Sciencemadness Discussion Board - The Electrochemical Oxidation of Ammonia to Potassium Nitrate

The Electrochemical Oxidation of Ammonia to Potassium Nitrate

WGTR - 27-12-2016 at 08:24

I've prepared a short document outlining my efforts towards producing potassium nitrate from ammonia and potassium hydroxide. They have met with some degree of initial success. This is a continuation of an earlier post here: https://www.sciencemadness.org/whisper/viewthread.php?tid=48...

It has a lot of room for improvement, but here it is for comments and criticism, in preparation for the next revision. Feel free to address any areas that you think need improvement or elaboration:

Attachment: The_Electrochemical_Oxidation_of_Ammonia_12_26_16.pdf (5.3MB)
This file has been downloaded 3085 times

I think the mathematical parts could use more elaboration. There's too much left to the reader to try and figure out.

The yield may be much better than reported, due to nitrite not being included in the final yield. From the original reference, it appears that nitrite is formed initially, then nitrate begins appearing. In this case, better yields of nitrate could be achieved just by running the experiment longer.

References:

Attachment: Traube_et_al-1904-Berichte_der_deutschen_chemischen_Gesellschaft.pdf (424kB)
This file has been downloaded 1326 times

Tsjerk - 27-12-2016 at 09:03

Very nice and clear write-up!

I think at least a part of the low yield can be explained by the less than perfect diafragma, maybe you can have a look of you can get agarose. This works really well to make a salt bridge when boiled in 2M NaCl (or KCl). It is sold as vegetarian gelatine. I made bridges with agarose solidified in a plastic tube.

Incorporating the cathode in the agar also helps against corrosion, I got that trick from guys using gelatine, but agarose is cleaner.

Chemetix - 27-12-2016 at 09:35

Very informative trial, clever use of cups. I have recently played with "water saving crystals" as gels. I think it's polyacrylamide but can't be sure, certainly wasn't CMC.
You can observe a distinct purplish tinge to the flame by the way.

[Edited on 27-12-2016 by Chemetix]

WGTR - 28-12-2016 at 09:51

Thanks for the comments!

Quote: Originally posted by Tsjerk

Very nice and clear write-up!

I think at least a part of the low yield can be explained by the less than perfect diafragma, maybe you can have a look of you can get agarose. This works really well to make a salt bridge when boiled in 2M NaCl (or KCl). It is sold as vegetarian gelatine. I made bridges with agarose solidified in a plastic tube.

Incorporating the cathode in the agar also helps against corrosion, I got that trick from guys using gelatine, but agarose is cleaner.

I've used a salt bridge before; the problem is that the resistance is usually too high to make it work, and overheating the bridge is a common problem. That's why I'm focusing on identifying a suitable thin membrane. That also helps with energy efficiency, as the voltage drop is lower. Worst case efficiency scenario, we're looking at about $2 of electricity per kg, assuming about $0.08 per kW/Hr. It should be far less than this, however. I'd like to see how efficient I can get the design with a paper membrane. That can't be beat for "cheap and available". I may have to go to something better later on, however.

I'm not sure if agar will handle the ammonia and hydroxide solutions. It might, but I haven't tried this yet to say either way.

The cathode actually produces a lot of hydrogen during the process, and doesn't seem to oxidize per se. There may be some rough copper deposits being deposited on it. I'm not sure how I'd incorporate the cathode into a solid electrolyte, because of the voluminous gas production. The anode is curiously absent from corrosion. During initial operation, there is almost no visible gas being produced at all, as it's being absorbed by the ammonia. Later on, some gas bubbles may be seen here or there, but not anything similar to the cathode gas volume.

There was a slight yellow tinge to the anode solution afterwards. This may be a small amount of iron contamination from the anode. After crystallization, however, the KNO₃ is perfectly white; and the needles are hair-like, and remind me of cotton.

Quote: Originally posted by Chemetix

Very informative trial, clever use of cups. I have recently played with "water saving crystals" as gels. I think it's polyacrylamide but can't be sure, certainly wasn't CMC.
You can observe a distinct purplish tinge to the flame by the way.

