## The trouble with neodymium...

Pages:  1  2    4  ..  6 Poppy - 12-11-2011 at 08:51

Hello, so has your method really succeded? good job!!
Thank you!

As in the case of the autoprotolysis I meant some Fe(OH)3 changes in situ into Fe2O3 because the autoprotolysys mechanism, as evidenced by the Schikorr reaction
Sorry for the misleading with the equation for O(2-).

Schikorr reaction starts at about 100ºC and increases as temperature goes up. At temperatures <100ºC the conversion of Fe(OH)3 into Fe2O3 is not even likely to occur, but my advice is to beware of high temperatures!
------------------------
Blogfast:
I get it now, Fe(OH)3 = Fe2O3.n H2O = HFeO2 = FeO(OH) in a different notation.
But powdered dry Fe2O3 would not turn into Fe2O3.n H2O by damping in water correct?
Heating 2 HFeO2 (the mud from reaction of FeIII ions with strong alkali) to 200ºC dehydrated it into Fe2O3.
Would this dry Fe2O3 work too???

For those who wants Fe(OH)3 but are not dealing with H2O2, you can prepare it without H2O2, following this long road production, starting with Schikorr reaction and then drying.
------------------------------------------
2 (Fe2+ → Fe3+ + e–) (oxidation of 2 iron(II) ions)
2 (H2O + e– → ½ H2 + OH–) (reduction of 2 water protons)

To give:
2 Fe2+ + 2 H2O → 2 Fe3+ + H2 + 2 OH–
Adding to this reaction one intact iron(II) ion for each two oxidized iron(II) ions leads to:
3 Fe2+ + 2 H2O → Fe2+ + 2 Fe3+ + H2 + 2 OH–

For electrical neutrality, OH(-) are shown:
3 Fe2+ + 6 OH– + 2 H2O → Fe2+ + 2 Fe3+ + H2 + 8 OH–
3 Fe(OH)2 + 2 H2O → Fe(OH)2 + 2 Fe(OH)3 + H2
Autoprotolysys mechanism: Its essential in the formation of the final product and occurs but only at temperatures about 100ºC or above.
OH– + OH– → O2– + H2O
acid 1 + base 2 → base 1 + acid 2, or also,
2 OH– → O2– + H2O
Then that all progresses as
3 Fe(OH)2 + 2 H2O → (FeO + H2O) + (Fe2O3 + 3 H2O) + H2
3 Fe(OH)2 + 2 H2O → FeO + Fe2O3 + 4 H2O + H2
3 Fe(OH)2 → FeO + Fe2O3 + 2 H2O + H2
The former Schikorr reaction being
3 Fe(OH)2 → Fe3O4 + 2 H2O + H2

I will make a test with Fe(OH)3 as soon as providing some HCl and H2O2 to see if it protolyses into pure Fe2O3!!
For that I'll just heat the ferric hydroxide mud until dryness and somehow check for pure Fe2O3 for its beatyful red color.
Another note, when preparing conventional Schikorr reaction for the production of magnetite, leaving the heating vessel with any opening causes oxygen to enter like crazy and get absorbed by the magnetite turning it into orange Fe2O3, which is a mixture consisting of mostly Fe2O3 and a little bit FeO.
2 Fe3O4 + 1/2 O2 --> 3 Fe2O3
The reaction with atmospheric oxygen was so strong that at 100ºC finely ground magnetite was actually burning!
The test for this orange Fe2O3 was made diluting it into 30% HCl and observing that some residue of FeO (black) was not atacked even with overweighted HCl.
Further burning of the oxide mix seemed not to oxydise anymore FeO.
The Fe2O3 can be dissolved in sulphuric acid and then put to react with strong alkali to produce Fe(OH)3.

[Edited on 11-12-2011 by Poppy]

[Edited on 11-12-2011 by Poppy]

Wizzard - 12-11-2011 at 10:51

Alright, I'm first going to admit I have absolutely no formal training in chemistry- I'm strictly a math major

But I noticed over a weeklong evaporation operation- The nd sulfate fell out first, in both cases, so much so that it was able to crystallize, and appears nearly Fe free.

Iron sulfate is VERY soluable at ~95*C, the Nd sulfate is not... Am I the only person evaporating the water out slowly at just below boiling, causing the Nd to drop out selectively?

I can post some pictures, I'm working on narrowing down the time spend evaporating- So far, I manage to drop out all the ND sulfates, and then the iron sulfates right down on top of it (but the layering is very distinct).

blogfast25 - 12-11-2011 at 13:39

 Quote: Originally posted by Poppy Would this dry Fe2O3 work too???

No, it needs to be sufficiently hydrated (‘freshly’ precipitated) to still react with low concentrations of H3O+: Fe(OH)3 + 3 H3O+ === > Fe3+ + 2H2O.

 Quote: Originally posted by Wizzard But I noticed over a weeklong evaporation operation- The nd sulfate fell out first, in both cases, so much so that it was able to crystallize, and appears nearly Fe free.

Welcome to the murky world of hobby chemistry, mathman!

Do NOT let appearances fool you: at least with the double salt (presumed Nd2(SO4).K2SO4.2H2O) I’ve obtained a nice, pink product which on further scrutiny contained… substantial amounts of iron! There seems to be co-precipitation going on. Whether that is also the case with ‘naked’ Nd2(SO4)4 I cannot vouch for. Test by converting your sulphate to hydroxide by treating with strong alkali. Filter off the precipitate and wash it, then dissolve it in clean (iron free) HCl. Add a bit of 9 % H2O2 to oxidise any Fe2+ to Fe3+. Test visually or with KSCN.

I wonder also about the iron content of Nd oxalate precipitated from iron rich solutions, see Mr Home Scientist higher up and kmno4 in a separate Nd thread...

£\$£\$£\$£\$£\$£\$

Well, here’s my ‘magnet chloride’ after heating and standing overnight:

Plenty of precipitated Fe(H)3 but the dark red-brownish supernatant liquid is the tell tale that plenty of Fe didn’t drop out. pH was less than 1 at that point, so no surprise there.

After that the work became more of a rescue operation than a chemical separation!

The slurry filtered only with great difficulty and adding some freshly precipitated Fe(OH)3 did not bring the pH up enough to precipitate the remaining dissolved iron. I had to resort to manually adjusting the pH to pH >= 4, first with small amounts of 5 M NaOH, the final adjustment with household ammonia. At that point the slurry became much thicker and filtered better. The filtrate contained all the Nd, but still tested slightly positive for Fe3+, mainly due to a bit of colloidal Fe(OH)3 running through the filter I believe. The raw NdCl3 has now been precipitated one last time and will be redissolved in HCl tomorrow.

In conclusion I’d say that separating Nd3+anf Fe3+ by playing to the Fe(OH)3 equilibrium works best for removing relatively small amounts of iron, not to remove iron in concentrations 7 times or larger than that of Nd3+.

[Edited on 13-11-2011 by blogfast25]

Wizzard - 13-11-2011 at 09:35

Will test once I get home It all dissolved in 15cc of very cold water (1-2*C), about .5g of the physically extracted Nd salt (scraped from the bottom- It's much softer than the very hard iron sulfate. Very faint purple color. I'm working on getting the rest out, by watching the evaporative process carefully.

I only was seeing purity because the Nd salt crystallized- I know some purity is needed for this. There were small 2-3mm crystals among the mess on the bottom of the container.

blogfast25 - 13-11-2011 at 09:47

Ok wizzard, what was your 'mother liquor' made from? 'magnet sulphate'? Did you simply dissolve a magnet in H2SO4? Then what do you do next?

Also, when you refer to 'iron sulphate', do you mean ferric or ferrous sulphate?

Wizzard - 13-11-2011 at 19:32

Mother aqueous solution was 'magnet sulfide', with the nickel COMPLETELY removed manually.

Then left to dissolve, in full, the remaining slush of unknown material and boron removed. (pic 1)

From this, total evaporation with small addition of distilled solution, but still acidic (I did not test pH).

From this, the layered substance in photo 2.

Scraping the purple Neodymium Sulfate into a glass (a small sample amount), dissolved and nearly frozen, then dried on a hot plate over the course of 6 hours, the resulting pure crystals formed- Picture 3 in 4 parts

My method seems low yield, but very easy! I prefer manual means- Not so much measurement (but I would like to learn). I'm a scientist for sure, but a chemist I am not!

The largest (centered) crystal cluster is about 7-8mm across.

@Blogfast- Ferrous, Iron(II) of course

[Edited on 11-14-2011 by Wizzard]

blogfast25 - 14-11-2011 at 05:12

So, if I understand well you simply slowly evaporated the ,magnet sulphate' until crystals of Nd2(SO4)3 started to appear? Certainly the Nd sulphate is much less soluble than ferrous sulphate but it's interesting you managed to isolate the minority constituent that way and of seemingly good quality too. AFAIK no one else has applied that approach here yet. Personally I had trouble dissolving the magnet in 50 % H2SO4, not sure why...

