Sciencemadness Discussion Board

Dry Pressurized Oxygen Generator

kilowatt - 15-10-2007 at 18:14

So here's the idea. Aqueous lead nitrate is electrolyzed to yield lead dioxide and nitric acid. Lead dioxide is placed into a tube surrounded by heating coils where it is decomposed upon heating (decomposition takes place at just 290°C at 1atm) to oxygen and lead(II) oxide. The liberated oxygen should be dry and of rather high purity (depending on how well the lead dioxide was washed and dried) and could be used right off the decomposition vessel or connected to a larger tank which could be pressurized by the reaction. The spent lead oxide from the tube could then be added again to the nitric acid bath to reform lead nitrate, and the process repeated.

I am assuming that the neccessary decomposition temperature increases with pressure, so it might not be practical or even possible to use this to charge a cylinder to very high pressure. I haven't been able to find any information on this but I might assume the temperature/pressure curve resembles a typical vapor pressure graph. Nonetheless it might provide a convenient source of indirect electrolytic oxygen at high purity and quite useable pressure. Approximately 80lbs of lead would have to be cycled in this manner to equal an 80 cubic foot oxygen cylinder in storage capacity.

12AX7 - 15-10-2007 at 20:17

Reminiscient of Lavousier.

An awful lot of trouble. What's the difference between filtering H2O out of direct electrolytic O2, and filtering PbOx particulate and NOx gas out of the pyrolyte*?

*Pyrolyte: product of pyrolysis. Not sure if it's a word, but it makes sense.

Note that you only get 3PbO2 <---> Pb3O4 + O2(g), and maybe more Pb(II) at even higher temperatures (where Pb3O4 decomposes). It's not very efficient.

If you want pressure, you're going to need a very high quality compressor anyway. O2 at 5 atm is nothing to take lightly, let alone 200 atm. You might as well decompose NaClO3, or use zeolite or something like that to isolate it from air. (Or buy it. :P )

Tim

Sauron - 15-10-2007 at 21:00

Oxygen generators and nitrogen generators are both commercially available. They simply intake room air and seperate the )2 from the N2 using a RO membrane and store whichever one the unit is designed to output in a tank. The purity is above USP and good enough for many purposes. As I recall the air intake is filtered to remove dust and humidity.

This is intended to replace tank O2 and N2 for many applications. The N2 is probably good enough for inert atmosphere work but maybe not for GC carrier gas. The O2 is good enough for a lot of reactions that require O2.

These things are relatively cheap being in the $1500 range including an integrated compressor and tank.

Otherwise it's commercial cylinders with attendant expenses of regulators and gauges, drayage etc.

not_important - 15-10-2007 at 22:59

A bit more detail on N2/O2 separators.

There are two common class, those based on pressure swing absorption, and those on permeable membranes. Both types commonly are designed to run off of "shop" pressurised air, 5 to 10 atmospheres, and deliver their output at 1 atmosphere on up to nearly that of the input. As Sauron said, the air is filter to remove dust (and oil, or use oil free compressors).

The pressure swing systems have several identical absorption tanks filled with zeolite for the oxygen producing systems or carbon molecular sieve for the nitrogen producers. The air is passed through one tank, where N2 or O2 are absorbed, and the desired gas removed at the other end. When the absorber is nearing saturation the gas flow is switched to another tank, while the saturated one is pumped down and sometimes heated to removed the absorbed oxygen and other gases. Nitrogen stream purity typically runs from 95% to five-9s, while the oxygen generators produce 90 to 99 percent O2.

The membrane systems are similar to reverse osmosis desalination systems, however different membrane types are used and the pressures are much lower. Nitrogen, argon, and methane do not pass through the membrane well, while water, hydrogen, helium, carbon dioxide, and oxygen pass through the membrane. Typically the O2 stream contains 30 to 50 percent oxygen, depending on needs. The nitrogen stream runs 95 to 99,95 percent N2, depending on the desired purity, and has a dew point of -60C or better.

Membrane plants can have produce more gas per unit time the the PSA type, but at a lower purity. Membrane plants can also easily simultaneously produce both a N2 output and an oxygen enriched output. Membrane systems have no moving parts outside of the compressors, and can run as long as 200 thousand hours with only a 10% reduction in output. They are sometimes used as from ends for a PSA (N2 or O specific), which increases the time a tank can be used before regeneration is required.

----------------------------------------------------------------------------------------------

As already noted, PbO2 isn't a vary efficient way to make O2. On top of that the hot lead oxides attack glass and react with some metals.

Sauron - 15-10-2007 at 23:12

Yes, I think the ones I was looking at a few years ago were indeed pressure swing systems. At the time I was frustrated by not being able to puchase furan, due to shipping hassles. The prep of furoic acid (from which to make furan) required O2. I was not completely confident that the O2 was pure enough, so in the end I shelved the furan project till later. Not before buying a couple case lots of furfural, still sitting around. I'll figure out something to do with it one of these days.