[Edited on 27-12-2016 by Chemetix]

I have a video of it, but it's not too interesting. Just a quick "woooosh!", and then me jumping in surprise with the camera.

I thinking about how to arrange the next experiment. Anode surface area is important, generally the less current density, the better. The cathode current density doesn't matter too much, except that if it's too high it can present a significant overvoltage, and the accompanying loss of energy efficiency.

WGTR - 7-1-2017 at 03:19

Update with some observations, in preparation for another round of experiments and document revision (posted in case I get Hit By A Bus):

In the original reference, it was mentioned that the experiment was conducted in an ice bath. I always assumed this was because the cell was self-heating due to ohmic losses, and needed to be actively cooled to stay close to room temperature. However, I've found it necessary to keep the electrochemical cell cold for what I think are a number of reasons: The dissolved copper catalyst sometimes precipitates out, or the copper anode corrodes, leaving a flocculant precipitate of copper hydroxide. So long as the temperature is cold, this precipitate is stable. It can be filtered out at the end of the experiment, and re-used. At mildly elevated temperatures, however, in a strongly alkaline environment, the copper hydroxide dehydrates to copper(II) oxide. This seems more difficultly soluble in the reaction solutions, and is harder to keep suspended in the solution. Secondly, both ammonia and oxygen are more soluble in colder solutions, and less inclined to escape. Finally, there may be an increase in anode overvoltage at colder temperatures, but I'm not sure if this has any effect on the results as of yet.

I've found it advantageous to keep the anode solution stirred with a stir bar. This ensures a fresh solution being continuously conveyed to the anode surface. Without this effort, the high current density areas of the anode will eventually start to evolve gas. With gentle stirring, no anode gases are observed. The stirring also keeps any copper hydroxide precipitate in suspension, and keeps the solution saturated with copper.

I've noticed different results when changing the order in which copper is added to the solutions. If copper carbonate is dissolved in concentrated ammonia, and then the dark blue ammonia solution is diluted out in the anolyte, the resulting anolyte has a clear blue coloration. Under cell operation, after about 1-2 hours the solution will turn clear, the copper being precipitated onto the cathode as a copper sponge. At this point the anode will start gassing in an unstirred solution. This is true even when using a Celgard membrane to separate the cathode and anode compartments, as the blue ammonia solution will pass through this membrane easily. In addition to the obvious hassle of removing the copper sponge, the increase in cathode surface area increases cathode efficiency for the reduction of nitrate back to ammonia, which we don't want. If the cathode current density remains high (perhaps 1A per sq. cm.), the reduction of hydrogen competes heavily with the other side reactions, giving better overall results for conversion to nitrate.

If ammonia is added to the anode compartment first, and then copper hydroxide under stirring, the color change of the solution is almost imperceptible, as not much copper dissolves. What does dissolve, however, appears to be enough to catalyze the conversion of ammonia to nitrate. Very little copper deposits on the cathode, not enough to affect cell operation.

Instead of steel (which can be used), copper was used for the latest anode configuration. This helps ensure that the solution remains saturated with copper, and provides an initial bit of copper to get started, if necessary. I observed that the copper anode would begin to corrode into a voluminous precipitate of copper hydroxide if the pH of the anolyte was allowed to drop below 13. As ammonia alone won't provide this level of alkalinity, periodic additions of potassium hydroxide should be made during cell operation. If the solution is alkaline enough, no apparent anode corrosion occurs, other than a black oxide coating.

A Celgard membrane was used instead of paper, to see how this would affect results. It turned out to work differently. As potassium migrated into the catholyte, the solution volume increased in the cathode compartment through osmotic pressure, until it was possible to overflow the cathode compartment if the experiment was left unattended. This kind of effect was not noted when using a paper diaphragm. A bleed hole to handle the overflow can circulate excess catholyte into the anode compartment, also serving to help keep the anolyte pH elevated, allowing a longer period of time between alkaline additions. Further experimentation will need to be done to determine whether it's possible to remove the diaphragm altogether, although keeping it will likely benefit the yield of nitrate.