I would still urge you to test one crystal for presence of iron. If contaminated you could recrystallise for instance, or did you already do that?

[Edited on 14-11-2011 by blogfast25]

Wizzard - 14-11-2011 at 05:52

Yes, after filtering out the insoluables, the Nd fell out of solution before the Fe did with slow evaporation at 90-95*c.

I will test for iron once I get my hands on some good hydrogen peroxide... My lab chemical supply is still expanding.

I have not recrystallized a second time, only a first. I could do it a second, but I fear there is little enough already! I'll wait until the larger batch I have going right now finishes.

blogfast25 - 14-11-2011 at 07:03

 Quote: Originally posted by Wizzard Yes, after filtering out the insoluables, the Nd fell out of solution before the Fe did with slow evaporation at 90-95*c. I will test for iron once I get my hands on some good hydrogen peroxide... My lab chemical supply is still expanding.

It's quite amazing you got crystals rather than microcrystals (powder). You must have gotten a bit lucky there, I'm guessing; right water-to-salt ratio and all that... It's also surprising your that your ferrous sulphate wasn't more oxidised by air and heat: it then has a tendency to drop out.

3 or 9 % H2O2 from a pharmacy will do. It's dirt cheap.

[Edited on 14-11-2011 by blogfast25]

Wizzard - 14-11-2011 at 07:59

Well I know from experience that ferrous sulfate oxidizes with contact to air in solution, unless kept mildly acidic... I've grown 1cm crystals in 2cm tubes I'm no chemist, but I do know a thing or two about keeping one's ingredients pure- I've been growing crystals of various salts since I was in high school.

And I was unaware that pharmacy grade H2O2 is good enough! Off I go!

Attached good macro pic, with scale.

[Edited on 11-14-2011 by Wizzard]

blogfast25 - 14-11-2011 at 08:32

Very nice crystals indeed: they look good enough to eat!

Do you recall:

* weight of the magnet
* amount and concentration of the sulphuric acid used?

Did you get any early precipitation (of Nd2(SO4)3) during dissolving?

Acidity slows the oxidation of Fe (II) down, that's very true. But you should have seen my acidic stockpile of FeCl2 crystals after a couple of weeks: it's a brown mass of Fe(OH)3, ferric oxychlorides and what not: I bet it would test negative for Fe (II)! Just for standing in open air...

Wizzard - 14-11-2011 at 09:43

The approximate weight of the magnets, before breaking up and removing the plating, was about 200g- I'd say 125-150g of magnet material went in.

I then added about equal parts distilled water and 98% sulfuric acid, maybe 50ccs total. When the boiling would stop, I would add more conc. sulfuric acid or water (4ccs at a time), depending on what the solution wanted, and I did this many times. It sometimes would run out of acid (and a light tinge of yellow iron hydr/oxide would form on the meniscus) or it would cease when it ran out of water. When it stopped and only small bubbles would come out of the insoluable slush at the bottom for some hours (and sometimes only when shaken slightly), then it was time to filter.

No visible precipitation occurred as the solution was made, but there could be some in the insoluable mess. It did sometimes have a tinge of purple. I am working on extracting what remaining material I can from it.

The solution was kept at room temperature while dissolving- Only in the end did I heat it (to wrap up any action of the sulfuric acid, at least 90*C) before I froze it (about 5*C), then filtered it cold (maximizing output of the Nd sulfate).

barley81 - 14-11-2011 at 10:07

 Quote: Originally posted by blogfast25 Fenton's reagent doesn't involve Fe (IV).

Really? I thought it did, after reading this:
http://woelen.homescience.net/science/chem/solutions/fe.html

blogfast25 - 14-11-2011 at 10:35

 Quote: Originally posted by barley81 Really? I thought it did, after reading this: http://woelen.homescience.net/science/chem/solutions/fe.html

I think it's a matter of debate, TBH. If Fe IV really does exist it must be very unstable.

@Wizzard:

I see. Well, maybe next time I'll try again with fully peeled magnets and a stoichiometrically determined amount of H2SO4...

[Edited on 14-11-2011 by blogfast25]

Wizzard - 14-11-2011 at 11:06

I'll run another batch, fully documented

I have stacks of old HDDs and all the motivation I need for beautiful, hexagonal air semi-stable (left one on the paper overnight- no water loss!!) color changing crystals of an element I have very little of in my collection.

blogfast25 - 14-11-2011 at 12:37

Lucky mathman! I asked a computer repair shop to keep any duff HDDs for me (for a token price) but it turns out they 'forgot'. Barstools.

The crystals, being of a compound that's not particularly highly soluble nor insoluble, should be neither particularly deliquiescent nor efflorescent, IMHO... My NdCl3 on the other hand deliquiesced like mad: very hygroscopic.

Poppy - 14-11-2011 at 12:50

Wizzard:
I would assume you just filtered the liquor out of the muddy magnet solution: this little ammount of liquor justifies the nicely dissolved Nd2(SO4)3 in there, and very small yield for 125g?
Also we must point out your evaporation method is a very good way for 1 stepped refining of neodimium sulphate., good job.

Wizzard - 14-11-2011 at 13:04

Thanks! Very small yield was due to my physical method of separating the Nd salt from the Fe salt- I merely scraped the easiest-to-get-at soft/wet and purple neodymium sulfate crystal from the hard, green ferrous sulfate.

Low yield, yes, but I'm still extracting the salts from that aqueous solution- This was just a quick and dirty trial. I'd also still say there's 50% more in my evaporation dish- There was a good quantity of smaller crystals, but I do need to remove the film which grew on the edges before I reattempt this (and recycle the dried and extracted but not removed Nd sulfate).

Partial evaporation yielded the following crystals- Such a lovely color! Recrystallizing tonite- It doesn't take long. Note the green tinge in the center mass, and along the edges- This is the start of when the iron sulfate crystallizes- Time to stop evaporating!

[Edited on 11-14-2011 by Wizzard]

blogfast25 - 15-11-2011 at 06:27

Wizzard:

Although I had come to believe that your crystals must be high purity and essentially free of iron, now I’m not so sure anymore or at least a little befuddled. A spammer inadvertently dragged up this old post on neodymium sulphate:

(go to the last post by kmno4)

The colour obtained by you seems highly unusual.

Wizzard - 15-11-2011 at 07:36

My lighting conditions are a normal 60W incandescent, and big-bulb commercial flourescent lighting. Perhaps I'll get a sunlight shot today- That will be nice for comparison.

My are certainly a bit more pink/red than those purple/pink ones you link to- But I will say I cant wait to grow a single crystal that size

[Edited on 11-15-2011 by Wizzard]

Poppy - 15-11-2011 at 08:19

On the the leftover liquor:
After separating in part the neodimium from iron from a magnet solution via the sulphate method, a leftover supernatant was obteined which seemed to contain, by its color, initially at least a big portion of Fe III sulphate, Fe II sulphate, unnoticeable by eye, and some neodimium as well.
Theoretical obtention of Nd2(SO4)3.8H2O would be slightly round 50g, but as there was obteined just 44g (oops, thats really close to the goal, and because some 5g were lost propositally =P) I am inclined to believe some neodimium was left in the supernatant solution (at least 6g) as well as some iron is present in my final powder.
The main reason is that the leftover supernatant gradually, and very slowly, formed crystals which took much longer than conventional ferrous sulphate crystals to form (6 hours for a supersaturated solution of FeSO4 and one week to badly precipitate some of the strange liquor).
Here you can see the leftover after about 2 days:

Here you can see the leftover solution after about 1 week:

Those crystals were filtered and dissolved again to see what happens: the result was a slightly pink solution from those green crystals!!!!

That after decanting (seeing pink depends on your mood on that day...)

The green crystals initially become decoloured when washed
then gives the solution above.

So there must be double salts of Nd, Fe in the solution which modifies how the precipitation of pure ferrous sulphate crystals would do, and my obteined Nd salt powder is highly contamined with Pr or Fe in order of having its yellow color, even after it was thorughly washed with hot water when it was prepared.

And this is the Nd sulphate under flash and daylight, theres no real difference.

[Edited on 11-15-2011 by Poppy]

blogfast25 - 15-11-2011 at 08:24

 Quote: Originally posted by Poppy On the the leftover liquor: After separating in part the neodimium from iron from a magnet solution via the sulphate method, a leftover supernatant was obteined which seemed to contain, by its color, initially at least a big portion of Fe III sulphate, Fe II sulphate, unnoticeable by eye, and some neodimium as well. Theoretical obtention of Nd2(SO4)3.9H2O would be slightly round 50g, but as there was obteined just 44g (oops, thats really close to the goal, and because some 5g were lost propositally =P)

Can you describe what you did slightly more in detail?
Also, literature describes an octahydrate (.8H2O), not a nonahydrate (.9H2O)...

