Sciencemadness Discussion Board

Ostwald style nitric production

 Pages:  1  

Chemetix - 15-12-2016 at 14:16

After much searching on the site I found lots of postulating and concluding that Ostwald style nitric is all but impossible without the use of platinum. And as such all but impossible to do back yard style, hence the considerable number of posts in the Birkland nitric thread. It seems more doable but not very practical.

I have had some recent success with an NO2 generator using urea as a source of ammonia and oxidising with air using cobalt carbonate on sand as a catalyst.

Now comes the next step, and some questions; is it better to just let the effluent gasses from the NO2 generator condense and form HNO3 with the water from the ammonia oxidation, or absorb them with a salt like sodium carbonate to form NaNO3 with a little NaNO2. Or continue to process the nitric acid to form 70%HNO3?

Can you use NO2 directly in some cases, instead of HNO3 and not bother about processing the 70%HNO3 to concentrated nitric?

Magpie - 15-12-2016 at 15:48

I have never done this so these are just my opinions:

I once worked in a place that recovered NO2 from a calcination. It was converted to HNO3 using a pressurized absorbtion column. This seems difficult for a home chemist.

However, if you could recover a dried (or damp) nitrate salt, say NaNO3, or possibly even better, Ca(NO3)2, then this could be treated with con H2SO4. This should allow you to recover fairly concentrated HNO3 by distillation.

j_sum1 - 15-12-2016 at 16:32

IIRC (and I may be mistaken) the OP requires elevated temperatures and pressures as well as the catalyst -- around 5-10 atmospheres and greater than 250°C. Then a considerable volume of NO needs to be fed back into the process. (I seem to recall 90% feedback and 10% throughput.) And the catalyst does degrade relatively quickly -- hence the standard addition of rhodium to the platinum to make it more durable.

Together this presents a number of significant hurdles for the amateur chemist. It is not just finding a cheap catalyst substitute. The engineering looks to be challenging.

If you can come up with something workable, (a) awesome -- well done and (b) you will find a lot of interest in your work and maybe a few who would like to replicate it.

Chemetix - 16-12-2016 at 02:56

Industrial scale nitric is going to use techniques that milk every last mol and kilojoule. In trying to compete with industrial efficiencies I'll admit defeat right now. However, the small scale has it's advantages. For one thing, the ability to cool the NO2 down and get more N2O4 in the effluent gas is a gain in efficiency, just the ratio of the volume of the apparatus compared to its' surface area means passive cooling is going to be more effective.

In the SM library I found Absorption of Nitrous Gasses, a 1933 text but an exhaustive tome on the dynamics of NO2 absorption and quite instructive.

3NO2 + H2O -> 2HNO3 + NO
Is the critical reaction and implies the feedback step necessary to make the most from the left over NO.
Choice- do you feed it back to the NO stage before the air inlet to get reoxidised or do you let it get absorbed in a tower arrangement and make nitrous acid (4% @ 25C) which can get oxidised with air to HNO3.?

A set of towers with broken glass to increase surface area would probably be the simplest engineering solution to the absorption of the left over NO. Because the reaction in the tower: 2HNO2 + O2 -> 2HNO3
would continue to convert the NO into HNO3.

Industry would prefer this reaction to occur:
N2O4 + H2O → HNO2 + HNO3
and the subsequent
2HNO2 + O2 -> 2HNO3
in the tower means no need to pump back your NO, but only if you have N2O4 which is favoured by lower temperatures (higher pressures).

This would mean costly active cooling (or pressure) and power consumption in industrial chemistry, but achievable by passive low energy schemes on small scale, which means more N2O4 at lower cost on the benchtop. And when you consider you are making nitric, despite the inefficiencies, you are going to be still miles ahead in terms of of buying retail quantities of acid. So what, you let a bit of NO go up the tail pipe, you've probably got 80% conversion to nitric from urea. An industrial operator would be put against the wall for that sort of efficiency, but a back-yarder? Laughing!

I figure 6n Urea makes a theoretical 11n HNO3 with reoxidation of the formed HNO2 to HNO3. Combined with the left over water from the oxidation of ammonia after conversion to HNO3 ( including the reoxidation of NO or HNO2) there should be 12n H20

11n HNO3 +12n H20 should be around 76% HNO3 by weight

Even if you achieve the azeotrope with water of 68%, this process is looking pretty good.

Aaaaarh! The catalyst, how long can you run the catalyst for? If you are using Pt/Rh at nearly 1000C you are going to have some problems with burn through of your gauze plus the volatisation of the metal.
Cobalt oxide on silica (or other refractory) I think is going to cope admirably. I've identified that cobalt seems slower in it's conversion rate and as such won't generate the rate of localised heating to keep the catalyst bed at self sustaining temperatures. A bitch for industry, but a few tens of Watts to keep the reaction zone at 700C for a pilot plant is nothing. I thought about using the waste heat from the reactor to power the urea degenerator. But I don't think there's even enough to do that at small scale.

I think Magpie's suggestion of CaNO2 might just be the best way of concentrating and using the NO2 without concentrating the nitric acid.

I'll try an absorption tower and see what is possible.

[Edited on 16-12-2016 by Chemetix]

j_sum1 - 16-12-2016 at 04:42

I like the way you think Chemetix. It all sounds good.

Chemetix - 17-12-2016 at 18:30

If I were to add a small amount of SO2 to the NO-> NO2 chamber, I'm thinking that NO2 + SO2 -> SO3 + NO would result.
Net effect is the NO would reoxidise, SO3 would suck up the water and become H2SO4 and leave you with Conc. HNO3

A byproduct of the process is concentrated H2SO4.

Quel dommage!

Fleaker - 21-12-2016 at 09:08

I know that at large silver refineries (like that of Eastman Kodak) where they dissolve metric tons of silver to make photograde silver nitrate, all of the nitrogen oxides go through something called a Johnson Scrubber. Which is just a big packed bed scrubber filled with dilute nitric acid and like 10% hydrogen peroxide. I think this is done to firstly convert any NO to NO2 which then becomes HNO3. The scrubber solution gets put into the next batch for dissolving the silver nitrate. Any leakage gets mopped up with alkaline sodium hydrogen sulfide solution.

Once I did an experiment with an autoclave and 75%Ag 25% Cu (m/m basis) and 30% w/v nitric acid. I loaded a stainless 2 liter Parr with the metal, put the acid in, and closed the hatch. The pressure of NO2/NO rapidly rose to several hundred PSI. I then introduced pure oxygen into the reactor and measured an exotherm and a dramatic pressure reduction to near ambient pressure.

I tried again by putting the metal in, pressurizing it all to 50 psi O2 and pumping in nitric acid. The pressure never got above the oxygen overpressure.

Chemetix - 21-12-2016 at 13:57

I wonder if under the high pressures you didn't end up making something like N2O3 or N2O5?

WGTR - 21-12-2016 at 15:07

Quote: Originally posted by Chemetix  

In the SM library I found Absorption of Nitrous Gasses, a 1933 text but an exhaustive tome on the dynamics of NO2 absorption and quite instructive.

I bought an original copy of that. Had to have it shipped from Australia to the U.S. I considered it well worth the price, and it occupies a place of honor on the bookshelf.

Quote: Originally posted by Chemetix  

I'll try an absorption tower and see what is possible.

If you want to, you can adsorb NO2 onto silica gel.

The NO just passes through unaffected. It's best if the gasses are dry, since water is also adsorbed. In this way, dilute streams of NO2 can be concentrated, allowing an easier conversion into nitric acid, and at a higher acid concentration. Also, the silica gel is a handy way of storing NO2. To release the NO2 vapors, the silica gel can be heated up. In the linked video, the gasses weren't dried, so some water was also getting adsorbed.

I'm working (off and on) on a Birkeland-Eyde reactor design. By drying the gasses going into the reactor, allowing the exit gasses time to oxidize, and then adsorbing the NO2 onto a silica gel column, I'm able to easily measure the efficiency by comparing the before and after weights of the adsorption tube.

Since you're producing water in the reaction (?), perhaps that could be adsorbed in a drying column right after the gasses cool down, before the NO has time to oxidize. After time spent in the oxidation chamber, then the resulting gasses could be adsorbed in a second column, that could be weighed...? Just an idea.

silica adsorption

Chemetix - 22-12-2016 at 02:03


If you want to, you can adsorb NO2 onto silica gel.

I saw your silica gel technique and have been ruminating away on the idea. Brilliant strategy I have to say, it offers so many possibilities? Can you cool down the NO2 soaked gel and the NO2 dimerizes in situ? How much can you store on the gel? Being able to analyse what's going on has been a concern.

I've been eyeing off an old GC sitting in the corner and wondering if I should resurrect it for analytical work on the reactor. Maybe some gel and a balance is the quick 'n dirty way.

I have recently taken the design beyond proof of concept and ran an absorption tower as planned (water with broken glass) and was happy with the results. No obvious NOx was leaving the tower. There were white wisps, but it seemed to be nictric fumes. The product in the tower water was very weak, but the condensate at the base of the oxidation chamber took paint off my retort stand when a ball joint leaked. It fizzed furiously with bicarb.

