Filling gas flasks at home using a fridge compressor

I bought a very old and quite large fridge compressor which originally ran with R12 (CF2Cl2).
I wanted to know how much pressure this thing can develop.

With some hard solder, copper pressure piping, a 60 bar manometer and some bought parts I threw together an adapter that allows the pressurizing of common gas flasks and displays the current pressure.

I turned on the compressor with closed valve first and couldn't believe my eyes: the manometer climbed straight to 50 bar!
I filled the gas flask to 40 bar (at this pressure the pump volume stream becomes very small and pressure rises extremely slowly, also leaks started to show at such a pressure- the parts were originally designed for 15 bar), which took 5 minutes (volume of the flask was 0,5 L).
Then I released a hurricane in my garage by opening the valve.

That thing is awesome, I can now fill my gas flask with gases like hydrogen (gonna have to purge the pump and flask with hydrogen first before letting it build up pressure to eliminate explosion risk) and ammonia (this would actually liquefy inside the flask and allow for huge amounts to be stored).

To fill the flask with gas, I would collect the gas inside a freezer bag (after washing and drying), turn on the compressor (also connected to the bag) as soon as the bag is full until the bag is empty. Then wait again until the bag is full and turn on again etc.

I could also make a pipe-bomb autoclave and pressurize with hydrogen to run catalytic hydrogenations- 40 bar should be enough for reactions with raney-nickel if the reaction time is long enough.

Pictures of my setup available here:
http://versuchschemie.donau-it.de/topic,4697,0,-Gasflaschen+...

Very nice. Yes, refrigeration regularly runs in the 200PSI (low side) to 500-600PSI (high side) range.

1 bar = 1 atm = 14.7 PSI so that's circa 41 bar. Sounds about right

Tim

That's pretty impressive. As you say it opens up all kinds of possible syntheses normally considered beyond our scope.

I have some questions: 1) those "gas flasks," - are they what are called lecture bottles? What are they rated for in pressure and volume. Do they come in different styles, wall thickness, etc, depending on the gas to be stored? Where do you get them?

Can compressors be found that are oil-less for compression at these pressures? It seems that would eliminate contamination with oil - if this is indeed a problem.

Could empty propane tanks be refilled with the gas of your choice? I do not mean the big white BBQ ones, but the small ones we all have lying around from our torches.

Would chlorine be rather difficult to store as it would react with most metals causing destruction of the bottle, or would the reaction passivate the metal surface stopping all futher reaction?

Very cool(pun intended) indeed.

Throw on some phosphorous pentoxide it looks like. Or should I say... metaphosphoric acid.

http://www.worksafesask.ca/files/ccohs/cheminfo/cie85.html

Very cool! Looks like my fridge compressor will soon be going to work

I don't think refilling old propane cylinders would be a problem as long as the pressure is stopped at 250psi (17 bar). That is the pressure propane cylinders are designed for, bring it higher is a unneccessary risk in my opinion.

I plan to test my homemade black iron (steel) ammonia cylinder hydrostatically 1st and then with air (at a lower pressure) to check for gas tightness.

Following mick's advice I stopped at a brake repair place to ask if they had a used master cylinder. I couldn't believe it - they did and gave it to me for nothing -what luck on the 1st try! It even has a vacuum assist which I can hook up to my aspirator if needed.

I don't want to use brake hydraulic fluid as a test liquid and think that leaving water in the master cylinder can't be good as it would rust. So I'm wondering if the denatured alcohol sold at the auto parts place for brake system cleaning would be a good permanent working fluid, something I could just leave in the master cylinder reservoirs.

I have completed assembly of my hydrostatic testing device. I have 2 jpg pictures that I would like to show but can't get them inserted in my post for some reason.

I've been trying to use the "Attachment:" and "Browse..." at the bottom of the page. This pulls up the photo addresses OK but they don't show up when I "Preview Post." I'm new at this and some guidance from forum members would be much appreciated.

Now I have successfully filled my gas bottle with ammonia.
If you tilt the bottle you can feel the liquid sloshing around, very nice!

Pics of my setup available here, with detailed description in german:
http://versuchschemie.donau-it.de/topic,4697,60,-Gasflaschen...

I used 1,4L 20% NH3 solution, in four portions of 350ml each. Each portion took one hour to extract the ammonia from.
The dimroth condenser was run with icecold water, and the drying column (custom made by a glass blower) is filled with KOH. This worked very well, only the first 10cm of the KOH became wet and the upper 20cm remained dry, indicating complete drying of the ammonia.
The 20% NH3 solution begins to "boil" at 50°C and the boiling temperature gradually rises as ammonia is given off. I boiled until it was at 95°C, at which point the NH3 evolution becomes too slow to be of use.

The bottle got very warm as the ammonia condensed inside it, and the pressure over the liquid which was at 10 bar initially later rised to 20 bar due to this.