Edit:

Dang!! Quite a bit of confusion on the colour of neodymium sulphate: higher up in this thread:

 Quote: Originally posted by DerAlte All references I have seen say Nd sulphate is red (my emph.) and Sm is light yellow, Ce white (colourless); and these were the colors I got. All these sulphates are very difficult to dissolve at 0C, the most soluble point. Once dissolved, heating to 100C usually will precipitate them. Better crystals can be obtained by slow evaporation. Regards, Der Alte

This photobucket entry which shows crystals very much like Wizzard’s:

http://media.photobucket.com/image/neodymium%20sulphate%20cr...

And here, slightly more pinkish:

http://www.metall.com.cn/ndso.htm

[Edited on 15-11-2011 by blogfast25]

[Edited on 15-11-2011 by blogfast25]

Wizzard - 15-11-2011 at 11:59

It's quite possible there's some other Lanthanide impurity involved there- Neodymium magnets dont need absolute purity, as far as the Nd is concerned, I think (and was this also mentioned previously in this thread?).
blogfast25 - 15-11-2011 at 12:27

 Quote: Originally posted by Wizzard It's quite possible there's some other Lanthanide impurity involved there- Neodymium magnets dont need absolute purity, as far as the Nd is concerned, I think (and was this also mentioned previously in this thread?).

Actually I believe it was praseodymium that was mentioned by kmno4. Absolute purity of Nd for magnets is probably not a prerequisite but to affect colour you'd need substantial contamination (I think).

Other things that might affect colour could be:

* light (UV content): already accounted for
* powder (microcrystaline) v. macro crystals
* variability in hydration (.xH2O). It's well known the lower hydrates (generally speaking - not spec. to Nd) tend to form at higher temps. and vice versa.

Looking at the body of evidence, yours seem to have the 'right' colour though...

Wizzard - 15-11-2011 at 13:12

What I'll experiment with next will be how much alcohol I can add to the water solution of high-Nd/low-Fe and plot the soluability tables with different mixtures of alcohol/water down to -10*C (the limits of my scientific "Wal-Mart" mini-fridge).

Maybe with some alcohol, I can lower temperatures and dissolve more Nd sulfate, and drop out more Fe sulfate.

blogfast25 - 15-11-2011 at 13:36

You assume Nd sulphate is soluble in alcohol, or at least alcohol/water mixtures. Any evidence you've come across for that?
Wizzard - 15-11-2011 at 14:02

I know I read some of the other lanthanide sulfates are alcohol soluable- To what degree, I dont know. But I do know Lanthanium and Cerium both have similar soluability curve in water, with respect to temperature. I'll just do some experimentation, and get back to you on that!

http://chemicalland21.com/industrialchem/inorganic/LANTHANUM...
"Lanthanum sulfate is a white crystals. It is slightly soluble in water, soluble in alcohol" down the page a bit.

This could have some interesting data, if I could view it...

[Edited on 11-15-2011 by Wizzard]

blogfast25 - 16-11-2011 at 07:04

So in a sense, assuming reasonably that this applies also to others in the 'cerium group' (as they were once called), you're risking to acually increase solubility!

But it's always, always, always worth a shot...

[Edited on 16-11-2011 by blogfast25]

Wizzard - 16-11-2011 at 07:39

Oh, yes- I'm looking to increase solubility below 0*C to push ever more ferrous sulfate out of solution.

Just reading through some other threads, came across a gas chemistry link, this may be of interest to this thread- I think it may be applied to Nd2O3 to possibly generate Nd metal!

http://mattson.creighton.edu/PipetteRxn/index.html

[Edited on 11-16-2011 by Wizzard]

[Edited on 11-16-2011 by Wizzard]

blogfast25 - 16-11-2011 at 08:53

 Quote: Originally posted by Wizzard Just reading through some other threads, came across a gas chemistry link, this may be of interest to this thread- I think it may be applied to Nd2O3 to possibly generate Nd metal! http://mattson.creighton.edu/PipetteRxn/index.html

Unfortunately, the reduction of Nd2O3 with hydrogen is far from possible. It is possible with the oxides of Fe, Co, Ni and a few others but not with the REs.

Allow me to explain. Take a generic reduction by hydrogen:

MO<sub>n</sub> + n H<sub>2</sub> === > M + n H<sub>2</sub>O

This can only proceed if the change in Gibbs Free Energy, ΔG = ΔH - T ΔS is negative (< 0): systems seek a state of minimal Free Energy. Assuming for argument’s sake that the entropic effects for this reaction are small (so T ΔS ≈ 0), then ΔG ≈ ΔH < 0.

ΔH can be found by adding the Heats of Formation of the reaction products and subtracting from this the sum of the Heats of Formation of the reagents. In the cases of Fe2O3, CoO and NiO this ΔH is indeed negative and these reactions can proceed.

But in the case of Nd2O3 its HoF is too large and ΔH > 0 (in reality we should take the entropic effects into account). It’s also for this reason that Nd2O3 cannot be reduced by aluminium or magnesium, both far stronger reducing agents than hydrogen.

REs are generally produced by electrolysis. But reduction of the trifluorides with magnesium is also possible (for the same reason: ΔH < 0). It’s this latter reaction Mr Home Scientist and me will endeavour to produce Nd metal. If we ever get round to making enough NdF<sub>3</sub>, that is!

[Edited on 16-11-2011 by blogfast25]

Wizzard - 16-11-2011 at 09:14

Well then I learn something every day! Thanks for the write-up.
MrHomeScientist - 16-11-2011 at 22:12

In the midst of all this talk about separation and purification, I wanted to get back on track with the final step of this process to obtain Nd metal. I have a few crops of NdF3 already made, so I wanted to try one out to see how it went.

These are my samples of NdF3 (under fluorescent lighting), each vial from a different batch of magnets. The one on the left is the one I showed in a previous post (distinctly green), the middle one is a very nice white, and the righthand sample is a bit more brown. The rightmost one is what I used in this experiment - since it's a bit more contaminated it was a good candidate for initial experimentation.

That righthand vial contained 7.1g of NdF3. I mixed this with 1.4g of Mg powder - somewhat more than the stoichiometric amount, but I wanted it in excess. I placed this in a fused silica crucible in which the reaction was to take place.

Since the total mass of the reactants was only 8.5g, I decided that normal thermite procedures wouldn't be sufficient (as per an earlier suggestion). So, I placed this crucible on a ring stand and heated it from below with a propane torch. I was hoping to get to the initiation heat just from this, but after about 15 minutes of blasting the bottom of the crucible with the torch there was no change. So, to get things started I grabbed a small strip of Mg ribbon with tongs, lit this with the torch, and dropped it into the hot crucible. This did the trick! The whole interior lit up with a nice orange glow and a good amount of smoke as the reaction took place. It reminded me of chromium thermite I've done in the past - lots of light and smoke but no sparks. I was very excited to see something happen Here's what it looked like shortly after the smoking stopped. You can see the ring stand and blue propane bottle beneath it.

I waited about an hour to let this cool, then took it inside to see what I could find. Disappointingly, I didn't find much at all. The vast majority was just a dark grey powder.

Clockwise from top left, here's what each pile is.
Top left (TL): crust from top of the reaction. It's mostly white MgO.
TR: Bulk of the material below that. It's pretty uniformly grey.
Bottom right (BR): Much blacker material from the bottom of the reaction products.
BL: A small amount of scrapings from stuff stuck to the bottom of the crucible.

Now here's where things get interesting.

In scraping away the black material from the bottom, I revealed an extremely thin layer of a very shiny metal!

I scraped some of this out and picked out some of the larger pieces to put in a vial to see them better. "Larger" of course meaning extremely tiny. This can't be more than 50mg. There were more shiny dots in the bottom left pile of the previous picture.

What's encouraging is that these flakes are not attracted to a strong magnet! That means they can't be iron, so I tentatively think that these are actual (miniscule) chips of Nd metal!
Some of the rest of the reaction material did seem to be affected by a magnet, so as I suspected there's certainly some iron contamination in this batch. Magnetite could explain some of the black coloration.

Now to test this metal, I placed a small amount of the black powder with shiny chips in a test tube and added a few drops of sulfuric (battery) acid. It immediately produced a good amount of gas and a large amount of heat. Some of the black material didn't dissolve. I was hoping to get a pink solution to confirm Nd, but it still looks colorless to me. It was such a tiny amount though I bet it'd be impossible to see if it was there anyways. I'd think testing with oxalate should show me something, even at low concentrations. I have yet to try this. It's hard to say if I actually made Nd metal because it's nearly impossible to separate the shiny bits for testing alone since they are so tiny. What's in the vial is by far the best of what I could collect.

So I suppose this confirms the difficulty of getting this reaction to work at small scales. I imagine the whole mess didn't get hot enough, despite my additional propane torch heating, to react all the way or produce molten metals. Something obviously happened to the reaction mixture, but it just didn't make an appreciable amount of metal. Just a lot of grey and black powdery garbage, plus that promising tiny amount of shiny metal.