I'm making the most of a few days off to work on the latest incarnation.
Improved flow rates for a start; that sand is too fine, there's only so much you can push through, so a coarser more porous catalyst support. Larger oxidation chamber (now with vortex action! oooohh!) and a condenser element in the chamber.

I'd be stoked if I could try a boyhood fantasy or running a N2O4 - Hydrazine liquid fueled rocket engine, just a static test would be awesome. It's a seriously toxic little cocktail if things go to shit; this will need some thinking about. But with NO2 at hand the project is now less of a fantasy.

Fleaker - 22-12-2016 at 10:20

What about just absorbing some chloroplatinic acid onto aerogel or molecular sieve and reducing it with hydrogen?

Chemetix, you ever read that book Ignition! in the SM library?

Chemetix - 22-12-2016 at 12:27


What about just absorbing some chloroplatinic acid onto aerogel or molecular sieve and reducing it with hydrogen

This would work I'm sure; I can at any time use a quartz frit and coat it in platinum lustre paint all of which I have at hand. I have a quantity of palladium chloride which would reduce onto a frit as well. But the mission has been to avoid the PGMs. This is science madness after all.

There is no greater joy than being told you can't have something and you get to say fine! I'll make it from air, water and dirt! Except perhaps, when you find a piece of literature that reports a synthesis using some exotic complex of iridium in an even more exotic solvent. And you replicate the results and improve the yield using hair gel and a dissolved coin.

This is going to be industrial chemistry McGuyver style, sans mullet and rayban aviators. Disappointing to hear the remake is tween Bond. There goes any new inspiration for the next generation of tinkerers.

Sorry aga- that's why there's a trend to talk more than do. You cant tinker with something unless you have a genius degree and work for a government department as far as popular culture is concerned. In fact how do you get a screenager to do anything AFK?

Ps. I'll look for ignition! I hadn't seen or had overlooked that one.

[Edited on 22-12-2016 by Chemetix]

Fleaker - 22-12-2016 at 13:00

PGMs get no love :-/

Sad state that platinum prices are at now.

WGTR - 22-12-2016 at 15:11

Quote: Originally posted by Chemetix  
There is no greater joy than being told you can't have something and you get to say fine! I'll make it from air, water and dirt![Edited on 22-12-2016 by Chemetix]

I think we share a fascination with this. There's something to be said about making a salt (NH4NO3) out of air and electricity (even the water vapor can be condensed from air).

If you include molten carbonate electrolysis, (I've done some of this), then one can isolate solid carbon from the carbon dioxide in air. From there, carbon monoxide can be synthesized, and then with hydrogen, a whole slew of hydrocarbons.

[Edited on 12-22-2016 by WGTR]

Benchtop Ostwald Pilot Plant Success

Chemetix - 24-12-2016 at 02:06

Here is the pilot plant.

It's a photo shoot of the unit in operation.

PilotPlant.JPG - 270kB
In a converted TV stand you can see the urea decomposer in the mantle front left. The black lump in the middle is the air compressor. The reactor tube runs into the converted 2L sep. funnel which admits air via the custom condenser fitting. The condensate runs into a reservoir where excess gasses run to an absorption tower.

PilotReactorSml.JPG - 124kB
This shot shows where the ammonia air mixture is fed into the reaction zone, the nice red glow is transmitted up the quartz.

PilotRedFumes3Sml.JPG - 177kB
Red fumes; condensation forms on the walls and the condenser. That's nitric, baby! You can see the tower behind it and the take off tap to collect the tower absorption.

PilotReservoirSml.JPG - 313kB
The condensate ends up in this collection flask and you can see the splash head type design to let the outlet gasses pass into the tower. There is still a lot of fumes, enough to be visibly red, but once they go to the tower there is no evidence of anything leaving the tower. Kind of important not filling the room with NOx.

PilotFizzSml.JPG - 199kB
Does it fizz with bicarb? You betcha!

This Christmas. Give the gift of nitric!

[Edited on 24-12-2016 by Chemetix]

[Edited on 24-12-2016 by Chemetix]

j_sum1 - 24-12-2016 at 02:40


Realy well done. I'm gonna examine those pix closely when I get on a larger screen.

Merry Christmas.

symboom - 24-12-2016 at 06:56

This needs a sticky for the form this is a breakthrough
Not sure if anyone agrees with this reason nitric acid is a big topic

Also did you incorporate silica beads in the process

[Edited on 24-12-2016 by symboom]

Magpie - 24-12-2016 at 10:11

Very nice and potentially very valuable to those in the world who have been denied this essential reagent.

Tell us about the air compressor please. What is the pressure in the oxidizer? Is that column open to the atmosphere? What is the strength of the acid? Can you titrate it?

[Edited on 24-12-2016 by Magpie]

Chemetix - 24-12-2016 at 15:17

Yesterday was it's first real run, I'll do some do some titrations soonish. But the air compressor might as well be an aquarium pump, it's a 24V dc thing that came out of something, I can't recall what, but it's low power and chugs along on 12V.

There's no real pressure in the system, just the back pressure of the small amount of water being bubbled in the tower which is open to the air. It's actually a simple system. Don't be put off by all the complicated glassware, I did it this way because I can. I'm sure this could be made with found items and a bit of creativity. No silica used, yet! I might find a use for it, but drying the gasses seems more effort than it's worth. Although the ammonia generator spits out a fair amount of water vapour and only dilutes the acid. I could put a condenser coil in the mixing flask to reduce that.

PilotCatalystSml.JPG - 345kB
I modified the catalyst in the pilot plant. It's crushed and screened expanded clay balls used as a growing medium. I sieved out the fine stuff and used another screen to take out the bigger more irregular pieces. A spatula's tip loaded with Co carbonate and a little demineralised water and a good churning to get it coated. That's it! Better flow rates and seems to work at lower temperature, the current draw was only 1.7A at 24VAC in the heater

PilotEndShot.JPG - 203kB
This is the tower, packed with broken glass and a little water. The reservoir with the blue tap holds the concentrated stuff and the exhaust gasses pass into a vertical anti suck back column before running along a small tube to the base of the tower. The red tap can draw off the dilute acid as needed.

One other thing to note is the syringe attached to the reactor inlet. (not in this shot by the way)

Slowly draw back 10ml, disconnect the syringe and put a finger over the end quickly, then put the tip into some water. The ammonia dissolves into the water and the water fills the syringe to replace the volume. You now know what ratio of ammonia to air is being fed to the reactor.

I'll keep you informed of process specs and mods as I make them, the concentration of the acid is going to be the first test.

[Edited on 25-12-2016 by Chemetix]

Marvin - 24-12-2016 at 15:49

Quote: Originally posted by Chemetix  
Slowly draw back 10ml, disconnect the syringe and put a finger over the end quickly, then put the tip into some water. The ammonia dissolves into the water and the water fills the syringe to replace the volume. You now know what ratio of ammonia to air is being fed to the reactor.
This blows me away, just a brilliant bit of lateral thinking.

Magpie - 24-12-2016 at 20:02

This is a really nice project. I have the feeling that you are just getting started on optimizing.

Your ammonia generator is just charged with urea and a little water. Wiki says that a method suitable for lab generation of ammonia is to use urea mixed with Ca(OH)2. This ties up the CO2 as CaCO3. Would this be an advantage to your process?

[Edited on 25-12-2016 by Magpie]

Chemetix - 24-12-2016 at 20:26

Magpie, that's a fair point, one that I considered early on but discarded for a few reasons.

Making calcium carbonate is a right royal pain in the arse, the generator gets clogged with an insoluble mess that interferes with generation and then it means either time or chemicals to clean it. The purist in me decided that no unwanted byproducts would be better. There is the fact that the rate of production would vary and controlling the air ammonia ratio would be harder. The same goes for sodium hydroxide, which I tried; the two sit almost unreacting in the flask and then give it a little heat VOOOM, too much ammonia.

The Water/Urea reaction is a text book first order reaction, very easy to control the rate with temperature and water concentration. It does generate a fair bit of water vapour and I think that minimising that will help the final concentration a good deal.

WGTR - 24-12-2016 at 21:15

If you preheat the gas feed, instead of applying all of the heat to the catalyst area, you can probably operate at a lower temperature overall.

Carbon dioxide is also very soluble in water. Keep that in mind as you sample the gas feed, if this gas is present.

I'm very impressed with your setup; jealous, in fact, but in a good way. Good job!

What do you estimate the gas feed to consist of (even if you think the number isn't very accurate)? It would provide a starting point for future efforts, at least.