A 6L freezer bag was used as the intermediate storing container, it was connected to the gas generator and the compressor via a "T" adapter.

The brass valve of the bottle was attacked somewhat, the filling construction too, as seen in the last two pics.
Copper and its alloys are definately not a good metal for use with ammonia!

When I was done filling, I unscrewed the bottle from the adapter and a huge amount of ammonia came out. The copper pipe and the adapter became cold as ice, and smoke was produced from the intense cold!
The cotton oil mist filter was completely soaked in liquid ammonia and continued to smoke for a while.
My garden must have stunk like hell... but I smelled nothing, thanks to my fume hood.

Now I need dry ice to condense some ammonia and produce some solvated electrons by adding sodium metal!
I also want to make NaNH2 by adding a catalyst (iron compound) to the blue solution.

I have a pressure reducer designed for oxygen, I'll have to see how good it will hold up to ammonia. Maybe add a few drops of oil to minimize corrosion?
I tested it without oil and it produced a blue hue on the brass inlet pipe of the reducer.
I hope the valve doesn't get attacked.

When I bubble ammonia into water, it gets absorbed very vigorously, the bubbles collapse immediately as they form, producing a peculiar sound.

[Edited on 3-11-2005 by garage chemist]

Here's the 2 pictures of my hydrostatic testing device made from a brake master cylinder. Turning the 3/8" nuts gets me about 200 psi/rev. Although my pressure guage only goes up to 1000 psig I'm sure the device is good for at least 2000 psig.

I have it loaded with DOT3 brake fluid and will likely keep it that way and just fill my cylinders with water. The when done I'll just bleed some brake fluid through to purge the device of any water.

Thanks Tacho for your help with the attachments - it seems I had them all along but was faked out as I did not see them on the "Preview Post."

[Edited on 30-1-2007 by chemoleo]

Oops! Here's the 2nd picture for the above post.

I wanted to mention that the vacuum assist (that big bell shaped item) is really not needed for this application. You get plenty of force with the clamp like affair that I assembled from 1/4" (6mm) steel plate on the right side. This is driven (against itself) with 2 each 3/8" nuts on bolts about 5" (12.5 cm) long. As the screws are tightened the far right plate drives the 1/4" rod attached to the brake pistons.

[Edited on 3-11-2005 by Magpie]

Sorry for so many posts on this thread but I'm imbedded here by now. Be sure to see garage chemists last post where he has great pictures of his successful liquid NH3 bottling.

The picture below is a pressure test of my homemade high pressure reaction vessel. The fittings are "150 lb" which are normally considered good to 300 psig at room temperature. I inadvertantly tested it at 350 psig. Messing with high pressures is a little exciting. Makes you feel like a kewl again.

[Edited on 30-1-2007 by chemoleo]

Here's a final picture of my hydrostatic pressure testing: the testing of my gas (NH3) storage cylinder. As you will note I had to add a check valve (right before the pressure guage) to build pressure. With the other vessel I could do this with the in-line ball valve.

I accidently ran this one up to 430 psig.
Will now test to 150 psi with air to test for gas tightness.

[Edited on 9-11-2005 by Magpie]

[Edited on 30-1-2007 by chemoleo]

It is a good question Garage Chemist. For my normal household (water) pipefitting I use Ace's TFE (Teflon) dope. I've stopped leaks with it that Teflon tape wouldn't.

But for the ammonia I talked to a technician who maintains a large ammonia cooling system for an ice-rink. He said not to use the PTFE dope but get a dope especially made for ammonia. So I asked my industrial supplier and he recommended Leak Lock, which is blue. IIRC it is good for acetone also - it better be . I have also found a supplier of valves and guages specifically sold for ammonia - good prices too (agricultural supplier - not scientific). But I think any steel will be good. The main thing is to avoid copper and its alloys (brass) for any long term use.

[Edited on 12-11-2005 by Magpie]

I have an inquiry into Highline (who makes Leak Lock) about compatibility with acetone. Perhaps I should have used my own research conclusion which was that Gasoila-100 pipe joint compound would have been the best. Its specifications are pretty impressive.

[Edited on 12-11-2005 by Magpie]

I received a reply from Highline Chemicals on the applicability of Leak Lock (the blue paste) to -80C acetone for 1/2 day at 5"H2O WC pressure. He said that, over time, Leak Lock is affected by ketones. But at my conditions it would be OK. So I will give it a try. (He offered to send me free samples of Leak Lock for testing if I requested them on my letterhead ).

I finally have achieved success in pneumatic (air) testing of my homemade pressure vessels for gas tightness. I learned some things along the way: (1) dope both surfaces and use plenty of it (2) turn the joint down "all the way home".

When the tests were failing I noted that the pressure loss rate was similar on both vessels. This made me think that the leak was in the test manifold. So I replaced a suspect nipple in the manifold and remade other joints on the 1/4" piping.