Chemistry Alchemist - 17-11-2011 at 01:30

Could the Nd that formed be from the heating with he torch? im sure if it was the thermite reaction, Nd metal will be spread through out the mixture but if it was only at the bottom, could that be from heating?
blogfast25 - 17-11-2011 at 06:23

Mr Home Scientist:

Wow! Much to talk about here, especially since I was going to U2U you advising NOT to proceed: much more about that later on.

You seem indeed to have obtained reaction and Nd metal.

Igniting the mixture by just heating would probably have required at least 800 C (my estimate). The high initial temperature might be what saved this reaction though. It was definitely the right idea to heat externally.

You claim to see MgO but the main reaction product should be MgF2 of course. I suggest in any case to grind down the slag and treat it with strong HCl. If the reaction took place then the dark material is likely caused by fine Nd metal dispersed through sintered MgF2. It should still react with strong acid though. With an MP of 1263 C your orange glow may have been just a little short of melting the MgF2 down.

One possibility is that the metal you found is unreacted Mg (Fe seems unlikely, as it was only a contaminant) and the only way to prove it is neodymium would be to try and dissolve it in as little H2SO4 as possible and see if you get a precipitate at BP, or by saturating the solution with K2SO4. If you go Nd2(SO4)3, it can be tested with oxalic acid (redissolve it in iced water first). MgSO4 is of course highly soluble over a wide range of temps. But it’s unlikely to be Mg, IMHO…

And here’s the strange news that I was going to U2U about.

Getting much closer to producing NdF3 now (I took receipt of the NH4HF2 this morning), I decided to do a full thermochemical calculation for NdF3 + 3/2 Mg === > Nd + 3/2 MgF2 (1), last night, with a view of estimating potential end temperatures. Well, that didn’t take very long! For some years I’ve been convinced that this reaction would be thermodynamically favourable (ΔG < 0) and with more than enough reaction enthalpy (ΔH << 0) to spare but that must have been based on an erroneous calculation or datapoint!

Wolfram Alpha coughed up a value for the standard heat of formation of NdF3 of - 1657 kJ/mol (which ‘sounds right’) and for MgF2 of - 1124 kJ/mol. Immediately it becomes apparent that the Standard Heat of Reaction for (1) = 1657 - 1.5 x 1124 = - 29 kJ/mol which really is negligibly small, almost ΔG ≈ 0. No entropic effects are taken into account but for reactions involving only solids these are usually small.

It really means that if the reaction proceeds AT ALL, strong external heating would be needed to get the metal and slag to melt and separate (whereas various thermite-like reactions generate enough heat to achieve this)

But it seems to me you’ve provided invaluable evidence that the reaction does proceed, albeit w/o generating much extra heat, otherwise you’d have obtained a neat slag puddle with the metal at the bottom.

The other point is that I found two literature references claiming NdF3 is purple/violet. I looked at your procedure again and believe your product(s) are relatively pure samples of neodymium trifluoride and that your colour is the right one, bar maybe slight distortion by FeF3. But that should have stayed in solution because it’s soluble though, bar a little occlusion/co-precipitation…

All in all, a big ‘thumbs up!’

[Edited on 17-11-2011 by blogfast25]

Wizzard - 17-11-2011 at 09:25

Second-round extraction and recrystallization of my magnet sulfate soup produced 5.45g of neodymum sulfate, plus .50g in 2 large single crystals- pics attached

The small one is 7x4x3mm, the large one is 9x6x6mm.

blogfast25 - 17-11-2011 at 09:36

Nice whoppers, Wizzard!

Are you gonna attempt some really giant ones?

Wizzard - 17-11-2011 at 09:54

You know it! I'm going to go a real nice writeup with at least 150g of magnet material, and walk-through the whole process, step by step... With the help of my new scale and other measuring devices. There's even still more Nd to be extracted from my soup. Noy much, but maybe another 3g or so I'd guess. 2nd round was very successful, though. I wish I had more hot plates...
Poppy - 17-11-2011 at 12:01

blogfast25 - 17-11-2011 at 12:17

 Quote: Originally posted by Poppy I'm jealous of your crystals

Thou Shall't not Coveth another Man's Crystals!

Wizz, I bet you could sell these, especially even larger ones, to the 'alternative medecine' crowd, you know? 'Astral spectrum changes', a bit of 'quantum this' and 'quantum that' and some good pics and hop: 20 bucks a pop!

Wizzard - 17-11-2011 at 12:56

Quite possibly. I dont take advantage of other people ignorance, however Plus I dont know if this sulfate is toxic or not.
Poppy - 17-11-2011 at 17:40

 Quote: Originally posted by Wizzard Quite possibly. I dont take advantage of other people ignorance, however Plus I dont know if this sulfate is toxic or not.

Yes every crystal really just filters incoming radiation and gives off specific ones of their kind, that could be useful if you manage to deal with some divine spectrometric relationship.
I was even planning on polishing a "ambient temperature thermal lens" crystal haha
Back on the neodimium issue, the sulphate method is trustworthy. My final powder was assumed to contain neodimium sulphate, but it dissolved so slowly in water you couldn't know. But diluting this with some hydrochloric acid gave a yellow brown solution under a tungsten lamp, and a green solutoin under a fluorescent lamp. There might just be some iron contaminants, but the sulphate method gives neodimium as the prevailing component of its admitted double salt.

Poppy - 17-11-2011 at 17:51

All products are going into a rebatch to see what happens in a schikorr like reaction with Nd hydroxide.
This product will then be slowly crystalised hopefully reaching wizzard's one quality.

MrHomeScientist - 18-11-2011 at 10:07

@Chemistry Alchemist: Yeah I was thinking the same thing too - since that's where the torch was hitting it was definitely hotter at the bottom. That, I think, fits well with blogfast's comments about this reaction needing a bit more heat to get results.

 Quote: Originally posted by blogfast25 You claim to see MgO but the main reaction product should be MgF2 of course. I suggest in any case to grind down the slag and treat it with strong HCl. If the reaction took place then the dark material is likely caused by fine Nd metal dispersed through sintered MgF2. It should still react with strong acid though. With an MP of 1263 C your orange glow may have been just a little short of melting the MgF2 down.

Indeed. I think just the top layer of white is MgO from the burning magnesium strip reacting with air. That's why the white is only on top but not further down. In the second to last picture you can see a lot of white powder on the sides of the crucible from this.

So you're thinking react all the products with acid, and it will leave behind the MgF2? Or dissolve everything and test with oxalate? MgOx appears to also be rather insoluble.

 Quote: Originally posted by blogfast25 One possibility is that the metal you found is unreacted Mg (Fe seems unlikely, as it was only a contaminant) and the only way to prove it is neodymium would be to try and dissolve it in as little H2SO4 as possible and see if you get a precipitate at BP, or by saturating the solution with K2SO4. If you go Nd2(SO4)3, it can be tested with oxalic acid (redissolve it in iced water first). MgSO4 is of course highly soluble over a wide range of temps. But it’s unlikely to be Mg, IMHO…

Right, I would think that any Mg metal down there would have been heated enough to react away. I thought that it might be Fe because the fluoride looked fairly contaminated, plus my recovered metal was such a tiny amount anyway.

 Quote: Originally posted by blogfast25 The other point is that I found two literature references claiming NdF3 is purple/violet. I looked at your procedure again and believe your product(s) are relatively pure samples of neodymium trifluoride and that your colour is the right one, bar maybe slight distortion by FeF3. But that should have stayed in solution because it’s soluble though, bar a little occlusion/co-precipitation…

I think that last bit is pretty likely. When making the fluoride I used a fairly concentrated solution of the sulfate. I can see why more dilute is better, and I'll make sure to do this next time. I've got another batch of sulfate ready to recrystallize and convert to the fluoride.

 Quote: Originally posted by blogfast25 And here’s the strange news that I was going to U2U about. ... It really means that if the reaction proceeds AT ALL, strong external heating would be needed to get the metal and slag to melt and separate (whereas various thermite-like reactions generate enough heat to achieve this) But it seems to me you’ve provided invaluable evidence that the reaction does proceed, albeit w/o generating much extra heat, otherwise you’d have obtained a neat slag puddle with the metal at the bottom.

That's quite interesting. It's good that you rechecked the theory! I think that is definitely going to be the way to go with this - heat the crucible up high enough, then initiate the reaction as I did before. To this end, I just ordered some fire bricks from Home Depot's website and I'm going to build a (crude) furnace around the crucible to keep the heat in. Hopefully this will be enough to get me to the melting points. Exciting times ahead!