Chemetix - 24-12-2016 at 22:01

The syringe technique relies on the ammonia absoption being a good deal quicker than the CO2 absorption. Doing a quick dunk in water fills the syringe in about 30 seconds. I'd say leaving it there for an hour or more would take up most of the CO2. But the quick and dirty result came back with around 6% by volume. I think that around 12% in air is stoichiometrically balanced. It just meant that I turned off the air inlet to the oxidation chamber and let the excess O2 oxidise the NO. It was obvious that the air bleed wasn't needed at the oxidation chamber because the chamber went a much deeper colour without it turned on.

It does highlight that the rate of ammonia production is going to be the step that will allow higher production rates, and drying it will mean more concentrated product.
So far the CO2 doesn't seem to hinder the process aside from being dead volume in the reaction stream.
I think a future project will absorb the ammonia into water to get rid of the CO2 and use the saturated solution to generate ammonia, a passive recirculating system attached to the urea reaction will be the best way to go, and possibly use the waste heat from the reactor to generate ammonia from the saturated solution.

I'm liking the suggestions and I think this is going to get a few brains thinking of better ways. I'm really hoping to see some experimentals come to pass from other members.

It has struck me as I write this that I could do a kit of sorts.
If you PM me I can send a small piece of quartz tube with or without inlets to you for a few dollars US$40-$50. Shipping is variable so I could only guess at a value. I wish it was cheaper but quartz costs a bit and working with it is slow. (if you were a commercial customer the value would be easily 3 times the SM price.) I'll send some pre-prepared catalyst if you like.

Sorry if this setup is too much glass-porn for some of you, @WGTR. I work with it and I still love a sexy bit of glassware.

[Edited on 25-12-2016 by Chemetix]

WGTR - 25-12-2016 at 02:15

OK, so the gas feed appears to be oxygen-rich...that's what I was wondering about.

I like working with glass, too, but i don't have all the fun, expensive, tools. I'm pretty much limited to working with tubing that I get from the science store in town. Soda-lime is pretty easy to join and work with various hand propane torches, and I can fuse boro if I'm careful. But quartz...I just can't! I would love to have a glass-blowing lathe!

The kit idea is a good one. My first thought, however, is to see if the reaction temperature can be lowered enough that borosilicate can be used. That's why I was thinking of preheating, and whether it would give better heat distribution in the catalyst, and hopefully allow a lower maximum temperature. Quartz, however, is the Cadillac of glass. It doesn't get any better. It's just, as you say, expensive and difficult to work with.

I'm wondering if the blue-indicating silica gel might work as a catalyst. It normally contains a cobalt salt in small amounts. Sorry, I keep obsessing about silica gel. For ethanol dehydrogenations, I've prepared the catalyst by co-precipitation of the active copper hydroxide with the inactive magnesium hydroxide support. I think cobalt could work the same way. Starting from a soluble salt makes the starting materials easier to define for a hobbyist, I think. Maybe something like this would work, and be easier to define and repeat. Just an idea. I'm trying to think of ways to minimize variables and avoiding the possible issue of "I just did it the way you did, and I couldn't get it to work". If something like "builder's sand" or clay particles are used, then ideally those materials need to be defined as to purity and provenance, so that others can duplicate it. I can try to prepare this precipitated catalyst if you want, to see if it has the structure that I hope it will. Let me know the inner diameter of the catalyst tube.

Magpie - 25-12-2016 at 08:00

Here's some recollections from my lab experience, FYI:

1. A porous catalyst support can be made from landscaping lava rock. I made some 4-8 mesh as a support for H3PO4 catalyst. It's extremely hard, however, and difficult to reduce to the desired particle size.

Pumice is available as an abrasive at pool supply stores. This might make a good catalyst support.

2. I made anhydrous NH3 by boiling it out of a water solution then passing it through a column loaded with KOH flakes.

Chemetix - 25-12-2016 at 12:36

What I have learnt from this build is the ammonia oxidation reaction is a robust one. It doesn't need a finely tuned set of operating parameters. And there are probably dozens of suitable catalysts out there, I had a few lined up to test and it turned out cobalt was active enough and so I stuck with it. Nickle oxide was my next bet.

The support bed can be just about anything- dirt! The catalyst doesn't seem to care what it's on so long as it holds up at the temperatures and gives you enough support, but volcanic rock or pumice (also a volcanic rock) would be perfect. I was looking to smash up and grind a piece of kiln furniture, and I did try breaking a fire brick but got more dust than screenings.

The tube - how many other options are there?
I'd say the reaction can be lowered to 500C, which is borosilicate range of working temps, just pack more catalyst into a longer tube to allow for the slower rate of reaction. I have eyed off a piece of tubing used for thermocouples, a pyro-ceramic of some sorts. A suitable alternative would be something like a copper tube with some glass tape wound around it( automotive exhaust shop), make a paint with sodium silicate and some silica flour(inhalation hazard) from a ceramics supply. Then some nichrome wire around that and more insulation over the top.

I'd say if you get fairly anhydrous NH3 without the CO2 at an optimum air mix, the reaction might just self sustain the heating. Nickle oxide I strongly suspect to be more active than the cobalt, but that's a hunch at this stage.

[Edited on 25-12-2016 by Chemetix]

Chemetix - 25-12-2016 at 12:51

Quote: Originally posted by WGTR  
Let me know the inner diameter of the catalyst tube.

Sorry I forgot to answer that; 8mm ID.

Chemetix - 25-12-2016 at 17:23

[Edited on 26-12-2016 by Chemetix]

Magpie - 25-12-2016 at 17:49

Quote: Originally posted by Chemetix  

The reactor tube runs into the converted 2L sep. funnel which admits air via the custom condenser fitting.

I'm a little confused here: are you admitting air into the 2L sep funnel absorber as well as ahead of the reactor?

Quote: Originally posted by Chemetix  

This shot shows where the ammonia air mixture is fed into the reaction zone, the nice red glow is transmitted up the quartz.

Clearly you are admitting air here ahead of the reactor. I assume this is the slightly pressurized source from the compressor?

Chemetix - 25-12-2016 at 18:13

Yes I can add air ahead of the reactor, in fact you need to:

4NH3 +5 O2 => 2NO + 3H2O

In the sep funnel the reaction:

2NO+ O2 => 2NO2

means you should need to add air as well. The ammonia air mixture was quite oxygen rich and so the unreacted O2 formed the needed O2 in the sep funnel. This meant I was only diluting the reaction with more air and slowing the next reaction down

2NO2 => N2O4

The rate of dimerization is proportional to concentration.
So I turned off the secondary air inlet and noticed the colour became darker. Edit- sorry, that makes it sound like the N2O4 is the darker product... it's just there was more NO2 by volume and hence darker.

[Edited on 26-12-2016 by Chemetix]

[Edited on 26-12-2016 by Chemetix]

phlogiston - 27-12-2016 at 04:37

Do you have any idea if some of the ammonia is able to pass the reactor unreacted? That would result in acid containing dissolved ammonia nitrate.

New Catalyst

Chemetix - 27-12-2016 at 09:18

Quote: Originally posted by phlogiston  
Do you have any idea if some of the ammonia is able to pass the reactor unreacted? That would result in acid containing dissolved ammonia nitrate.

In the preliminary trials I did with the catalyst on glass fiber, the support melted and there was little surface area to react with and ammonia started coming over past the reaction zone. White fumes appeared as the acid and base began to react.

This happened again today as I had made some modifications to the set up.
I dried the ammonia/air mixture with a reflux condenser this time and noticed there was more glow coming from the reactor and then things started to change. The oxidation chamber lost colour and there was no condensation forming. I lifted the condenser coil out and whiffed the emanating fumes lightly - ammonia. The catalyst had died.

I shut everything down and cleaned out the reactor; the support had a slightly glazed look and a greyish colour. There was oxide residue on the walls and I used some conc. HCL to remove it. What was telling and confusing was the smell of sulfide. Sulfur had killed the catalyst but where did it come from? It gave me the chance to try another variation I had in mind.

Pilot Ni.JPG - 316kB
Broken bath tile support and nickle oxide.

Pilot Ni Reactor sml.JPG - 209kB
Not only did the NiO work, it worked well. The reaction zone glowed much hotter and pulsed hotter with the higher flow rates from the air/ ammonia generator. I'd bet that this would self sustain once it has got to this temperature. Will try next run.

PilotReflux.JPG - 373kB
This is the condenser to dry the ammonia/air stream.
The pulsing glow happens due to the concentrated ammonia solution in the condenser falling back into the flask as drops, the cold concentrated solution emits gas as it hits the hot solution of urea.

The dried air/ammonia "burns" hotter than with the water rich vapour I was using. 1- it meant the high temperatures could have caused contaminants in the expanded clay balls to react with the catalyst or fuse with the catalyst, killing it.
2- it makes more concentrated acid without the introduction of water into the stream.

PilotConcentratedAcidFumesSml.JPG - 474kB
Concentrated nitric fumes like crazy in moist air. The oxidation chamber was filled with acid mist this time. Concentrated acid fumes can be seen leaving the absorption tower. I now understand the need for multiple towers used in industry.

I ran out of time to titrate the product, but it took more Bi-Carb to neutralise a similar quantity of the last batch, and the tower solution gave a more pronounced reaction with bicarb despite far less operating time.