YT2095 - 16-10-2007 at 02:04

make DMF with it, apparently it`s even better than Biodiesel.

Tacho - 16-10-2007 at 02:21

Quote:
Originally posted by YT2095
make DMF with it, apparently it`s even better than Biodiesel.


Could you elaborate? Good for what?

YT2095 - 16-10-2007 at 03:18

as a fuel to run engines on.

kilowatt - 16-10-2007 at 13:15

My original idea was to use the thermal decomposition of silver oxide to powdered silver metal for this; it occurs at even lower temperature and is more complete. The silver would be re-oxidized in powder form over an anode in aqueous NaOH. Then I calculated it would require 1500 troy-oz cycles to equal an 80CF oxygen cylinder. I'm not willing to sit there and recycle an ounce of silver 1500 times any more than I am willing to buy 1500oz of silver.:o I thought lead could make a worthwhile substitute but perhaps this is not the case afterall.

Quote:

An awful lot of trouble. What's the difference between filtering H2O out of direct electrolytic O2, and filtering PbOx particulate and NOx gas out of the pyrolyte*?


Well PbOx should be much easier to filter out than water, since it is non-volatile. I was just going to use a short condenser pipe where it could settle out, and a fritted filter or something after that. Any that did make it out into a storage tank would simply settle out harmlessly. I'm not too concerned about NOx in the product because this oxygen is for use as an oxidizer, for torches and stuff. Still, by thoroughly washing and then drying to PbO2 product very little nitrogen impurities should be present.

The other and more important factor is that electrolysis cells themselves require rather closely balanced pressure on each side of the cell or it will physically displace the electrolyte. The same problem applies to even a good membrane cell, as the membrane would be probably a thin sheet of tyvek which could not handle much pressure. I wish to obtain a usable, possibly even a storable pressure. This has to be obtained directly from the reaction since no sort of compressor I could obtain would be able to work with oxygen. The lubricating oil in them is a fire/explosion hazard in the presence of concentrated oxygen.

The thing that really kind of shoots this whole deal to hell is if only Pb3O4 is obtained at the low temperatures. To get the high yields needed to make this practical I need to get all the way to the +2 oxidation state. The oxide is bulky enough even then requiring 80lbs of lead to equal an 80CF tank. This requirement would go up to 120lbs with Pb3O4 as a final product. I assume in reality it would be a mix of the two, and my actual need before a practical regeneration cycle is far less than the amounts described. Anyways, if anything over about 400°C is needed it would really start to make this impractical as not only does it become more difficult to heat and insulate (hey, I could heat it with an oxy-fuel torch (kidding of course)), but the yield strength of the steel tube goes down so much at those temperatures that it cannot handle hardly any pressure safely. Corrosion by the oxygen will also increase drastically and a nickel plate is probably required in any case.

Quote:

If you want pressure, you're going to need a very high quality compressor anyway. O2 at 5 atm is nothing to take lightly, let alone 200 atm.


Absolutely, but 5atm is far more than enough pressure for any end-use for gaseous oxygen that I can think of (note this does not include liquid rockets which utilize cryogenic LOX). Torches and other oxygen fed burners (which is my main goal for this) only require typically 20psi max. If I wanted to make LOX I would obviously just condense it off an electrolysis cell.;) If I can get a couple atmospheres for less than maybe 350°C with the decomposition I'd still like to try this, but that doesn't look like it can happen at least not with a practical yield. Regular compressors can't handle the oxygen stream because of the lubricating oil in them. The decomposition reaction should be able to develop a couple atmospheres directly without too much trouble, but an oxygen separator fed by pressurized air sounds like a better approach.

Referring to oxygen generators:
Quote:

These things are relatively cheap being in the $1500 range including an integrated compressor and tank.

Cheap to who?:o A small business/shop maybe. I don't have that kind of money to be throwing into such things; I could get a hell of a lot of 80CF welding cylinders filled for $1500. Still, even they are expensive, on the order of $50 for a refill plus a high initial cost and all the trouble of a yearly cylinder lease. In contrast, the same amount of electrolytic oxygen requires only about 7.5kWh to make in a 3V cell. That costs something like $0.06. Even with an inefficient cell dropping 6V or more, I like that much better than the cylinder exchange. Of course as I mentioned before direct electrolytic oxygen is of little real use, as the cells can't develop any pressure.

It should not be difficult to make my own membrane type oxygen concentrator (if I could find a suitable membrane) but as not_important said, and as I have heard before, they are not good for high purity or high concentration. A manual (homemade, jerry-rigged) pressure swing system running off a shop air compressor could be almost as labor intensive to operate as say the silver based decomposition system I mentioned at the start of this post, since the sieve becomes saturated and needs regenerated frequently. With any system I use it is important that the pressurization comes before the concentration since as I mentioned I do not have a compressor that could safely work with oxygen. For now I will continue researching zeolite based seives.