[Edited on 11-18-2011 by MrHomeScientist]

Endimion17 - 18-11-2011 at 10:44

If you're about to try that melting scenario, consider argon, because if you don't use it, you will fail miserably, meaning the yield will be dreadfuly small, and you'll end up with your precious starting batch heavily contaminated.
Wizzard - 18-11-2011 at 12:11

Argon may work well. Just don't try nitrogen or CO2 Speaking of which, anybody know anything about lanthanide nitrides?
blogfast25 - 18-11-2011 at 13:58

 Quote: Originally posted by Endimion17 If you're about to try that melting scenario, consider argon, because if you don't use it, you will fail miserably, meaning the yield will be dreadfuly small, and you'll end up with your precious starting batch heavily contaminated.

Nope. Failure is by no means guaranteed w/o argon. See aluminothermic reactions of which I carried out more than a dozen on different metal oxides (Fe, Cu, Si, Ti, Cr, Mn, V, Nb and combined ones as well): despite end temperatures of around 3000 K, you usually get good quality metal (if you get everything right! - they’re not as easy as you might think) because the once molten slag (alumina) forms a perfect protective barrier for the metal regulus. No need for argon here neither, assuming you can get the MgF2 to actually melt (1263 C, not even that hard).

@Mr Home Scientist:

Yes, I suggest trying to recover as much Nd by extracting it from the slag. Nd and Mg should be easily separable with the saturated K2SO4 method. Careful not to use too strong acid (32 % HCl will be OK) or your MgF2 will start acting like fluorite: a HF generator!

Way to go with the furnace. I’m thinking of carrying out my reaction with my dinky toy charcoal furnace: that melts copper but only just…

There is however another ‘theoretical’ aspect that I haven’t touched upon.

Assume we heat a stoichiometric mix of NdF3 and Mg to above the MP of MgF2 and reaction occurs. Now we hold the crucible there at that temperature: in the melt the equilibrium 2 NdF3 + 3 Mg < === > 2 Nd + 3 MgF2 rules.

The equilibrium constant is given by Nernst:

ΔG = - RT lnK with ΔG the change in Free Energy (at temperature T) and K the equilibrium constant (at T):

K = ( a<sub>Nd</sub><sup>2</sup> x a <sub>MgF2</sub><sup>3</sup> / ( a<sub>NdF3</sub><sup>2</sup> x a <sub>Mg</sub><sup>3</sup>

With each a the activity of each species (we cannot use concentrations here)

We would have to calculate ΔG (at T) from the Standard change of Free Energy, ΔG<sup>298 K</sup> but examples with similar reactions show that the temperature correction, even for T > 1000 K, is usually small.

Although we don’t know the exact ΔG, we know from the above that it is likely to be quite small. Well, from Nernst we know that if ΔG ≈ 0, then K ≈ 1 or in plain English: there should be considerable free Mg in the metal phase. I thus predict that obtaining high purity Nd with this method is probably impossible.

In practical news, I’ve finally finished precipitating, filtering and washing the first batch of NdF3. Tomorrow it will be dried. So far it is very clearly lavender coloured.

And another 25 g of magnet material has been dissolved in HCl. I’m going the ‘sat. K2SO4’ route again, just to explore that a bit more. The double salt has been precipitated and will be further processed, hopefully tomorrow…

[Edited on 18-11-2011 by blogfast25]

blogfast25 - 19-11-2011 at 11:16

After filtering and drying (on a hot plate), I obtained a disappointing 6.3 g NdF3. TY should have been about 20, I was hoping for about 10 g and got 6. On the bright side, the product is clearly pink/lanender in tungsten light and switches quite dramatically to this eerie green under TL. With what I’ve made today I should have about 10 g or more, enough for an ‘all in’ attempt.

So with the last magnet (28 g) I revisited the K2SO4/Nd sulphate double salt method, to try and clear a few things up.

Everything was uneventful at first. I set the first filtrate (after precipitation of the Nd double salt) but before any washing, aside as ‘Filtrate A’, then washed with cold, saturated K2SO4 and obtained this ‘pretty in pink’: the NdK double sulphate (hydrate):

But it was during washing I stumbled on a mistake I made in the previous runs and which explains I found a lot of iron in the washed double sulphate: without controlling the pH in the washing solution iron becomes oxidised to Fe3+ and at the higher pH (4 - 5 or above) starts dropping out. That then get caught in the filter cake and there’s your contamination! Remedy: simply acidify the washing solution of sat. K2SO4 slightly.

The liquid obtained from washing was labelled ‘Filtrate B’.

I then proceeded to test both A and B for presence of Nd with oxalate. Obviously this can’t be done in the presence of ferrous material because FeOx is poorly soluble (and there’s a lot of Fe). Therefore 50 ml of both solutions were oxidised on ice bath with 9 % H2O2, simmered to destroy any excess H2O2, then calculated amounts of oxalic acid dehydrate were added. This complexes the Fe3+ to soluble trioxaloferrate (III) (FeOx<sub>3</sub><sup>3-</sup>. This for instance is B after that treatment: the trioxaloferrate complex is emerald green:

What is very interesting is that in neither the cases of A nor B after treatment does any insoluble neodymium oxalate precipitate, none whatsoever, indicating the separation is complete.

The double sulphate was then converted to Nd(OH)3 with KOH, filtered and washed and dissolved in 10 ml of glacial acetic acid and 50 ml water, chosen because my HCl contains traces of Fe3+. The hydroxide dissolved well in that solvent but needed a little heat. And that caused precipitation! First I thought Fe(OH)3 but it’s whitish. Remnants of Nd sulphate? A hydrolysis product of Nd acetate? To be figured out tomorrow…

Wizzard - 21-11-2011 at 12:34

Here's a video I made of the last round of separation by evaporation of the ferrous sulfate form the neodymium sulfate.

http://youtu.be/UoWkqdHnEQc

Play it at high resolution, and if you look close at the lighted side, you'll see the crystals form and grow before being covered in the green iron sulfate.

This is how I separate the two, but this sample was not run hot enough, and was not stopped once the iron sulfate crust forms- Simply stopping evaporation at a certain point and filtering out the precipitated crystals of neodymium sulfate from the solution of (almost pure) ferrous sulfate is my process. This video was more a time-lapse demo- I have rehydrated the mess, and I am already recrystallizing.

[Edited on 11-22-2011 by Wizzard]

Wizzard - 22-11-2011 at 10:50

A slightly better video:
http://youtu.be/hY51xRO7SjA

Anybody else still working on this? I've accumulated almost 10g of Nd2(SO4)3

blogfast25 - 22-11-2011 at 14:05

 Quote: Originally posted by Wizzard A slightly better video: http://youtu.be/hY51xRO7SjA Anybody else still working on this? I've accumulated almost 10g of Nd2(SO4)3

Oh yes. Now drying another batch of hopefully 10 g of NdF3.

Endimion17 - 22-11-2011 at 15:17

 Quote: Originally posted by blogfast25 Nope. Failure is by no means guaranteed w/o argon. See aluminothermic reactions of which I carried out more than a dozen on different metal oxides (Fe, Cu, Si, Ti, Cr, Mn, V, Nb and combined ones as well): despite end temperatures of around 3000 K, you usually get good quality metal (if you get everything right! - they’re not as easy as you might think) because the once molten slag (alumina) forms a perfect protective barrier for the metal regulus. No need for argon here neither, assuming you can get the MgF2 to actually melt (1263 C, not even that hard).

That's true for larger batches, but the amounts we're dealing with here are laughably small. If it was measured in kilos and heated in cylinders (like uranium was initially made) in furnaces, it could end up well because layers would form, assisted by the weight of the mixture.

Things I see here, though very interesting and commendable, produce dispersed, impure metals poorly embedded in the crust, prone to oxidation unless argon is applied... Not that there's anything wrong with it ...

Wizzard - 22-11-2011 at 17:16

Neat thing I tried- I reduced the Nd2(SO4)3 with an amount of KI of equal molar, in about 1cc of water, and I believe K2(SO4) (clear, soluble), NdI2(green) and Nd2O3 (pale blue purple precipitate, turns blue under flourescent light) was produced by the colors of the precipitates and solution.

Evaporating now I believe all of the Nd2(SO4)3 was consumed.

blogfast25 - 23-11-2011 at 06:03

Endi:

I doubt if you've run many a thermite reaction: I've done reactions with as little as 20 g of batch size and obtained metal of very good quality. Bigger is definitely better and gives higher yield but kilos just aren't necessary at all.

AFAIK, nuclear grade uranium metal is still produced by UF4 + 2 Mg. And yes, for optimum results vacuuming the reactor and back filling with low pressure argon is recommended. But it's not essential.

The REAL problem here is not argon/no argon but the low reaction enthalpy. I had wrongly banked on about (-) 300 to 400 kJ/mol.

kmno4 - 24-11-2011 at 10:13

To make it clear, for reaction:
3Mg + 2NdF3 -> 3MgF2 + 2Nd
enthalpy = +14 kJ
gibbs = -7 kJ
You can forget about thermite-like reaction ( but the reaction can give Nd-Mg alloy)
Values taken from experimetal papers (by DOI: 10.1021/j100792a041 and 10.1063/1.438183), not from somebody's assumptions or "Tungsten Gamma" similar sites.