I think I'm ready to go now and will assemble the equipment in the fume hood for NH3 liquifaction.

Here's a picture of the last test. I kept it in original format so people could read the data in my notebook if desired. Sorry for any seasickness.

Lest anyone call foul it is still at 142 psig after 4 hours.

[Edited on 16-11-2005 by Magpie]

[Edited on 30-1-2007 by chemoleo]

[Edited on 30-1-2007 by chemoleo]

NH3 liquidfied and stored today!

I knew this experiment was going to teach me a lot about high pressure and cryogenics - I just didn't know how much.

The liquifaction train is shown in the picture below. I used a 30mm glass tube 2 feet long for a drying column per garage chemist. I put 13 " of flake KOH in the tube supported on some SS scrub pad. Probably way too much KOH.

RBF is 500 mL. In the 1 liter dewar is a cold trap used to protect vacuum pumps. Sorry the picture is out of focus.

I used acetone/dry ice to get the temperature down to about -38C in both that dewar and an auxiliary dewar consisting of a wide-mouth Thermos bottle. In the Thermos I had the gas storage cylinder protected by a small poly bag from the acetone. This was at about -45C per #1 son who was helping me.

After about 1/2 hr of gas generation we could see the cold trap gas vent burp occaisionally. This worried me so I terminated the run. We couldn't smell any ammonia in the lab but sure could smell it outside. I conclude that I was generating NH3 too fast and/or didn't have my bath temperature low enough.

There was about 15 mL of crystal clear NH3 in the cold trap. (My target was 50 mL.) I poured this into the gas storage cylinder and screwed in a pressure relief (liquid) valve set at 350 psi. It is shown in the 2nd picture (next message).

The main thing I learned is to have all your acetone and dewars pre-cooled before you even start NH3 generation. I learned this the hard way by frothing over a lot of warm acetone when I added the chunks of solid CO2. Once you get everything cold the dewars do a nice job of keeping them cold.

[Edited on 18-11-2005 by Magpie]

In reference to my message above here's the 2nd picture:

Garage chemist: I will set up the apparatus (without loading the vessels) and take another picture.

When I first started the NH3 liquifaction I was disappointed as it was messy and high maintenance to get the temperature down. Now I know that my whole problem was due to failure to pre-cool my acetone, dewars, cylinder, funnel, etc. Once you get down to temperature it stays there. If I would have turned my heater down and kept at it I could easily have gotten the 50 mL needed to fill my small cylinder.

Would I do this again. Definitely, once I have a definite need. This was mostly a "proof-of-principle" experiment and to get some NH3 to upgrade my reagent shelf NH4OH. It is too expensive to do casually. I saved the acetone and can reuse it for this again. But I have to buy a minimum 2.5 kg of dry ice at $7.

Keeping the dry ice well insulated (say by wrapping with cloth towels) will keep it from disappearing too fast.

Was there any floating dry ice? Well, I didn't think so but when I poured out the dewar 1.5 hours later there was an active bubbling - so I guess there was floating CO2. Also the temperature was still down at -35C.

When I first started I was thinking how much superior your method is to mine. But then I kept thinking: "But you get to see the liquid NH3!"

Garage chemist here is a picture of the cryogenic components. The NH3 generator and KOH water absorber were pretty much the same as yours so I didn't show them. (If you need info on those let me know.) The cold trap is the smaller glass tube that fits inside the larger (30mm) glass tube and is sealed to it via ground glass fitting. This assembly is inserted into the 1 liter dewar flask. The thermometer has a range of -100C to 50C. Any questions just let me know.

[Edited on 30-1-2007 by chemoleo]

One last report:

I poured ~40 mL of 7N NH4OH into the iron reaction cylinder. Then connected this to the NH3 storage cylinder via hydraulic hose. When I opened both valves I expected to see about 110 psig on the pressure guage. It spiked up to 50 psig then quickly dropped off to about 20 psig. Over the next 1/2 hr it dropped to 0 psig.

The gas storage cylinder became very cold indicating evaporation of the liquid NH3. The reaction cylinder became warm indicating the heat of dissolution/reaction of NH3 with the NH4OH solution.

I titrated the product NH4OH at 10.9N.
It was rusty colored from reaction with the iron cylinder.

Although smaller quantities than desired were produced, this series of experiments met all of my objectives for "proof-of-principle." 2 pictures follow:

[Edited on 30-1-2007 by chemoleo]

2nd picture from post above:

[Edited on 30-1-2007 by chemoleo]

Have you ever noticed that if you back away from a problem for awhile the right (intuitive) side of your brain will often solve the problem? Well, I now know why the liquid ammonia was burping in the cold trap: I had it hooked up backwards!

A similar thing happended at work once due to the plant constructor mixing up two 2" pipes. This manifested itself through severe water hammer - lucky no one was hurt.