[Edited on 1-7-2017 by WGTR]

XeonTheMGPony - 21-3-2017 at 19:12

Would a Mercury bridge work?

WGTR

Sulaiman - 22-3-2017 at 00:47

I did a little experimenting with electrochemical primary cells,
mostly based on variations of the Daniell cell,
which has the problem of the membrane (porous pots in my case) fouling up.

For the Daniell cell the membrane can be dispensed with if a vertical rather than harizontal electrode separation is used.
See 'Gravity Cell' here https://en.wikipedia.org/wiki/Daniell_cell

So, if your two electrolytes have different densities,
you may be ale to use this method ?

[Edited on 22-3-2017 by Sulaiman]

WGTR - 22-3-2017 at 07:32

Quote: Originally posted by XeonTheMGPony

Would a Mercury bridge work?

I think in this case we want ionic conductivity, but not electrical conductivity. Otherwise the mercury becomes the electrode itself, which defeats the point of a separator, if I'm understanding your suggestion correctly. We actually want to keep the alkali at the anode, which is the opposite of what normally happens. During operation, I periodically have to take some catholyte, and add it back to the anode compartment.

Quote: Originally posted by Sulaiman

I did a little experimenting with electrochemical primary cells,
mostly based on variations of the Daniell cell,
which has the problem of the membrane (porous pots in my case) fouling up.

For the Daniell cell the membrane can be dispensed with if a vertical rather than horizontal electrode separation is used.
See 'Gravity Cell' here https://en.wikipedia.org/wiki/Daniell_cell

So, if your two electrolytes have different densities,
you may be able to use this method ?

Thanks, I'll give some thought to this, to see if it's possible. I think the anolyte needs to be stirred, however.

I haven't forgotten about this project, but lately I've been clobbered by car problems and a heavy work load. Life, and all that... I have, however, prepared an insulated container for the next experiment, so that I don't have to keep adding ice to it every couple of hours. As soon as I get some breathing room, I'll get back on it.

XeonTheMGPony - 22-3-2017 at 08:01

In the mercury chloride cell the mercury is actually an ion filter.

it bridges two separated cells, - in the water side and positive on the chloride side.

- generates sodium hydroxide and H2 gas, + side produces chlorine and no loss of water other then evaporation.

but it works on the base sodium is soluble in mercury, not sure if such soluble ions are generated in this action.

do you have schematics? of the cell

[Edited on 22-3-2017 by XeonTheMGPony]

WGTR - 22-3-2017 at 09:27

I think I understand what you mean, but I'm not sure it will work.

With a mercury separator, the potassium (or sodium and probably copper) ions will get reduced and form an amalgam, which, if stirred vigorously enough, will be available to the other side of the cell to act as the anode, allowing fairly high purity alkali hydroxide to be formed. That's the theory. However, sodium amalgam will also directly reduce the nitrate that was just formed at the anode, causing a loss in efficiency.

I think it's better to use a mechanical separator in this case.

Melgar - 17-9-2017 at 19:59

@WGTR: I just remembered that I found this paper and figured that SOMEONE here would be able to use it. It never generated a ton of interest, but it seems perfect for what you need:

http://www.sciencemadness.org/talk/viewthread.php?tid=73150

Polyvinyl alcohol is extremely cheap if you know where to look, because the Chinese government subsidized its production, since it's a type of plastic that doesn't require oil to make. (It can be made from coal, via propylene, I believe.) It's in all sorts of stuff, including this really shitty dollar-store glue that's sold all over the place here in NYC, but probably not everywhere.

knowledgevschaos - 11-6-2024 at 20:51

Copper carbonate dissolved in ammonia sounds a lot like Schweitzer's reagent (Cu(NH3)4(OH)2), which is used to dissolve cellulose. Is it possible that this is dissolving the cellulose in your membrane, damaging it?