[Edited on 24-11-2011 by kmno4]

blogfast25 - 24-11-2011 at 10:58

 Quote: Originally posted by kmno4 To make it clear, for reaction: 3Mg + 2NdF3 -> 3MgF2 + 2Nd enthaly = +14 kJ gibbs = -7 kJ You can forget about thermite-like reaction ( but the reaction can give Nd-Mg alloy)

Which is what I've been saying: I computed a reaction enthalpy of - 29 kJ/mol Nd.

[Edited on 24-11-2011 by blogfast25]

MrHomeScientist - 8-12-2011 at 20:23

Just a quick update, I haven't forgotten about this! It's getting colder here and I tend to go into hibernation, so posting's probably going to slow down a bit.

I built a crude furnace out of fire bricks and a propane torch that I hope will be able to hold the heat in and get me to the temperature required for this reaction. The lid is on in the picture, but of course below that is the rectangular heating chamber. I drilled a slanted hole in the bottom of the chamber that the torch nozzle sticks through to point directly at the crucible.

I'll try melting aluminum with it this weekend and see if it does anything for me. Nothing's mortared together or anything, so I'll certainly lose a good bit of heat through the cracks but we'll see how it does.

blogfast25 - 9-12-2011 at 05:49

Testing with aluminium is a good idea.
Poppy - 11-12-2011 at 15:13

Brand news:
The neodimium sulfate obteined from the sulphate method was relatively impure because of its slight yellowish color but not purple ~ dissolving it in water was improductive as it would take a week stirring. So to check for coloration differences a little hydrochloric acid was poured just to see if it would dissolve the sulphate and reveal its propoerties under flashtube and tungsten lamps: it did. The solution is yellow green under FL light and olive brown under tungsten lamp.
The interesting part is: different eyes are differently opened for different views. While showing the substance to my grandma, she noticed the precipitate (because just a little HCl was added to the Nd sulphate contaminated salt) was from white (under FL light) to purple (under Tungsten lamp light). The changed runned unnoticed to me until she told, so I must assume her eyes were much better prepared than mine..
As a consequence I supposed the following reactions took place unpredictedly:
Nd2(SO4)3 + HCl --> NdCl3
As a result the predicted Nd/ Fe double salt contaminants reacted with HCl but, slowly, the Nd went back along the sulphate anions and re-precipitated, while all the iron seemed to stay in solution in the form of iron II chloride together with some Nd chloride.
Very luckly, this way a purer Nd2(SO4) was believely obteined. I recall it luck because the dissolution of insoluble salts with acids is not supposed to be obvious, but follows some pH rules to happen, so I was lucky to add enough HCl into it to happen the so said reactions.
The complete analisys of the accomplishments can be described with analitical chemistry formulae. I was studying it recently so I dispose myself to publish it here as soon as possible, along with pictures of the product I have.

Till next c

LanthanumK - 12-12-2011 at 09:39

Upon dissolution of a neodymium magnet in hydrochloric acid, precipitation with sodium bicarbonate, oxidation with hydrogen peroxide, and redissolving the precipitate, a blood red solution is formed. Is this a coordination complex with iron?
blogfast25 - 12-12-2011 at 10:09

 Quote: Originally posted by LanthanumK Upon dissolution of a neodymium magnet in hydrochloric acid, precipitation with sodium bicarbonate, oxidation with hydrogen peroxide, and redissolving the precipitate, a blood red solution is formed. Is this a coordination complex with iron?

Not really. It's simply concentrated Fe<sup>+3</sup> which tends to hydrolyse to FeOH<sup>2+</sup> (which gives it the colour) in high concentrations. Add a lot of oxalic acid to that solution, heat and the iron will complex to FeOx<sub>3</sub><sup>3-</sup> which is green and soluble and the neodymium will precipitate out as the insoluble oxalate. That's one separation method...

Wizzard - 12-12-2011 at 12:17

Noticed while processing my second batch- The nickel coating has much copper in it. I reiterate my suggestion to remove it as fully as possible.

My 60g batch of magnets (latest) had only about 1/2 the coating or less removed, and the magnet sulfate soup is a different shade- I hope the copper sulfate doesn't become a problem :C

blogfast25 - 13-12-2011 at 05:41

Copper isn't very soluble in H2SO4 or HCl without an oxidiser present. Small amounts of CuSO4 or CuCl2 should remain in the mother liquor, as both salts are highly soluble.

Judging a solution of coloured salts is very subjective: concentration is a big factor.

[Edited on 13-12-2011 by blogfast25]

Wizzard - 23-12-2011 at 08:35

Grew out my second batch- 60g of magnets, 500mL water and 100mL 98% sulfuric yeilds about 24g of neodymium sulfate, and an unknown amount of waste material!

I have extracted 18.5 grams into LARGE crystals, the largest of which I have photographed here. I have about 4 g of material being reprocessed and purified. They are a very, very beautiful rose red when large and transparent (in sunlight).

Notes:
1. In the initial purification by crystallization, small Nd sulfate crystals grew, which were then overtaken by pink crystals, which did not change color- I do not know what they were. I think I saved some, but the rest went to reprocessing.

2. While the large batch was purifying (slow, high-temp evap), there was a growth of very light, flakey crystals, like Mica, when gathered, looked silver/white. I have saved a small vial of these crystals, what did not fly away when I gathered them from the top and sides of the vessel, outside of the water line (where water may have collected as steam and evaporated).

3. Also noticed while evaporating, the mixture took a dark brown appearance, but the Nd sulfate crystals grew just the same.

4. Crystal growth over 5 days stopped on the 5th- At this point, the solution was nearly devoid of the Nd sulfate, but was still not oversaturated with Fe sulfate.

I've also made Nd iodide, if anybody is interested Only a very small amount, but the shift in color of the crystals from green to red (in different lighting) is quite lovely.

1cm grid in photo- Small bagged crystals are the largest from the previous batch. All crystals seem to split on their own- Unknown cause- maybe they are naturally twinned?

MrHomeScientist - 23-12-2011 at 09:55

Wizzard,

Nice crystals! I'm currently recrystallizing some of my Nd-sulfate as well, and its forming just as yours look. I have also noticed the color change in the solution when hot - when heated to near boiling, my 'magnet soup' became much darker, nearly black, and lightened again after cooling.

I'm curious - what separation method did you use to get your Nd-sulfate from your magnet solution? A number of ways have been discussed. Sounds like you heated it below boiling to speed evaporation, but if you have any more details I'd be interested. I'm still searching for the easiest way to separate good purity Nd-sulfate from the rest of the magnet crud.

Wizzard - 23-12-2011 at 10:47

To extract like I do, I first take the soup, add some hot water, heat to 'warm' (so as little as possible solubles are left out of solution), and filter with a fine filter paper.

Then take as much Fe sulfate out as possible- After first heating, seal the vessel (not airtight, just enough to little vapor travels in r out) and stick it in the freezer- LARGE crystals of Fe sulfate will grow, and then pour out the rest, add a bit of water, filter, and start the next step.

This filtered liquid is then heated to about 90-95*C, and a thick paper towel put over the top- The heat of the liquid pushes Fe sulfate solubility way up, and Nd sulfate way down- As the liquid SLOWLY loses water through the paper, the Nd sulfate is all pushed out of solution as evaporation occurs, and the Fe sulfate is left in, not quite saturated

blogfast25 - 23-12-2011 at 13:00

The darkening of the solution on heating and lightening upon cooling is very typical of Fe<sup>3+</sup>, which hydrolyses to FeOH<sup>2+</sup> (simply put) and that is strongly favoured by temperature: ferric solutions are thermochromic. Now I know that the Fe is supposed to be there as ferrous sulphate but it's possible you've oxidised it (with air) to ferric sulphate which at low pH (definitely less than 3, depending on concentration) will stay in solution (it's very soluble, as long as the pH is really low).

Without calculation, 500 ml water + 100 ml conc. H2SO4 would sound like an excess of acid and the Fe concentration would be roughly 2 M, based on 60 g of magnet. So that kind of fit the bill. Of course concentrated solutions of Fe<sup>2+</sup> are quite dark too but not thermochromic...

Very nice crystals, wizz!

Wizzard - 24-12-2011 at 18:06

Interesting note I made today- Nd2(SO4)3 is paramagnetic, just as ferrous sulfate is!
Poppy - 25-12-2011 at 14:15

Might bother telling us how did you pick the measure of ferrous sulphate paramagnetism? Can ferrous sulphate just be attracted by a magnet?
Wizzard - 25-12-2011 at 14:18

No measure, but holding a large Neo magnet up to the crystal causes medium crystals (5mm or so) to stick- smaller ones aren't strong enough, larger ones too massive. You can also put a magnet to the crystals when in a bag- the bag will sway towards the magnet, strongly sometimes!
Poppy - 14-1-2012 at 13:34

As pomissed I've reached expertise in metalic ion separation via pH control. I've purchased a pH meter which will fit best for the purpose of determining the correct scales involved. I just need to resuply my chemicals this month, becuse a car crash took my mney away last month :p . Data shall be uploaded very soon. C you guys

Wizzard: Could you please post pictures of your neodimium iodide, it should seem pretty!