12AX7 - 16-10-2007 at 14:36

So build your cell with one float to meter water and one to balance pressure. The H2 vent pressure tracks O2, and you could even use a phase control power supply (think lamp dimmer) to throttle electrolysis to regulate the O2 pressure.

You don't even need expensive tanks of acetylene or whatever, as you get the perfect fuel in direct proportion.

Note that you do need tens of kilowatts of electrolysis to support a large flame. Pumping it into a high pressure storage tank is a better option in that case, so you run into that problem anyway.

Tim

garage chemist - 16-10-2007 at 15:10

Decomposition of potassium chlorate would be quite a bit more advantageous than heavy metal compounds with regard to oxygen content of the solid and attainable pressure.
The KCl can be recycled in a chlorate cell. The resulting oxygen is very pure.
Also, the decomposition temperature of KClO3 would not rise a lot with pressure since KClO3 is thermodynamically unstable under all conditions.
You have to make sure however to heat above the decomposition temperature of KClO4 as well since a part of the chlorate will form perchlorate initially.

kilowatt - 16-10-2007 at 15:35

Quote:

So build your cell with one float to meter water and one to balance pressure. The H2 vent pressure tracks O2, and you could even use a phase control power supply (think lamp dimmer) to throttle electrolysis to regulate the O2 pressure.

I could do that, but a cell made of glass or plastic would not be very suitable even to get maybe 30psi. A cell with metal parts would be susceptible to hydrogen embrittlement. I suppose it is possible I could make coaxial cell with PVC pipe, though. That could actually be a good idea for pressures up to 250psi. That's still a little low for storage, but better than nothing.

I won't be able to supply enough power to sustain a decent flame by itself, though. With 12kVA maximum on a breaker it is really unlikely I will be able to beat 8kW or so of electrolysis power no matter what I use for a power converter. I do have a large number of old computer power supplies, which were cheaper and more efficient than a large 3V transformer though. A number of hacked car alternators in parallel, or a few similarly hacked car alternators running a 3ph high frequency transformer might do the trick, but I have yet to complete the infrastructure of such a thing.

Quote:

Decomposition of potassium chlorate would be quite a bit more advantageous than heavy metal compounds with regard to oxygen content of the solid and attainable pressure.

Couldn't that possibly explode? My understanding was that it is in no way safe to heat any chlorates to decomposition. On top of that, the idea of a molten chlorate in a steel (a pretty decent reducer at these temperatures) vessel makes me cringe.

Quote:

You have to make sure however to heat above the decomposition temperature of KClO4 as well since a part of the chlorate will form perchlorate initially.


I don't see why; perchlorate is not formed in a chlorate cell. What "part" of sodium chlorate is sodium perchlorate? The chlorate contains less oxygen, which is given off when the chlorate decomposes. Would that oxygen tend to react with more chlorate to form perchlorate or something?

I found somewhere that manganese dioxide is a good catalyst for this decomposition.

garage chemist - 16-10-2007 at 15:54

When heating (pure) chlorates to decomposition, a certain amount of disproportionation occurs:
4 KClO3 ---> 3 KClO4 + KCl
The amount of perchlorate formed depends on heating speed and time. With quick heating, almost no perchlorate is formed, with slower heating more perchlorate is formed, although always accompanied by simultaneous decomposition of the chlorate to oxygen. Therefore it is important to exceed the decomposition temp of KClO4 when oxygen is to be quantitatively liberated from KClO3 by heating.

Chlorates do not explode by heating, only when large amounts are heated there is a danger of runaway since the decomposition is exothermic.
Metals catalyze the decomposition. Especially MnO2 very strongly catalyzes oxygen evolution, and a ground mix of KClO3 and a few percent of MnO2 gives off oxygen already far below the melting point of KClO3, no perchlorate is formed then. The oxygen is contaminated by a small amount of chlorine, but this is easily removed by leading the oxygen through powdered hydrated quicklime (Ca(OH)2).

kilowatt - 16-10-2007 at 16:21

Is there any danger of runaway when using the chlorate/MnO2 mixture? I'm assuming if this is done in a long narrow pipe, it will be relatively safe from such events.