[Edited on 27-12-2016 by Chemetix]

Jstuyfzand - 27-12-2016 at 15:09

Looking great Chemetix, great work!

Have you tried MnO2 as the catalyst?

Fulmen - 27-12-2016 at 15:31

Outstanding work, truly inspiring. Cobalt seems to be the ideal catalyst for this, from what I can tell it's in commercial use today. And a heck of a lot easier to get hold off than platinum/rhodium. If nickel was anywhere near this good, wouldn't we've heard about it by now?

Anyway, the biggest challenge as I see it is the ammonia-generation. I like your approach, but I still can't help thinking there's a better one out there. A kipp-style generator would be perfect, but that's not as easy as it sounds.
I'm not big on glassware, but I wonder if it isn't possible to construct a compact design from metal, at least the ammonia and reaction zones.

j_sum1 - 27-12-2016 at 15:59

My standard ammonia generator is NaOH drain cleaner and ammonium sulfate fertiliser. Neither is too expensive. I don't see why a different ammonia feed would be problematic.

Chemetix - 27-12-2016 at 19:32

Quote: Originally posted by Fulmen  
If nickel was anywhere near this good, wouldn't we've heard about it by now?

.... but I wonder if it isn't possible to construct a compact design from metal, at least the ammonia and reaction zones.

It's funny, I've sort of made a career out of doing things that are assumed to be obvious to everyone else as either 'they've tried and it doesn't work' or 'if it worked like that they'd already be doing it that way'. Nickel seemed an obvious choice really; when you can't have Pt or Pd the next on the list is Ni.
But because of the German patent I tried cobalt first. Actually, I tried red iron oxide first and it didn't seem to do anything. The problem with these alternative catalysts is after shut down there is concentrated nitric acid around to basically chew off your catalyst in the reactor. Platinum puts up more of a fight in this regard.

And on to the other point, the glass porn. The short answer to 'is there a way to do it in metal' and I'd say yes. 316 Stainless would have to be the next best thing, I have this in my workshop and I'm handy with a TIG. But that said I might give a more grass roots approach a go too. Bits of copper tubing- glass bottles with holes ground into them- lots of teflon tape. Maybe I'll leave that to some inventive backyarder to complete, because it's entirely doable. The glassware gives a very educational approach to the process. You really can see what is happening and when and get a feel for how much. And I hope the pics are encouraging and informative for the forum.

I think the ammonia generator could have a few improvements made, I did a literature search for anything that can catalyze the urea decomposition reaction. Nothing useful so far, the research is focused around engine emission control measures. I found using glycerol or sugar with water I can raise the temperature and the rate of evolution...and the energy cost of generating it. I like the idea of just pour in the urea and water and no by products if you keep the ratios right.

I'll draw up a schematic using my M.S.paint- 'Fu'; I should get a better sketch app one day...

ps "Have you tried MnO2 as the catalyst?"

Het spijt me, ik weet niet dat die MnO2 werken. I should think it would work, I'm starting to suspect my suspicions about there being many available catalysts is correct.
Someone can give it a go.

[Edited on 28-12-2016 by Chemetix]

[Edited on 28-12-2016 by Chemetix]

Herr Haber - 28-12-2016 at 04:32

I absolutely love the "let me prove everyone wrong" mindset. Especially in this case !
How many pages in this forum alone saying this or that process for making HNO3 is not doable ?

The only sad thing I see here is the timing. A few more days and you would have definitely gotten my vote for Mad Scientist of the year :)

Fulmen - 28-12-2016 at 04:47

Quote: Originally posted by Chemetix  
The problem with these alternative catalysts is after shut down there is concentrated nitric acid around to basically chew off your catalyst in the reactor.

Good point. Shouldn't be hard to avoid as long as one is aware of the problem though.

As for metals I agree that 316 is the obvious choice, it should work for the absorption tower as well as long as the concentration and temperature isn't too high. Copper sounds like a poor choice, it could perhaps work for the reaction chamber assuming you can produce dry ammonia gas?

Jstuyfzand - 28-12-2016 at 05:58

Mixing in some Dutch, love it Chemetix!
I look forward to seeing more, especially the Titration results.

Fulmen - 28-12-2016 at 08:31

I missed your post where you tested nickel, seems like we have several catalysts at our disposal. This simplifies thing even more as I already have nickel salts. As for the ammonia-generator it's hard to beat urea as a source, although a kipp-style generator would be nice.
This might be useful: (Electrochemically Induced Conversion of Urea to Ammonia)

It might be possible to design a pressure regulated generator this way, using a gravity fed reservoir and back pressure to regulate the electrode area.

[Edited on 28-12-16 by Fulmen]

Attachment: ECS Electrochem. Lett.-2015-Lu-E5-7.pdf (287kB)
This file has been downloaded 594 times

Magpie - 28-12-2016 at 09:45

Quote: Originally posted by Herr Haber  

The only sad thing I see here is the timing. A few more days and you would have definitely gotten my vote for Mad Scientist of the year :)

Yes, this is the most exciting project since Pok's making of Potassium. On an importance scale this has to rank very high.

Urea seems a very good source of ammonia: compact, dry solid, just add water and heat - what could be easier. Regulation would be nice - I guess that's what a Kipp would give you. Does the CO2 cause any problems other than dilution?

Also, urea is dirt cheap. I bought a 50 lb bag for $10.

Jstuyfzand - 28-12-2016 at 09:48

Quote: Originally posted by Magpie  
Quote: Originally posted by Herr Haber  

Also, urea is dirt cheap. I bought a 50 lb bag for $10.

Where did you find such deals?

Magpie - 28-12-2016 at 09:55

"Weed & feed" places. That is, agriculture and garden suppliers.

ecos - 28-12-2016 at 12:41

did you try to use copper wire as catalyst?
I found some videos showing that it works fine
plz check attachment.

Attachment: Media.mpg (3.2MB)
This file has been downloaded 924 times

[Edited on 28-12-2016 by ecos]

WGTR - 28-12-2016 at 13:46

So what's the longest period of time that you've used a particular catalyst? Or how much product can you currently obtain before needing to change out the catalyst?

I'm the type of person that likes to do chemical reactions in stages, so naturally I'd suggest making some dry ammonia gas ahead of time and storing it in a bag, a "gas bag", if you will. I'm not describing the mother-in-law after a chili cook-off, but rather a plastic bag with a weight on top, to regulate the flow of gas. Even if it isn't used for bulk storage, some kind of bag like this can work as a regulator, to absorb pressure fluctuations from your ammonia generator. Air can be supplied the way you already do. Perhaps I could demonstrate it if I have extra time. But then again, if I had spare time, I might spend it taking rides on the pet unicorn that I'll never have either. But I can try.

A cubic foot of gas would be around a mole of ammonia, and that would make quite a bit of nitric acid, if system-wide efficiencies are good.

Chemetix - 28-12-2016 at 13:48

Quote: Originally posted by ecos  
did you try to use copper wire as catalyst?
I found some videos showing that it works fine

You did see my original setup with your thread.

and your reply :


I am really surprised that copper works as a catalyst . I thought we need only Pt to oxidize ammonia.i found a link that shows copper as a catalyst . it also has videos.Link :[Edited on 9-11-2016 by ecos]

And it works too well and over-oxidises the ammonia to N2 and H2O if you recall my original assessment.

CopperNH3oxidation.png - 165kB
copper glows alright- just no discernable production of NO or NO2.

Jstuyfzand - 28-12-2016 at 15:40

The copper is too active, does this mean that the Ammonia gets converted to N2 and H2O immediately or does it go to NO first and then N2 and H2O?
If its the second, higher flow rates could take care of this problem, I presume.

Fulmen - 28-12-2016 at 15:40

Electrochemical urea to ammonia (eU2A) sounds like a promising alternative. The paper used nickel electrodes which had a catalytic function, luckily nickel strip for battery assembly is easy to get hold of. The method used constant voltage; 1.65V applied to 30% urea + KOH.

Chemetix - 29-12-2016 at 00:43

As to the question of whether the ammonia goes to NO then N2, it is largely academic. I can't say by observation alone. I'll leave that to analytical chemists to solve one day.
But there just hasn't been enough time in the day for me to give the unit a good run down the highway so to speak, open her up and see what she can do.
Holidays have consumed me with family duties and finishing off commercial work as well.
Why won't someone pay me to be a hobby chemist? Dammit!

So as yet I haven't been able to get an estimation of catalyst life expectancy. All I can say is the cobalt carbonate survived several shutdown and restarts until a change in the system design ultimately caused it to die. Maybe using those clay balls was fine so long as there was moisture in the stream to prevent the catalyst being poisoned somehow.
Maybe the ceramic tile fragments are cleaner and will run at higher temperatures and drier conditions and would have allowed the cobalt to survive longer.
But I'm sure you can all appreciate that the goal is to find a set of parameters that allow the production of the highest volume with the highest concentrations in the shortest amount of time. I'll get a few days soon to let it have a good run and get an estimation of efficiency and final concentrations.

eU2A does sound like a faster way to get ammonia, but then, so is a bigger pot. The latter is the simpler approach, I'll leave electrochemistry for those braver than me.
The gas bag idea sounds a great way to do analytical studies on the reaction. I have my chemical engineering hat on at the moment, it's all about increase of production for minimum costs.

