## oxygen removal

imsmooth - 3-10-2012 at 14:03

I've tried to find a simple method to remove oxygen from air at a flow of about 3 scfm. I've considered burning and I've read about a toxic chemical pyrogallol.

I also am familiar with membranes, which are expensive. The packets used for food seem to be too slow.

Any suggestions?

plante1999 - 3-10-2012 at 14:08

Copper powder is the original procedure to remove oxygen from air. You might want to try to pass air on heated copper powder.
m1tanker78 - 3-10-2012 at 15:11

Could you be more specific? What will be the total volume of gas needed? If you just need an inert atmosphere, why not use argon?

Tank

imsmooth - 3-10-2012 at 18:13

 Quote: Originally posted by imsmooth I've tried to find a simple method to remove oxygen from air at a flow of about 3 scfm. Any suggestions?

From my original post the volume/min is given: 3 scfm

The source has to be from the air. I am not buying prefilled cylinders of gas.

How hot does the copper powder have to be?

watson.fawkes - 3-10-2012 at 19:14

 Quote: Originally posted by imsmooth I've tried to find a simple method to remove oxygen from air at a flow of about 3 scfm. [...] I also am familiar with membranes, which are expensive.
Seems like you might want to do some math first if you're worried about expense. Start with the molar amount of oxygen at STP (standard temperature and pressure) (hint: it's 22.4 L / mol), express your flow rate in liters per minute (1 cfm = 28.3 L / min), multiply that by the percentage of oxygen, assume that 1 mol of your reagent takes away 1 mol of oxygen, and then estimate the cost per mole on the ridiculously cheap side, say, 0.10 USD. I'll let you do the math.

Clue: Tank nitrogen is cheaper, by about a factor of five.

And now it's time for the standard question: What are you doing with this stream of deoxygenated air?

ItsAChitzen - 4-10-2012 at 07:04

I agree with watson... nitrogen or even argon will be WAY cheaper than trying to get some chemical removal via reaction with copper. The companies use pressure swing methods to get pure gas, something that you would have to spend an industrial amount of money to do. Getting the gas cylinders is super easy and cheap, and not weird. Where I live, super pure nitrogen is like 40 USD for a standard sized cylinder. This should last you until the sun collides with earth, unless you are simply buying the gas to vent to atmosphere lolololol...

[Edited on 4-10-2012 by ItsAChitzen]

Lambda-Eyde - 4-10-2012 at 09:26

 Quote: Originally posted by ItsAChitzen Where I live, super pure nitrogen is like 40 USD for a standard sized cylinder. This should last you until the sun collides with earth, unless you are simply buying the gas to vent to atmosphere lolololol...

Where I live it would cost about the same, but buying or renting a cylinder is about an order of magnitude more expensive... Refer to my signature.

imsmooth - 4-10-2012 at 10:49

I am making liquid air by compressing air. I want to get just liquid n2. I am doing this as a project for the sake of seeing if I can do it. Normally, this is an industrial process. I have the means to remove the co2 and water. I would prefer to remove the o2 before compression, rather than fractionally distilling it afterwards.

While I can use a nitrogen tank as a source for the raw material, I would prefer to do as much myself.

And, yes, I am familiar with the number of liters per mole at standard atmosphere and temperature.

unionised - 4-10-2012 at 12:14

I think the cheapest way to remove oxygen is probably to use wet iron.
Unfortunately the reaction is rather slow.

You can get round that if the total volume of nitrogen you need isn't too big.
If you can store air in the presence of wet steel wool you will scrub out the oxygen fairly effectively.
You would need to dry the gas after wards
You could also use the same wool at red heat and it would be quick but the "ash" would probably clog up the system.

zed - 4-10-2012 at 15:29

Well, you could try processing automobile, stove pipe, or other exhaust. Just convert free O2 and hydrocarbons, to CO2, CO, and H2O...by combustion. It might not be a completely hygienic process, but It could remove most of your free O2, right away.
watson.fawkes - 4-10-2012 at 16:49

 Quote: Originally posted by imsmooth I am making liquid air by compressing air. [...] While I can use a nitrogen tank as a source for the raw material, I would prefer to do as much myself.
If you can get a shop-built air-liquefier working, we'll gladly help you figure out how to get rid of the oxygen. But really, first things first.

Industrially, small-scale liquifiers use a pressure-swing adsorption unit to remove oxygen before liquefaction. The small medical oxygen sources are PSA devices, but they use a different adsorbent and different cycle timings to extract nitrogen instead of oxygen.

imsmooth - 4-10-2012 at 17:19

Quote: Originally posted by watson.fawkes
 Quote: Originally posted by imsmooth I am making liquid air by compressing air. [...] While I can use a nitrogen tank as a source for the raw material, I would prefer to do as much myself.
If you can get a shop-built air-liquefier working, we'll gladly help you figure out how to get rid of the oxygen.