[Edited on 16-10-2007 by kilowatt]

garage chemist - 16-10-2007 at 17:46

I suggest you carry out small-scale experiments with this method of oxygen production to get a feel for the reaction. Can you get or make sodium or potassium chlorate?
A temperature sensor inside the decomposition vessel would be a good idea to get better control of the reaction speed, like stopping external heating if the temperature rises too rapidly.

kilowatt - 16-10-2007 at 18:11

Making sodium or potassium chlorate is rather easy in an undivided chlorate cell. The main problem is electrode corrosion, but I see there is a thread here on PbO2 anodes. Measuring the temperature inside with a thermocouple might be more of a problem since fused potassium chlorate must be conductive. It should work ok with the chlorate/MnO2 mixture though as it will be below melting. Some time ago I got a large K-type thermocouple for kilns. I don't know how it would like the oxidizing environment but I could dip it in some boric oxide before putting it in to protect it. I don't think it would react to form BCl3 or anything silly since that should require an additional reducing agent in the mix. If so I could try dipping it in borosilicate glass.

I see sodium chlorate decomposes at a lower temperature, but it is also less stable...

garage chemist - 16-10-2007 at 18:25

You could encase it in a narrow glass tube sealed at one end. The temperatures are no problem for glass, but the chlorine impurity in the oxygen will be a problem for unprotected wires.

kilowatt - 16-10-2007 at 18:55

That was the idea of dipping it in either glass or boric oxide. The coating should be thin and small enough as to not crack under thermal expansion. They would provide better coupling than the air gap of being inside a tube, which could be essential for monitoring this reaction safely. Otherwise I could run it into a steel tube that goes deep inside the reactor, with the very end filled with glass or boric oxide. I want good thermal coupling to get a good real-time reading. The thermocouple has ceramic supports/insulators on it except for the very end.

This reminds me of the thermal decomposition of ammonium nitrate to get nitrous oxide, a reaction which I have always considered more risky than it's worth. I suppose even that would be ok if it was managed in thin tubes with high surface area in a fluid filled jacket, but the consequences of a severe runaway are grave...

garage chemist - 16-10-2007 at 19:13

I have done the thermal decomposition of ammonium nitrate several times (glass RBF, heating with bunsen burner- takes some skill to keep the temperature right and is not suitable for more than 100g AN at once), works nicely, but the gas needs extensive purification (NO and NO2 removal) which is the biggest drawback.

There is a better method for N2O production which involves decomposition of AN in dilute HNO3 solution with chloride as catalyst, this is without danger of runaway. Search for microscale gas chemistry and go to the page about N2O.
I found the gas production with this method too slow to be usefull though. It is however todays standard method of producing N2O on an industrial scale.

12AX7 - 16-10-2007 at 19:34

Iron does not act as a reducing agent in the presence of molten chlorate. It makes an excellent crucible, in fact.

Tim

kilowatt - 21-10-2007 at 12:25

Is fused sodium chlorate not as strong of an oxidizer as fused ammonium nitrate, then? Everything I have found indicates that fused ammonium nitrate is an incredibly strong oxidizer and only a glass vessel is suitable for its containment. I would have thought chlorates would be similar, if not moreso. The chlorate should decompose below its melting point with manganese dioxide catalyst, but nonetheless this is a very oxidizing environment with some free chlorine present. I am after some 304 stainless pipes anyway for the chlorate cell, so perhaps some of that would be a more suitable vessel for the decomposition.

Even a small amount of chlorate stores a lot of energy, so I would like to be quite sure of what I am doing before attempting this. Even these "small-scale" experiments could cause a huge explosion if they were to get out of hand, such as in the presence of unsuitable materials.

If successful, it should take about 6.4kg of NaClO3 to equal an 80CF oxygen cylinder. This is opposed to 43kg of PbO2 with full decomposition to PbO.

[Edited on 21-10-2007 by kilowatt]

12AX7 - 21-10-2007 at 17:45

As you may or may not have read in the chromate thread, I once cooked off a whole pound of KClO3, resulting in its total, exothermic decomposition to KCl and oxygen (with some chlorine due to the small amount of acidic chromium I had added). It was quite vigorous, but not explosive.

I did this in our back yard, which happens to be shadowed by a large oak tree. If an acorn had fallen into the melt, I probably wouldn't want to stand too near.

Keep the reaction moderated with suitable cooling. A controlled supply of cold air over the reaction vessel should suffice.

Tim

kilowatt - 21-10-2007 at 21:48

No, I hadn't seen that. I just went back and looked, but could not find what sort of reaction vessel you used. Was it steel or iron?

12AX7 - 22-10-2007 at 05:29

Hmm, suprised if I hadn't mentioned it was a stainless steel crucible.

Tim

Oxygen from bleach

ciscosdad - 23-10-2007 at 22:00

Mellor Vol 1 P 354 indicates that dry Calcium Hypochlorite with a little Cobalt or Nicket salt admixed will yield Oxygen on heating with a "small" amount of chlorine contamination.
Does that help?
Unfortunately goes to CaCl2 so not cyclic.
And requires red heat.

[Edited on 24-10-2007 by ciscosdad]