as you stated.

Good luck.


Thanks.

I don't have the possibility of taking a NMR. Out of curiosity I found the NMR at the following site https://www.azom.com/article.aspx?ArticleID=11531

Apparently because of the symmetry of the molecule the MRI is simple.

704-nmr.jpg - 23kB

The high boiling point will be a challenge for me. Measuring the density should be simple so I will do that first.

Panache - 18-10-2020 at 10:40

all ultracentrifuges have a roughing pump and a diffusion pump as part of the system.
all the cost of the of these devices are in the rotors.
So you can buy an ultracentrifuge sans rotor often very cheaply.
its what i just did. shame i wanted to hear the thing spin at 120,000 rpm, instead i have to make do with listening to pumps.
$200 i got mine for, labx has heaps without rotors for cheap.
has to be an ultracentrifuge, it uses the vacuum to attain the spin speeds.

wg48temp9 - 9-11-2020 at 05:29

While searching for a cheap gate or butterfly valve I found a scrapped API-365-Triple-Quadrupole-Mass-Spectrometer and got it for a great price.

two-t-pumps.JPG - 43kB

I think its got two turbo pumps two large bore valves and probably two vacuum gauges and various vacuum flanges. Of cause I don't know if any of them work. If one of the turbo pumps works and I can get the driver working it will make my diffusion pump redundant. Some turbo pumps apparently have a drag pump on the output which allows them to operate at an outlet pressure of a few tor which my rotary vane pump can manage easily.

The quadrupoles should be interesting but it may not have the electronics.