I plan to document the entire process along with a tutorial so others can duplicate it safely. I will read about pressure-swing absorbers. I have considered making liquid air and storing the N2 that will evaporate first. I just don't have the means to store such a large volume. I also don't want to get into secondary pumps to push it through the compression/refrigeration unit.

My last project was a 12kw induction heater that a blacksmith eventually bought. That tutorial is here

unionised - 5-10-2012 at 04:00

Quote: Originally posted by watson.fawkes
 Quote: Originally posted by imsmooth I am making liquid air by compressing air. [...] While I can use a nitrogen tank as a source for the raw material, I would prefer to do as much myself.
If you can get a shop-built air-liquefier working, we'll gladly help you figure out how to get rid of the oxygen. But really, first things first.

Industrially, small-scale liquifiers use a pressure-swing adsorption unit to remove oxygen before liquefaction. The small medical oxygen sources are PSA devices, but they use a different adsorbent and different cycle timings to extract nitrogen instead of oxygen.

Actually, most of what they do to produce oxygen rather than nitrogen is switch the labels that say "Product" and "Waste".
But as far as I can tell, membrane systems seem to be more popular for nitrogen production.

watson.fawkes - 5-10-2012 at 06:34

 Quote: Originally posted by imsmooth I will read about pressure-swing absorbers. I have considered making liquid air and storing the N2 that will evaporate first. I just don't have the means to store such a large volume. I also don't want to get into secondary pumps to push it through the compression/refrigeration unit.
Look up the monograph Pressure Swing Adsorption, by Ruthven, Farooq, and Knaebel. It goes into all the engineering details needed to design and build one from scratch. Having looked into this a few years ago, the biggest part expenses are the solenoid valves and pumps. The biggest time expenditure would be developing control software.

For the scale you're looking at, though, I'd recommend refitting a medical oxygen compressor. All the plumbing, valves, and pumps are already there. You'd need to gut and replace the control system, and possibly the sensors. The solenoid drivers might be usable, depending on how the internal electronics are. Replacing the adsorbent shouldn't be a problem, since these are designed to be maintained and the internal canisters refilled.

FYI, I've had your induction heater project in my bookmarks for a while now.

watson.fawkes - 5-10-2012 at 07:00

 Quote: Originally posted by unionised Actually, most of what they do to produce oxygen rather than nitrogen is switch the labels that say "Product" and "Waste".
No. Thank you for playing.

Scientifically-accurate names for the output of a PSA unit are "raffinate" (product) and "desorbate" (waste). For an oxygen concentrator, the raffinate is air from which nitrogen is preferentially removed. The desorbate, though, is whatever the net adsorption is, which isn't just nitrogen. The adsorbent is only preferential to nitrogen, not exclusive to it. So the desorbate has significant oxygen in it. Raffinate recovery for small units is around 50% (give or take 10%), so the desorbate still has half the oxygen from the feed air, around 10%. This is why you have to change the adsorbent and the cycle if you want raffinate nitrogen, which uses an oxygen-preferential adsorbent.

Incidentally, one of the advantages of PSA for the amateur lab is that the argon content of air (~ 1%) comes out in the raffinate stream (at higher percentage). Separating liquid argon from LN2 isn't necessary for safety reasons (as it is with LOX), but it does provide a supplementary product.

unionised - 5-10-2012 at 08:22

That might explain why permeation systems seem more popular for N2 generators.
watson.fawkes - 6-10-2012 at 10:29

 Quote: Originally posted by unionised That might explain why permeation systems seem more popular for N2 generators.
The membrane systems also generate N2 as a raffinate. O2 diffuses through the membranes ~4 times faster and leave in the exhaust. The higher the flow rate (input pressure) through a membrane cartridge, the higher the non-nitrogen fraction of the oxygen. I checked the Parker website, a manufacturer of these things. One of the applications they highlighted was de-oxygenating mine tunnels. For that purpose, you don't need to get rid of all the oxygen, just most of it, and at the highest flow rates they operate at 95% nitrogen in the raffinate. They'll get up to 99.5% at significant lower flow rates through the same cartridge.

Membrane systems seem particular to raffinates which consist of the largest molecule in the input stream. That's the case for nitrogen in air. Oxygen concentration would require multiple stages for adequate pure product. On the other hand, one of the advantages of PSA systems generally is that if you can find a preferential adsorbent, you can remove it from the raffinate. For example, activated alumina within a PSA unit will remove water vapor from air very efficiently at ambient temperature.

The shortest answer is that these technologies have different area where each has advantages.