[Edited on 29-12-2016 by Chemetix]

Fulmen - 29-12-2016 at 02:33

It's not the speed of eU2A that interests me but the possibility of a constant pressure generator.

Magpie - 29-12-2016 at 10:27

It seems that a key goal here is achieving the ability to send the optimum air/ammonia mix to the catalytic reactor, smoothly and continuously.

With the air compressor you have control of the air feed rate. Now if you could match that with similar control over the ammonia feed you could achieve that goal. The challenge is in how to do this.

You don't have a batch process here but a continuous process. This is much trickier to design and control. Industry would use flowmeters, pressure controllers, and feedback loops. But for the home chemist the goal would be to do this in a much simpler and cheaper way.

Chemetix, you know all this. I'm just putting my thoughts on paper. You are doing a great job and I completely understand how time consuming all this experimentation is. Like the rest of us you have other demands on your time. At least you live down under and don't have to face an ice-cold lab like I do. ;)

WGTR - 29-12-2016 at 11:34

Quote: Originally posted by Magpie  
You are doing a great job and I completely understand how time consuming all this experimentation is. Like the rest of us you have other demands on your time.

Seconded. Your efforts are really appreciated!

Sometimes people ask me if I'm upset with them, because I seem argumentative. In actuality, they just did something that impressed me, so I'm trying to draw out every detail so that it can be duplicated by others later on (hopefully without sounding too demanding). The challenge is to take the piles of raw data (I added solution A to B, turned around and walked 3 feet in 0.25 seconds, waited 20 more seconds, sneezed, and then added some of B to C), and isolate all of the important variables out of all that. Sometimes what's important isn't what you'd think.

I recently developed a procedure at work that did exactly what I wanted it to, every time. Then someone else tried to duplicate it, and got the exact opposite results, every time. After some investigation, it was found that there was one step where I was rinsing a part with a spritz of DI water, whereas the other person was rinsing it under running DI water for several minutes. It made all the difference in the results. So, we had to go back and add a few ppm of "contamination" to get the desired results.

Another time, we developed a catalyst for a particular procedure. All of the samples worked more or less OK, but there was one sample that was dropped onto the ground. It was carefully scooped up off the tile and put into a bag. Just for fun, this one was tested also. It turns out that this one worked several times better than any of the others. Though a lot of effort was put into characterizing it, no one could find anything different about that sample, other than its activity. So all of the experimental details were documented (drop sample onto floor, dance in a circle, scoop up this way with this spatula, etc.). Maybe some day someone else will be able to take this information and make sense of the results.

A general rule around my lab is, the most useful bit of documentation is the part that didn't get recorded. Additionally, experimental details that I could expertly recall during an experiment, are a week later just a vague memory. Finally, too much caffeine brings on hallucinations of pet unicorns and talking cats.

Chemetix - 29-12-2016 at 14:21

The attention to details is what makes chemistry an art. Physicists take their superiority from working with first principles and mathematical logic. Chemistry gets its results from the messy chaotic world of reality. When you are trying to replicate a published work and you stand there after the third hour of slow addition with stirring, and you contemplate that the author reports, ' ...after half an hour the mixture becomes milky and changes viscosity...' while you are looking at a clear reaction.

Are you able to know that what you are looking at has all the signs of the reaction still working fine despite the obvious difference from what was reported. Are you able to draw on your dark arts of the work up and get the product and yield despite doing it differently from the published work. Are you able to intuit that because you used a technical grade reagent the impurity prevents the phases from separating like it was published.

And this is why I love running my own research projects, no lab manager or supervisor insisting I do it a certain way when I know I can cut out three steps. B.P-be damned!
And I love SM, being able to publish my work in almost real time. To peers and what seems like lab partners you can lean across the bench and say ' Hey, check this out' to.
Without the lab politics, time pressures, constantly justifying everything you do, staying on an agreed path of work when you know a detour might solve an impasse.
There are some real benefits to being a backyarder. And some real detriments. Being poor and not having the LC-MS you fantasize about putting in the corner where you just know it would fit. or a lab that is either freezing or sweltering. (I've had a few days recently where it just got too hot to work)

I'll continue to fill in as many details as possible. But I should have been clear about my mission statement much sooner. The project was focused on ' Make nitric acid from urea in a way that can be easily replicated '.

So in many ways I'm being deliberately rough, no measured amounts no specific grades of materials. And can I, under these conditions, have a robust plant that can still produce useful amounts to those who need only 100ml.

I might offend the research chemists, but I will have more appeal to those who are looking for a way of reliably obtaining their much coveted reagent. If I can achieve that, then I'll think about doing some kinetic studies on the alternative catalysts and publish their relative effectiveness. Or, someone else can, I'm not protective about this work.
In fact working collectively and openly justifies our existence as amateurs. We can generate good chemistry from a community model not a capitalist one.

Magpie - 29-12-2016 at 14:52

FYI, here's a schematic of a fluid control system I have found useful. It shows a way to control the feed rate using a valve located on a bleed line. This example is for control of an air feed. But I use it most of the time to control the amount of water sent to my condenser.

compressor feed control.bmp - 703kB

[Edited on 29-12-2016 by Magpie]

phlogiston - 29-12-2016 at 18:02

Very, very interesting, thanks! For some, this extremely useful reagent is very difficult to obtain. Seems worth a sticky to me.

So in many ways I'm being deliberately rough, no measured amounts no specific grades of materials. And can I, under these conditions, have a robust plant that can still produce useful amounts to those who need only 100ml.

There is definitely something to say for this philosophy also in a professional setting. I have a colleague that changes seemingly unimportant variables in experiments all the time when 'duplicating' it a second or third time. He deliberately takes the other bottle of a reagent if we have several of them in the storage room, uses a different pipette, takes a coffee break in the middle of the experiment, etc. His reasoning is that if an experiment is robust to these kinds of changes, it is more likely that it can be duplicated by someone else in a different lab. Most scientists would try to replicate the conditions of their initial experiment as closely as possible and only if time permits proceed to try different variations. Time being a very valuable resource however, this step is skipped more often than not.

I tried this reaction also using copper wire, but the heat of the reaction is sufficient to melt the copper wire very quickly, even without applying external heating. Do you adjust the heating to compensate for the heat generated by the exothermic reaction itself?

PHILOU Zrealone - 30-12-2016 at 05:41

Is there a risk of explosion inside the glassware?

NH3 + O2 + catalyst may be used as rocket propellant.

[Edited on 30-12-2016 by PHILOU Zrealone]

Fulmen - 30-12-2016 at 07:00

Quote: Originally posted by Chemetix  

The project was focused on ' Make nitric acid from urea in a way that can be easily replicated '

I like your philosophy. For amateur chemistry supplies and equipment are often a limiting factor, one mans "simple build" can be an acquisition nightmare for others.

This process seems quite robust, and the different parts can easily be modified to suit individual preferences and capabilities.

Schematic diagram

Chemetix - 30-12-2016 at 14:59

Quote: Originally posted by PHILOU Zrealone  
Is there a risk of explosion inside the glassware?

NH3 + O2 + catalyst may be used as rocket propellant.

I guess if there was pure ammonia and pure oxygen it's a potential risk. I'm fairly sure that ammonia in air and diluted with CO2 is highly unlikely to ever ignite back upstream.

I think I said earlier that I'd draw up a schematic.

OstwaldStyle NA Process.png - 34kB

[Edited on 30-12-2016 by Chemetix]

Magpie - 30-12-2016 at 16:00

Nice! Thanks for this. I'm loving the simplicity.

Wiki says that the explosive limits in air are 15-28% NH3. The somewhat heated (I presume) air/NH3 mix coming out of the NH3 generator is diluted with CO2 and water vapor. Are there any scenarios where explosion is a risk that you are aware of at this time?

Chemetix - 30-12-2016 at 17:52

You know, I tried to see how you could blow up the system and I don't think it's easy. If you managed to fill the biggest volume with air ammonia, and somehow managed to ignite it, it burns so slowly without a catalyst, so the worst that could happen is you pop the top on the sep funnel. I saw a wisp of blueish flame try to creep back from the copper oxidised reaction once, but it stopped suddenly. I read somewhere that ammonia is difficult to sustain combustion even under perfect conditions.

Wiki confirms this:
"The combustion of ammonia in air is very difficult in the absence of a catalyst (such as platinum gauze or warm chromium(III) oxide), because the temperature of the flame is usually lower than the ignition temperature of the ammonia–air mixture. The flammable range of ammonia in air is 16–25%.[22]"

As far as a typical fuel air mixture goes, this one seems fairly tame.