Poppy - 8-2-2012 at 19:34

Okay here follows the calculations involving the required pH to operate these chemical purification processes.
From a source entitled "The Thermodynamic Properties of Neodymium Hydroxide Nd(OH)3, in Acid, Neutral and Alkaline Solutions at 25°; the Hydrolysis of the Neodymium and Praseodymium Ions, Nd3+, Pr3+" the following values goes for neodimium hydroxide:
solubility: 4,8 x 10^(-3) mol per 1000g water.
Nd(OH3) (s) <--> (Nd3+) (aq) + 3(OH-) (aq) K = (8,7 +/- 4,47)x10^(-21)
Let us use the average 8,7x10^(-24) value for K, just for comodity.
The value of the K constant agrees well with the formula invonving the solubility of Nd(OH)3:

Nd(OH3) (s) <--> (Nd3+) (aq) + 3(OH-) (aq)

K = [Nd3+].[OH-]^3
as for [OH-], there will be a kind of common ion equilibrum with water molecules: Kw = [H+].[OH-]; Kw/[H+] = [OH-]
Thus follows that
K = [Nd3+].(Kw/[H+] )^3
K = [Nd3+].(Kw^3).[H+]^(-3)
considering the solubility of the compound is so low it wont move the kW value away from 1,0x10^(-14), and also considering the concentration of (H3O+) cations will be compensated by the acid behaviour of the Nd(OH2)6 aquacomplex formed by the dissolution of the compound, we can state for the sake of aproximation that [H+] = 1,0x10^(-7) too, the equation becomes:
K = 4,8x10^(-3) . 1,0x10^(-42) . 1,0x10^21 = 4,8x10^-24, which is close to the result the high grade guys obtained. They must have complicated a bit more this equation but I don't know how.
They also give another number if you guys find it useful:
Nd3+ (aq) + Nd(OH)3 (s) + (H+) (aq) --> (Nd3+) (aq) + 3 H2O (l)
K = 8.7 x 10^13 (thats about it because the article is a hard-to-visualise free version)
I had an doub't if K was in the order of 10^(-21) or 10^(-24) because of the bad visualisation allowed for a free article, but that equations above provided insurance.

So, considering the value of K, neodymium hydroxide, as well as iron III hydroxide, phydrolises too easily in solution depending on pH, it must be adjusted to prevent that. The pH at which neodymium hydroxide starts precipitating out of the solution can be calculated by switching [Nd3+] by some hypothesized value of Nd3+ concentration.
8,7x10^(-24) = [Nd3+].[OH-]^3, you find [OH-] and calculate kW=[H+][OH-], then use -log [H+] to know the pH maximum pH the solution of your Nd3+ ions can go before hydrolising and falling out of the solution, later being mistaken as "salts".
The K value for the dissolution of iron II hydroxide:
K = 2,79x10^(-39) = [Fe3+][OH-]^3 giving a somewhat lower pH of precipitation to start.
The different pH at which iron and neodymium precipitate could be used to separate the two via a pH separation, but for a series of reasons I wont do the calculations x)
So, first of all, when dissolving the magnets, keep in mind the solution must be very acidic, or otherwise very dilute for the final products, remember oxydising the Fe II into Fe III to evade problems and if you trying the sulfate separation method keep the pH high enough just to prevent hydrolisys. Nd(OH)3(s) will not add to crystal formation or color enhancing of the solution, because they neither get into solution.

Sorry for taking this long, :p
Now I gotta recycle my shitty Neodimium powder to get rid of the double salts and hydroxides left mistaken by Nd3(SO4)2 and do it all over again.
SO i hope the crystals so formed will get pretty much like wizard's ones

Muahahah!

Attachment: arquivo ruim de ler.htm (93kB)

[Edited on 2-9-2012 by Poppy]

blogfast25 - 9-2-2012 at 07:23

Poppy:

I think differential (by pH) precipitation of Nd and Fe hydroxides will be very difficult to do. Concentration effects need also be taken into account, although the high molar ratio Fe/Nd = 7 would be an advantage here.

In an accidental sense I've already tried this, see above, w/o success. Filtering problems with large amounts of Fe(OH)3 (which peptises easily) would also need to be taken into account. Co-precipitation problems might also show up.

This is why the use of poor solubility of hot Nd sulphate or cold (Nd, K) double sulphates remains preferred industrially, or so I believe.

[Edited on 9-2-2012 by blogfast25]

Poppy - 9-2-2012 at 16:12

Agreed, see it would be quite difficult.

The sulfate was recycled back into hydroxide form, dryed in an oven, weighed and put to react directly with 98% sulfuric acid. The reaction proceeded very well, releasing heat and with the formation of a strong pink solution. the yield was 28g. This was then dissolved in a 1L beaker filled with 530mL water and sulfuric acid at pH 1.7. Initially there was added twice the necessary ammount of anhydrous sulfuric acid to dissolve the neodymium/ iron hydroxide because a gel was forming which turned to complicate further dissolving. So after dilution the pH was even lower than 1.7: that to prevent iron III species to fall as a precipitate. The 530mL were measured out to dilute the 28g of sulfate at 30ºC, then heating to 100ºC would bring 20g of the neodymium sulfate dissolved to come as crystals. But they didn't. Maybe the very excessive sulfuric acid is playin' a role on holding the solublity of the Nd sulfate up.
The solution at the moment is slowly evaporating and hopefully will give crystals.
How many surprises are we yet to see!

Wizzard - 10-2-2012 at 05:36

@Poppy- I've found the heating and evaporating the solution will not separate the two sulfates at any temperature if the mixture is too acidic.
blogfast25 - 10-2-2012 at 07:31

 Quote: Originally posted by Wizzard @Poppy- I've found the heating and evaporating the solution will not separate the two sulfates at any temperature if the mixture is too acidic.

Explain a bit more, if you please?

Wizzard - 10-2-2012 at 07:51

@Blogfast- Certainly! During evaporation, trying to recover the last bit of Nd sulfate from solution, what I've found is left after evaporation is a hard, brown/green/white amorphic mass at the bottom of the vessel, with or without unevaporated liquid solution.

I found this solution is nearly always VERY acidic- I would also imagine the solution it precipitated from is as well. When rehydrated, the mixture makes a sparkley liquid, and the mass is difficult to get into solution, it takes much stirring and heat. The sparkley stuff (white/yellow, very fine) is unknown, but I have isolated it (it comes off in the steam during any evaporation phase, and collects and crystallizes white, in solution it develops a yellow tinge, likely due to iron oxide on contact with the water). The material then still will not crystallize and separate Nd from Fe sulfate, it will only dry out and reform the same very hard mass.

Adding water to the mass, heavy stirring, some heat and a minimal of extra magnet material (I will not add something else to the mixture to only neutralize the acid) results in a solution which looks much more like the original solution. Of course, now filter and clean the soltion of junk and insoluables.

This solution will now act mostly normal- It will cool (from 80*C to 0*C) and crystallize large iron sulfate crystals, and upon removal of those (or leave them in, it doesn't matter), when heated to 90*C and allowed to evaporate, will result in crystals of Nd sulfate. The material which crystallizes is much greater than the mass of magnet material added to neutralize the acid- This hints that the problem (I think) is something about the pH of the solution.

blogfast25 - 10-2-2012 at 08:03

Wizzkid:

Yes, I’ve seen your ‘sparkley stuff’ too, when working with quite concentrated iron solutions. It’s a bit like metallic paint, right? These must be hydroxy ferric compounds, possibly Fe(OH)SO4 or similar. By that time your ferrous sulphate will have been oxidised mostly to Fe (III) because of heat and air exposure.

In my case sometimes I got the stuff to dissolve again, sometimes not. The sparkly effect must be due to shiny, platelet-like crystals, acting like mirrors, like in a flaky aluminium powder-resin mixture (metallic paint).

[Edited on 10-2-2012 by blogfast25]

Poppy - 10-2-2012 at 16:56

You guys mean there should be kept about a half ammount of the hydroxides to replace them to the final solution thus neutralizing it?
What about just throwing ammonia in there? anyone ever tried this? Will the Nd still behave the same manner when heated, i.e., precipitating?
Has anyone tried this? I'll if none did.

Wizzard - 10-2-2012 at 21:05

@Blogfast- Yes, like paint I have a small vial of the material, perhaps I should test it's composition... But I would not know where to start.
blogfast25 - 11-2-2012 at 06:01

Wizzard:

You'd have to isolate the 'shiny' material first and I bet it filters very badly.

Poppy:

Ammonia is great for gradually adjusting pH. Not sure what you mean by the rest.