And by the way- Ostwald got his process, Haber got his process, then there's Solvay, Weldon, Burton, Raschig, name a few. Does this constitute a unique process by industrial standards? Do I get to name it?

[Edited on 31-12-2016 by Chemetix]

plante1999 - 30-12-2016 at 18:47

In the past I have done my experimentation on the ostwald process, and got success. I was, however, using platinum. As a substrate I would recommend either asbestos or the sintered oxide itself.

To generate a continuous flow of ammonia, the best would be to prepare it and store it under pressure with a flow regulating valve, however, since this is inconvenient (in my personal experience at least) Direct production is more desirable.

One could make a setup in which urea is decomposed, the gas passed through triethanolamine and the resulting gas should be quite pure ammonia. A flow meter could then be installed to monitor the output.

A side note about the air/ammonia ratio: In my experimentation and research I found out that if too much ammonia is present one would only get nitrogen and water because ammonium nitrite would be formed _in situe_. On the other side, if too much oxygen is present, the nitrogen oxides tend to decompose back into oxygen and nitrogen. One should try to maintain a slight oxygen excess for better yield of the nitrogen oxides and put another air inlet after the catalyst tube.

For optimal absorption in an amateur setting, 3 column in a row can be used. A strong cooling of the exit gases and the tower is recommended for superior efficiency.

Chemetix - 30-12-2016 at 20:06

" One should try to maintain a slight oxygen excess for better yield of the nitrogen oxides and put another air inlet after the catalyst tube.
For optimal absorption in an amateur setting, 3 column in a row can be used. A strong cooling of the exit gases and the tower is recommended for superior efficiency. "

That's good advice.
Would you recommend having the highest concentration of catalyst per surface area and generating the highest reaction temperatures you can, or as I have been doing, diluting the catalyst over a larger area and running at a lower temperature than would be seen in conventional Ostwald plants?

I will aim to have some form of monitoring system of the gas composition before and after the reaction zone. I hope it's not necessary for a good yield. That there will be a fairly wide sweet spot for optimum production. One that is easy to stay within for an amateur and still get, say, 80% yield based on the urea. I suspect it's possible but like most things, they turn out differently in practise.

[Edited on 31-12-2016 by Chemetix]

plante1999 - 30-12-2016 at 20:59

Well, try to limit the flow to keep the temperature on point with a concentrated catalyst.

As for the yield, I'm afraid it won't exactly be what you expect. Thermodynamics alone makes it hard to reach your expected yield. By the design of the process alone you will lose a good deal of ammonia as hydrogen/nitrogen/water. In the disproportion of the nitrogen oxides you will also loose a fair bit.

Fulmen - 31-12-2016 at 02:05

While it's good to cover all the points I don't think a gas explosion is very likely. Ammonia isn't very flammable and small diameter tubes reduce the risk further. A flame arrester made from metal mesh could be installed if you're worried, but I don't think it's necessary with this ammonia/CO2-mixture.

j_sum1 - 31-12-2016 at 20:34

I am assuming that you have already come across the (rather lengthy) thread -- the Drunken aga challenge for producing nitric acid. If not, here it is:

One of the most promising ideas to come from that was a gem of a paper that deltaH found. His link seems to be broken but I found an alternative:
[edit: only part of the paper available. Jump forward two posts for actual file. /edit]

As you can tell from the title, manganese dioxide has yielded results in this process before. I just thought I would throw this idea in the hat for when we get to playing with different catalysts. MnO2 has to be one of the most accessible chemicals around and if it can be made to work reliably I think it would meet your overall aim.

I think that your apparatus is vastly simpler than the setup that Snowden Piggot came up with.

apparatus.jpg - 69kB

[Edited on 1-1-2017 by j_sum1]

Fulmen - 31-12-2016 at 21:27

The apparatus will be as complex as funds and supply allows ;)
(Nerd rules applies).

[Edited on 1-1-17 by Fulmen]

Chemetix - 1-1-2017 at 03:10

" MnO2 has to be one of the most accessible chemicals around and if it can be made to work reliably I think it would meet your overall aim."

I suspected it might work, and I might just have to open an old dry cell just to see.
Jsum, that apparatus from Piggot seems like a liquid reaction scheme, i'd be very interested in seeing the details. The link to the book gave an excerpt that said something about preventing inhalation from carbon monoxide...I'm confused.

Today I had the inevitable run that seemed to not work as well as the last few. I changed the level of doping on the substrate this time to make a higher concentration and it didn't seem to react as hot or as fast as the last time I tried nickel oxide.
Will try again soon.

j_sum1 - 1-1-2017 at 06:33

Ok... here is the actual paper.

Attachment: manganeseincatal00piggrich.pdf (1.3MB)
This file has been downloaded 779 times

plante1999 - 1-1-2017 at 09:43

Try to use nickel nitrate impregnated catalyst and reduce it in the tube with the usual procedure, the catalyst is going to be much more active this way. Also, ammonia/air explosion can happen, however it's extremely unlikely, plus, it doesn't combust very fast. If you are worried, put a pressure check blow valve if you increase the pressure. It could literally save your life.

[Edited on 2-1-2017 by plante1999]

Chemetix - 1-1-2017 at 14:12

Thanks Jsum. The Piggot paper is a very similar scheme to what I have just built. I'm glad to have this, it pretty much outlines the problems faced with the oxide catalysts. Variable activity and quite sensitive to how they are run. After reading this it struck me that changing how you start the reactor could affect the effectiveness of the catalyst. And a detail that might explain why this run was quite different from last time. I waited for the ammonia generator to come up to temperature and production before introducing it with air to the fresh catalyst. The last few runs the catalyst was heated, then the air was started and at the same time the ammonia generator switched on. The rate of ammonia slowly increasing up to about 6% the total volume. The slow introduction of ammonia could be critical to how the oxide performs as a catalyst.
@Plante " use nickel nitrate impregnated catalyst and reduce it in the tube "

I'm curious, if platinum and copper work as a catalyst while metallic, do the oxides work as true oxides or do they get reduced to a metallic state?

2NH3 + 3MO => 3M + N2 + 3H2O

I might run some hydrogen through the tube at heat to reduce the oxide, then see how it behaves.
First I'll replicate the slow startup procedure and see if this is a variable.

See below; I like how I'm able to edit the post in the past.

[Edited on 1-1-2017 by Chemetix]

Fulmen - 1-1-2017 at 14:57

Fascinating read. It's becoming apparent that many metal oxides can work as a catalyst, but most plants still use platinum. So it's safe to assume you are working with a disadvantage. If readily available oxides were great we'd seen more commercial use by now. But for amateur use it still seems promising, as long as the catalyst isn't expensive or hard to make a bit of wear might not be a major issue. Nobody's going to run such plants for days anyway.

plante1999 - 1-1-2017 at 20:50

Well, in the case of copper, it's the oxide/metal doing the work. Just like with silver, or many other metal catalysts.

My knowledge of what goes on exactly with platinum is too limited to tell you what happens then.

However most catalyst of this nature work by a constant reduction/oxidation. You can see this in action by putting copper wire over an acetone beaker. If you heat the wire and lower it you will notice it starts glowing from the heat and move from "black oxide" to "copper" back and forth.

Calcination is better done under an oxidative atmosphere for these catalyst. If you put ammonia in you could get nitrides, although this is far fetched. Mainly the porosity will be lowered.

The thing that you want is active oxides. In the case of copper, it's easy to get activity is not very surprising considering the purity in which you can get copper, and the metal properties.

Chemetix - 2-1-2017 at 00:07

The mechanism of catalysis is of course a complicated affair, platinum seems to work by pulling off hydrogen from the moiety where copper seems to offer an active oxygen. It's the idea that the metal does the transfer of electrons while the oxide becomes electron deficient and activated that is the way I see this working. Whether this is correct is speculation on my behalf at this stage. But I see this is where the activity of the oxide catalyst comes from, small sections of the oxide have exposed metallic sites to transfer electrons.

I tried the same start procedure as earlier than yesterdays run and it seems like there are subtle differences. The catalyst seems to build to a maximum efficiency where as the quick start seems to have a lower stable level of efficiency. I really need to have some analytical hardware to make these calls, but for now perception is all I can go on.
But todays' run with a new catalyst made with 10g light tan house brick support and 1.0g NiO filling the tube to 5cm, ran pretty much the same as yesterdays run. This time it ran continuous for 5 hrs and generated 13ml of condensate and 15ml of tower absorption. I realised I have no indicator on hand, otherwise they would have been titrated. An estimate based on the way they took a certain amount of bicarb to neutralise would be, condensate 50% conc. - tower solutions 30-40%.

The urea is so slow to decompose, I think doing your gas in a bag trick WGTR is the way to get the controlled flow and more of it. I am torn with the electro-accelerated decomposition scheme, sounds effective, but do I want more complicated equipment to build...not at the moment. Maybe I'll just get a 10 litre pot and fill most of it with urea and water and bang on a reflux system, stick the whole thing on a gas stove.