Wizzard - 11-2-2012 at 06:28

@Blogfast- It doesn't need to be filtered, it can be isolated in the evaporation condensate on the sides and top of the vessel, free from water. I'm very sure it's 100@ 'shiny' Picture in a bit.
Poppy - 11-2-2012 at 11:38

Blogfast: I mean if Nd-ammonium double salts would precipitate the same way neodimmy sulfate does.

Nvm, ammonia has just been added into the solution, bringing the pH from 1.3 to 1.8-1.9 with a pH meter. Ionic strenghts seems to be playing a role there so the calculations useing this information are kinda complicated for me.

For all of you guys, is it just me but neodymium compounds has like a somewhat sweet smell? That be a wonder of nature, of course!!

The solution was set to evaporate at room temp and the salts remained the same color.

Wizzard - 28-2-2012 at 10:44

New crystal pictures- Evaporative growth over 2 weeks (in 2 stages to remove iron) lead to these crystals from the smaller, 'scrap' Nd2(204)3 stock I have prepared.

The largest single crystal at the bottom, double-terminated (but flat on one side) is 1.8cm in length!

All large crystals of this Nd sulfate show the same fracturing seen in previous large crystals- I do not know what causes this, but the previous largest crystal (I would guess 2.2cm!!) broke on removal from the growing glass.

And I can get some pics of the Nd iodide soon, also- I'll be making another batch! The oxide is also a by-product, which is just as pretty.

blogfast25 - 28-2-2012 at 14:09

Very nice crystals, as per usual.

How did you synth. the NdI3?

Poppy - 28-2-2012 at 19:50

the ammoniacal Nd solution I`ve spoken about is as yet unchanged. I'll keep this until finding the proper evaporatio carryin devic es Crystal trips via boiling coprecipitates Fe2O3 (or maybe we talking about lanthanide impurities?)  mainly in the secondary preci pitation steps. ~ Very nice thx or the images
Wizzard - 28-2-2012 at 21:29

Poppy, can you describe your ammon. Nd solution? I'm building a beautiful color-changing display for my RE salts
Poppy - 5-3-2012 at 11:59

Wizz,

The ammoniacal solution was made this way:
1 mol Neodimmy hydroxide was mixed with 2 mol conc. sulfuric acid. The solution was afterward neutralised with 1 mol ammonia which brought pH to 4 - 5. The thick syrup did evaporate a little bit and the crystals formed changes from lavander through pink depending on lighting. It followed that the once precipitated ironIII hydroxides redissolved as the concentration raised in the solution, leaving a brownish supernatant liquid over the crystals.
=P
the ammonia helps the crystal growing fat
Oh sorry to the other chemicals water was added to makeup 530mL, with the initially poured 20g Nd(OH)3

[Edited on 3-5-2012 by Poppy]

[Edited on 3-5-2012 by Poppy]

Poppy - 5-3-2012 at 17:02

Considering the mass of the supposedly double salt thus obtained I can't tell if ammonia entered the crystal at any degree. If the mass surpasses the expected ratio then ammonia definitely played its role. The very pure Nd salt seems relatively more soluble now though.
Poppy - 7-3-2012 at 16:59

Photos of 99.9999%-iron free Nd sulfate. The ammoniacal salts remained in the crystalization liquor.
Neodimium sulfate under large fluorescent tubes:

" under small fluorescent tubes:

"under tungsten lamp lighting:

[Edited on 3-8-2012 by Poppy]

[Edited on 3-8-2012 by Poppy]

blogfast25 - 8-3-2012 at 06:35

 Quote: Originally posted by Poppy Photos of 99.9999%-iron free Nd sulfate. The ammoniacal salts remained in the crystalization liquor.

Poppy, can you explain again (briefly) where the ammonia comes into things?

Also, how can you substantiate your '99.9999 %' claim? On what grounds?

Poppy - 8-3-2012 at 06:45

Blogfast,

Ammonia comes into nothing, apparenttly, it was used to neutralise a fraction of the excess sulfuric acid

The purity is a relative 99.9999% iron free (which means no iron at all), not 99.9999% pure Nd sulfate. The acidic washing bath with sulfuric acid should have reached this, or maybe its just wrong?

I reallty guess it was too straight an assumption, but considering that by througly sinkings the products allowed for a high quality trustworthy final product.

99.9999% was simply too much way out the strathosphere.

blogfast25 - 8-3-2012 at 07:20

Poppy:

Ah. Test your product with KSCN or NH4SCN by leaching some of it with cold water, add a bit of H2O2 to the leachate (to oxidise any to Fe (III)). Fe3+ tests positive for red FeSCN2+ if any iron present. It's a fairly sensitive test.

Your product looks very neat though...

[Edited on 8-3-2012 by blogfast25]

Poppy - 8-3-2012 at 09:36

I'm gonna bring this to my university, they got KSCN there, buying this is not likely affordable in normal circunstances.
elementcollector1 - 12-3-2012 at 19:38

I've been reading through this thread, and I'm confused as to the oxalate route. I recently acquired some "Davy's Oxalic Acid Crystals" which I assume are pure, and added this to a gray (not purple, for some reason) solution of 'magnet chloride'. I now have a green supernatant liquid, and a greenish-gray precipitate. What did I do wrong, and how can I fix this?
blogfast25 - 13-3-2012 at 06:03

EC1:

As I’m not sure as to your exact procedure it’s a little difficult to comment. I’ll just say this.

Your ‘magnet chloride’ should be emerald green, of FeCl2.

The method relies on the fact that Nd oxalate is very insoluble but of iron exists a ferric trisoxalato complex - FeOx<sub>3</sub><sup>3-</sup> - bright green, slightly fluorescent and very soluble. To convert all iron to the complex, firstly the Fe (II) needs to be oxidised with H2O2 to Fe (III). Do this by adding the required amount of ice cold peroxide to iced magnet chloride, slowly!

At that point, iron (III) (hydr)oxide probably started precipitating (or it may not, depending on pH of the solution). Now add a calculated excess of oxalic acid [note: you need to account for the neodymium oxalate too!], dissolved in warm water, slowly while slowly heating the solution until the brown Fe(OH)3 dissolved to green FeOx<sub>3</sub><sup>3-</sup> (this can take a while to be fully achieved). The off-white precipitate formed is Nd<sub>2</sub>Ox<sub>3</sub>. Hot filter to avoid excess oxalic acid to crystallise out and confuse you. Wash Nd oxalate copiously…

Careful: oxalic acid is toxic! Wear suitable gloves and goggles at all times!

All in all the straight 'sulphuric acid method' is probably simpler but the one time I tried it, I couldn't get the magnet to dissolve easily because the dissolution 'stalled' and I had to resort to HCl to complete it... Others haven't had that problem.

[Edited on 13-3-2012 by blogfast25]

[Edited on 13-3-2012 by blogfast25]

elementcollector1 - 13-3-2012 at 14:59

Just filtered my solution, and you're right, it is emerald green. Why were my previous solutions purple?
Anyway, from what I'm inferring from your post, upon slow addition of chilled peroxide to chilled magnet chloride, the iron hydroxide should start precipitating out. What happens to its chloride ions? Then, add mucho oxalic acid in (what is warm, 50 degrees? near boiling?) until no more precipitate forms, and do this while heating the other mix (this will be tricky on the home stove, and even trickier on the camp burner). Correct?

blogfast25 - 13-3-2012 at 17:46

Due to suspended particulate matter, unfiltered solutions may appear very different from filtered ones. Don't worry about it.

Add the warm (50 to 100 C is OK) oxalic acid solution slowly while heating the oxidised magnet soup to simmering. After some time the red-brown slurry-solution will clear to green (the precipitated Nd oxalate tends to accumulate at the bottom). Careful doing this on the stove: a dedicated second hand domestic elecrtical hot plate is much to be preferred because of the potential for boil overs and the toxicity of OA!

The chloride ends up basically as dissolved HCl.

[Edited on 14-3-2012 by blogfast25]

elementcollector1 - 13-3-2012 at 20:06

I'll see if I can find or borrow a hot plate.
So, the oxalate precipitate is purple? That's what I got online, anyway. My first precipitate, which I assume is a mix of neodymium and iron (II) oxalates, was a grayish-green.

blogfast25 - 14-3-2012 at 05:37

Neodymium oxalate is off-white, probably with a purple tinge, under incandescent light. But speaking of its colour is a bit meaningless: remember that Nd salts look decidedly different under incandescent light or TL light (or saver bulbs) because traces of UV cause the salts to fluoresce greenish! It's one simple way of recognising Nd salts...

Let us know how you get on.

[Edited on 14-3-2012 by blogfast25]

elementcollector1 - 1-4-2012 at 20:41

I did the steps slightly out of order: I added the H2O2, and then chilled it in a freezer. An hour later, nothing had changed, no precipitate of anything. The current color of the solution is yellow, different from the emerald green. Should I add more H2O2? I added a whole ton already.
Pages:  1  2    4  ..  6