I wonder if copper could be 'poisoned' to prevent it over oxidising the ammonia back to N2.? Lindlars version of copper.

Fulmen - 2-1-2017 at 03:27

Quote: Originally posted by Chemetix  
but do I want more complicated equipment to build...not at the moment.

That's understandable, you have more than enough to deal with at the moment. Still, for a plant capable of producing practical amounts of acid a better method for ammonia is needed. Perhaps someone else could start experimenting with the eU2A, by the time they have a working method I'm sure you've made headway with the catalyst.

j_sum1 - 2-1-2017 at 05:05

Quote: Originally posted by Fulmen  

That's understandable, you have more than enough to deal with at the moment. Still, for a plant capable of producing practical amounts of acid a better method for ammonia is needed. Perhaps someone else could start experimenting with the eU2A, by the time they have a working method I'm sure you've made headway with the catalyst.

What is wrong with the standard ammonia salt and sodium hydroxide?

Fulmen - 2-1-2017 at 07:42

Nothing, but it will consume more chemicals. Urea is both cheap and readily available, and the CO2 doesn't seem to cause any problems.

WGTR - 2-1-2017 at 12:08

I alluded earlier to how I like doing experiments in stages. There's a practical reason for that. I think this project has a good start and a lot of potential, but it's possible that too many variables are being tackled at once.

Many people here have made ammonia at one point or another. It's a pretty easy thing to do. That's not the important part of your process. What's important, is your use of alternative, cheaper, catalysts for the oxidation of ammonia to nitric oxide. The ammonia can come from many different sources, even a pressurized cylinder.

If you can store enough of it ahead of time to allow your system to reach steady-state operational conditions, you can make measurements of the ammonia/air mix and flow rate, residence time on the catalyst, and make a determination of the product (so many grams per hour, etc.). If the reactor is always ramping up or down, it's difficult to make these measurements.

The "gas-bag" referred to is just a Zip-Lock bag. The plastic on these is somewhat tough, much more so than a garbage bag or something, but it still has some elasticity. I connect to one by poking a small hole in the inflated bag, and inserting a slightly larger diameter glass tube, maybe a couple of inches in length. This glass tube is connected to some flexible tubing, and the glass tube is taped down to the bag, like a nurse tapes an IV to your arm, to keep it from moving around at the bag seal. Sometimes I'll have trouble with the seal, but usually the seal is a surprisingly good one. A few days ago I put one of these bags together, and floated it on water with a weight on top. It didn't appear to lose any volume overnight. Of course, something like this can't provide more than a fraction of a psi, but it's an easy way to store a gas and use it at some adjustable rate.

Fulmen - 2-1-2017 at 12:36

Quote: Originally posted by WGTR  
Many people here have made ammonia at one point or another. It's a pretty easy thing to do. That's not the important part of your process.

For this prototype I would have to agree, but if one wants to scale things up to a decent production volume this becomes a big issue.

Bringing it up at this time could be a distraction, but at the same time the work shows such promise it's impossible for people to stop thinking of improvements.

Magpie - 2-1-2017 at 12:57

I don't know about Australia but buying or filling a cylinder of NH3 in the US might well bring the DEA a knockin'.

Chemetix - 3-1-2017 at 02:40

Ok, here's the details of the latest run.

Catalyst: 1.0gNiO on 10.8g house brick support, screened to pass 7.0mm square mesh and retained on 2.0mm square mesh. Used enough to make a 5cm reaction zone.
Run time: 5hrs.
Condensate collected: 13ml
Tower collection: 15ml

titrated against 0.5M NaOH
1.00ml of sample in 10.0ml of demineralised water, 0.5ml of red cabbage indicator. (what else!)

Condensate: needed 12.6ml 0.5M NaOH to neutralise. Concentration of sample = 6.3M
Tower: needed 3.23ml 0.5M NaOH to neutralise. Concentration of sample = 1.6M
reference sample of commercial 70% HNO3: needed 33.5ml 0.5M NaOH to neutralise. Concentration of sample = 16.7M
NiRun4TitrationStart.JPG - 288kB Start of titration

NiRun4TitrationEnd.JPG - 222kB End point

As expected, but not what I hoped; its pretty mild stuff coming from the process. But not too far from industrial results- I think older style plants would get 30-40% from the oxidising chamber.

The visual observations of the oxidising chamber have been noteworthy. Each reaction has so far behaved differently, subtle differences and not so subtle.

NiRun4RxnZn.JPG - 718kB
Note the reaction zone, no pronounced glow like the first run using nickel

NiRun4Obs.JPG - 356kB
And not the dense fumes. Condensation appears more prominent but what does it mean? I will have to do some yield calculations next to see the sort of efficiencies the catalysts are displaying.

The results are not terrible. This was going to be difficult to get a result and I have come further than I expected with this project. I'm now intrigued by the possibility of making different catalysts, with promoter and activator additives. Also I want to try palladium just to see the difference between transition metals and a PGM. But increasing the ammonia to get a decent flow rate, and one that can be measured and controlled, will solve the issue of catalyst efficiency and conversion values.

Fulmen - 3-1-2017 at 03:02

Do you have any idea/guess of how much ammonia was produced/consumed? From what I can tell you've collected appr 0.1moles of nitric acid.

Chemetix - 3-1-2017 at 03:16

All I can say about the ammonia production is that it is slow. Just by looking at the fumes that enter the chamber it's a wisp of a flow rate. So really, the conversion might be moderate. More work to do.

Fulmen - 3-1-2017 at 03:28

Luckily there are many ways to measure gas production, I would probably focus on that next. Sooner or later you will need some efficiency data.

WGTR - 3-1-2017 at 06:54

I'm not sure when I can do this, or if I'll have the time at all, but I'd like to help design the metering on the input side of the reaction.

The idea is to use gas bags containing ammonia or air, and seal these bags inside of larger chambers/tubes. A bag would be connected to the outside world through a fitting in the outer chamber. Gas flow could be controlled by a second fitting in the chamber, that would apply slight air pressure to the chamber. This would force ammonia or air out of the bag, through the output port. A MEMS flow meter on the input port can measure gas flow, and this can be integrated with time to give gas volume. As long as the pressures are low in the system (fractions of a psi), I don't think much error would be added to the results. Except for the flow meters, everything else is a DIY construction project, and fairly cheap. Since the flow meter is separated from the gas mixture by a diaphragm, it doesn't have to tolerate anything other than air.

I thought of pre-mixing ammonia and air into one bag, as this would require only one flow meter. However, I'm not convinced this is a good idea (BOOM!). Figuring out the flow rates can give an idea of the residence times of the gas mixture on the catalyst as well as the optimum gas ratio, and then it will be possible to tailor these parameter for best results.

I appreciate the pictures and your efforts at describing the results, by the way. This is turning out to be a very interesting and informative project.

Fulmen - 3-1-2017 at 07:09

I like the bag-in-bag design, starting from a known volume you really don't need a flow meter. As a first approximation you could simply fill bags of known volume using the same setup&power as your experiment.

Magpie - 3-1-2017 at 12:22

Here's a flowmeter that I occaisionally use for metering argon. You can sometimes pick these rotameters up fairly cheaply on eBay. This one reads in SCFH (standard cubic feet per hour).

rotameter.jpg - 107kB

WGTR - 3-1-2017 at 21:41

Here's a brief video (greatly sped up, of course) illustrating the rough concept of a bag-in-a-bottle. The Zip Lock bag acts like a flexible diaphragm, and isolates the gas from the air supply that's driving it. This allows changing the gas flow rate on the fly. There is less likelihood, also, of an ammonia release into the room if the bag breaks.

Attachment: gas_bag_in_a_flask.mp4 (4.1MB)
This file has been downloaded 1045 times

Chemetix - 4-1-2017 at 01:44

The bag in a flask is an awesome way to do it really, scaleable with ease. The urea water thermal decomposition is slow; my god it's slow. But as a feed stock you have to concede it's cheap and easily handled. With some collective problem solving the ammonia generator can be made cheaply and effectively I'm sure.

So how much ammonia was the generator producing per hour? Not a lot. Today I got the generator running, then bubbled the outlet into a volumetric flask with about 100ml of DM water. After an hour of running the ammonia was removed and switched off. The 200ml volumetric flask was made up to the 200ml mark and titrated against a known concentration of HCL.

The actual procedure was done by taking a 10.0ml aliquot of ammonia from the 200ml standard solution flask and titrating with 0.455M HCL.
0.455M just happened to be the concentration of a 50.0ml pipette of hardware store HCL added to a 1000ml volumetric flask and titrated with the same 0.5M NaOH solution I used yesterday to find it's actual concentration.

10.0ml of NH3 solution was added to a flask with the same indicator as yesterday (red cabbage) and then titrated with 0.455HCl. until a lightly acidic point was reached.
This was a source of error with the titration as NH4Cl is slightly acidic at low concentrations so the end point needs a judgement as to the point of neutralisation. I took it to a mauve rather than amethyst which is the pH7.0 point. A low concentration of NH4Cl hs a pH of around 5.

Note: have edited out the calculations until I get better data, I wasn't happy reporting speculation.

[Edited on 4-1-2017 by Chemetix]

[Edited on 4-1-2017 by Chemetix]

Lefaucheux10 - 5-1-2017 at 00:57

Hi guy,

Very interesting post and i will try to built a similar setup

Here a proposition, why not use ammonium nitrate wich is like urea a pretty cheap fertilizer.

As far as i know it causing troubles to distill off nitric acid from sulfuric acid and ammonium nitrate but ....

for running the Ostwald process you can do

NH4NO3 + NaOH = NH3 + NaNO3 + H2O

and the ammount of NH3 passing threw the catalyst can be controled by the rate of addition of NaOH soln into saturated boiling NH4NO3 soln

after you can boiled the resulting solution to obtain NaNO3 wich can be mix with H2SO4 to make more HNO3

with this you only lost the NaOH from the fist step, some enery to boiled of the soln of NaNO3 and some sulfuric acid

very cheap chemicals ! doubled amounts of HNO3

Another think : Why not absorbate the gaz into some H2O2 for accelerate the oxydation of NOX gases ?

Wait for your comments ;)

Chemetix - 5-1-2017 at 04:28

There is a problem with getting ammonium nitrate these days. In fact just about any nitrate is hard to come by without paperwork and expense. Many people on this site have thought about and tried different ways of making nitric acid without a nitrate salt probably for that exact reason. Ammonium sulfate is readily available to generate ammonia by use of hydroxide. Even urea and sodium hydroxide makes ammonia quite rapidly and effectively. But they are extra expenses and extra steps.

An ostwald style plant is a continuous process, a switch it on let it run kind of thing, so urea and water can be left to boil and decompose without having to check on it. A reaction with an ammonium salt has to be run and monitored as a batch process, and once generated you need to then find a way to add it in a controlled way to the reaction zone. Despite WGTRs ingenious bag in a flask technique, I'd rather avoid it if I can.

And you are right about using H2O2, it does maximise the absorption of NOx. I think that step will be one for getting the last bits of efficiency from the process, I'm concerned about getting the reaction running reliably and in moderate yield for a start. I have no idea how effective these alternative catalysts are in the broad sense. I've got a lot of research ahead. But I'm close to being able to state yields and operating conditions for a given catalyst.
I'll keep posting my progress for those interested in this process.

NH3 Flowmeter Sml.JPG - 330kBThis is a very sensitive flowmeter I put together from fish tank needle valves and some small irrigation fittings. The fluid is ammonium sulfate soln. and some food dye.

[Edited on 5-1-2017 by Chemetix]

WGTR - 5-1-2017 at 08:11

Good job on the flowmeter. As long as it doesn't change, relative measurements can be recorded from it. The calibration can be done later on.

ecos - 8-1-2017 at 01:02

I think this reference shows some results of using Cu/CuO as a catalyst :

I point to table 1( row 3 col. 5)

[Edited on 8-1-2017 by ecos]

Chemetix - 8-1-2017 at 02:43

An update on progress

I've learned a thing or two so to keep the flow of information going I'll give an update.
Run#5 Ni2O3 black nickel oxide
I have a flow meter decided to try the highest flow rates that will sustain the reaction.
I replaced the ammonia generator flask from a 250 ml to a 1L 1/2 filled, I cranked up the heat to the ammonia generator and had it refluxing quite adequately, this should be a maximum of NH3 supplied to the reactor. I ramped up the air flow and tried to see the most flow I could supply. A red colour and heavy fog persisted. As I brought up the air flow, there was a point where the red colour disappeared. After bringing the ammonia temperature back down and also the air supply, the reaction failed to start again. The catalyst had changed.

Ni2O3 after run4&5.JPG - 241kB
This looks to be NiCO3 and perhaps some NiOH. It fizzed with dilute acid.
Nickel is susceptible to being chemically changed

Run#6 tried a tile glaze that did nothing

Run#7 CoO black cobalt oxide
Started the reactor heater, turned on the NH3 generator turned on the air with manometer reading 40mm. A dark red gas began to appear at the outlet then as the ammonia generator began to reflux a fog appeared. Ran the system with air at 50mm for a few hours then shut down.

If the reaction was producing dark red gas without fog before reflux then the urea decomposition reaction might be generating too much ammonium carbamate when refluxing. This might be the source of the fog. Ammonium ions get through the catalyst zone unchanged?
I took some of the condensate and evaporated it in a beaker.

NH4 residueRun5.JPG - 436kB
Those long clear needles look familiar. It looks like ammonia is getting past in significant quantities when the urea water reaction is refluxing.

Run8CoO.JPG - 202kB
Run#8 CoO same as run 7
Cleared the condensate line and flask, refreshed the tower water.
Kept the air flow on 40mm to begin with and maintained the urea reaction below visible gas evolution, no bubbles. Dark red gas without fog in the oxidation chamber, the condensate flask and the vertical displacement tube, no visible fumes or red gas leaving the tower.
Left the unit run for 3 hrs.
Switched off the reactor but bubbled the ammonia generator into a 200ml volumetric flask with cold water for 1hr with the same air flow setting (50mm).

Noticed bubbling in the condensate flask.
This would have to be the reaction :
2NO2 + H20 -> HNO3 + NO

Attachment: NObubbles Run8 CoO.mp4 (1.4MB)
This file has been downloaded 771 times

Left the reaction to continue overnight.

Next day.
Collected 7ml of condensate and 18ml of tower absorption.

Titrated the samples
condensate 8.1M = 0.056n
tower 7.2M = 0.122n
Ammonia generator collected 0.4M x 0.200L =
3 hrs @ = .24n (NH3)

Theoretical amount of HNO3 from NH3 is n(NH3).024 x 11/6 = 0.439n

Yield (0.056+0.122)/0.439 = 0.4
Approximately 40% from the theoretical amount.

My 80% estimate was ambitious. But urea is cheap, and the rest can be scaled.
I need to reproduce these results a few times before I'll call it properly.

Flow meter Calibration.bmp - 669kB

Chemetix - 8-1-2017 at 03:05

Excellent find Ecos.
It does explain why copper is so erratic as a catalyst at atmospheric pressures, I could find no activity for NO conversion with the so called "copper" scourer I had at the time.
For a start I have no way of telling if the "copper" was pure.
Higher pressure suppresses the formation of the metal nitride which stops the catalyst working. Other metals which have active oxides but resist nitride formation would be good candidates to screen for this process.

Magpie - 8-1-2017 at 09:34

I like your flowmeter. I made a manometer just like it for delta P measurements when I was trying to see if I could use an old HVAC fan from my house for a fume hood. I also used it when trying to develop a venturi based air-mover. Neither of those experiments produced a suitable air-mover, however. The delta P vs airflow characteristics were not appropriate.

I'm surprised that your data plot is so linear.

Chemetix - 8-1-2017 at 12:25

I'm kind of surprised too. For a start there is no zero, the fish tank valve had no off point, just a slow leak. And I didn't run it a full open either, that would push the fluid over. The secondary valve controls the scale as I'm sure you know. I set it at something that gave me a small readable flow and a decent flow at around full scale which was around 90mm height. Values where taken by the time taken to fill a 1L volumetric flask.

The plot could be logarithmic, however, excel fits a fairly convincing linear regression.

If the value for 5mm height turns out to be somewhere around 0.5 - 1.0ml.s-1 I think I'm looking at a logarithmic plot. Which I suspect would be the case given how these things work.

violet sin - 8-1-2017 at 13:41

I was wondering if some sillica gell dessicant could be used to capture low concentration NO2 effectively, for later release by heat and production of concentrated acid? make for effective use of H2O2 in a separate container, instead of dumping it in the whole system and hoping for use vs. decomp. It was mentioned in another thread previously. At the verry least it could be used as a pre-exhaust scrubber, but I don't recall if there was any issue with NH3 absorbtion.

Ill try to look for that info latr, when more time is available

James Ikanov - 27-1-2017 at 14:40

I have a hypothetical that I was curious about for the acid production side of this question.

Could pumping in a close to stoichemetric mixture (for NO2) of ozone and nitrogen increase yield?

It seems a simple water electrolysis system hooked into a decent ozone generator could give you a low pressure source of fairly pure oxygen if arranged correctly. Pure nitrogen is a harder question, but I figured I'd get the easy stuff out of the way first.

I'd also be interested to know how much more effective H2O2 makes the absorption process, since such an electrolysis device offers the option of preparing high test peroxide before hand as a feedstock for the acid.

plante1999 - 27-1-2017 at 19:30

If you actually intend on keeping the nitrogen as nitrite/nitrate, lead dioxide is probably the best scrubber you can get easily. However, if you want to remove all the NOx fumes, I'd recommend sulfamic acid. For convenience, urea can do a decent job too.
 Pages:  1