Sciencemadness Discussion Board - The trouble with neodymium...

The trouble with neodymium...

Pages: 1 2 3 4 5 6

blogfast25 - 31-1-2014 at 06:28

Quote: Originally posted by MrHomeScientist

"organic sulfate of potash"

Nice euphemism

'Mineral sulphate of potash' would have been more accurate.

Brain&Force - 31-1-2014 at 18:47

I made some magnet nitrate today, dissolving a few small magnets in 6M nitric acid. They were unusually slow to dissolve and no visible amount of NO₂ was produced. They were not demagnetized and only partially stripped of their coating, so there is Ni(NO₃)₂ in solution as well. I added some potassium sulfate and heated the solution to encourage precipitation, but nothing came out. As with the terbium I've decided to wait for next week in case supersaturation occurred.

There were also white flakes floating in solution that eventually dissolved...I wonder what that was.

MrHomeScientist - 31-1-2014 at 18:56

White flakes could possibly be boric acid from the boron in the magnets. Somewhere in the thread I reported finding some white flakes after filtering, which tested (not conclusively) positive for boron. Yours might not have precipitated because there was not enough sulfate in solution, possibly. In my case at least I used sulfuric acid so there was plenty available for the double salt. Did you add enough sulfate calculated to saturate the soluton?

Brain&Force - 31-1-2014 at 20:07

I think I didn't use enough magnets (this was just a test with two tiny magnet fragments, about .5 gram). I'll try boiling the solution down further, or dissolving more magnets.

[Edited on 1-2-2014 by Brain&Force]

blogfast25 - 1-2-2014 at 06:10

Why waste precious nitric when you've got 6 M HCl (as I recall)? To dissolve also the Ni cover?

I always demag first (put them on a hot plate on full), then crack them with a hammer. Then 12 M HCl. When it's more or less done, I filter off the residue and subject it once more to a small amount of 12 M HCl because there's usually a bit of magnet metal lurking in the corners of the coating.

Then calculate the amount of K2SO4 to saturate the volume, add it, simmer for a bit and allow to cool to RT and the Nd precipitates out as the double sulphate. Filter and wash with pH < 3 wash water.

Because the double sulphate is not the end product there's no point in washing it further (after all Fe has been removed). I treat it with strong ammonia immediately, simmer for a bit, hot filter and wash the Nd(OH)3 copiously with water.

Brain&Force - 1-2-2014 at 08:28

I just ran out of HCl, but I still have plenty of nitric. The magnets only dissolve with difficulty, though, and NO₂ is not generated. One of the problems might be that the magnets have been broken for a long time and have corroded.

[Edited on 1-2-2014 by Brain&Force]

blogfast25 - 1-2-2014 at 13:07

If you do get to Nd(OH)3, make sure you check out its fluorescence too. It really looks eerily different under an incandescent bulb as under a saver bulb.

Brain&Force - 1-2-2014 at 20:13

Nd only fluoresces in the IR range. The effect is caused by absorption bands not lining up with the emission bands of light sources.

blogfast25 - 2-2-2014 at 06:39

Huh? See these pictures of NdCl3.6H2O under sunlight and 'black light':

http://en.wikipedia.org/wiki/Neodymium(III)_chloride

I've seen this myself even though my NdCl3 was far from pure and I only used a UV counterfeit detector. ndF3 does it too: blueish (sunlight) to greenish (UV rich light).

Brain&Force - 2-2-2014 at 10:45

It's not black light, it's a fluorescent energy saving bulb. The color change is caused by narrow absorption bands caused by forbidden f-f transitions.

See http://woelen.homescience.net/science/chem/exps/neodymium/in...

If it really is fluorescent, I would love to see photos.

The wiki explanation of the effect is dead wrong, a charge transfer only occurs in melts of NdCl₃.

blogfast25 - 2-2-2014 at 11:29

Quote: Originally posted by Brain&Force

It's not black light, it's a fluorescent energy saving bulb. The color change is caused by narrow absorption bands caused by forbidden f-f transitions.

See http://woelen.homescience.net/science/chem/exps/neodymium/in...

You are right, B%F. The colour change of Nd(OH)3 and NdF3 in incandescent v. saver bulb light is so striking I always believed it was due to UV fluorescence.

Brain&Force - 2-2-2014 at 12:02

Here's a video showing the same effect in holmium triflate: https://www.youtube.com/watch?v=GEK90hf49Jk

Is Nd fluoride supposed to be green or was there Pr contamination in your sample?

blogfast25 - 2-2-2014 at 12:31

Quote: Originally posted by Brain&Force

Is Nd fluoride supposed to be green or was there Pr contamination in your sample?

It's supposed to be blueish purple, which it was. It was magnet Nd, so Pr cannot be excluded but I doubt if there was any there.

Under saver bulb it looks greenish.

MrHS has some photos of his NdF3 higher up in this very thread, IIRW...

MrHomeScientist - 2-2-2014 at 12:59

Yes my first batch of NdF₃ was green under small fluorescent tube lights and pink in daylight. My second batch, where I was more careful about purity, was a clean white instead of green. The photos are in this thread somewhere.

Brain&Force - 5-2-2014 at 20:17

I evaporated the magnet nitrate to dryness, but I fear I have caused everything to hydrolyze.

I added some more nitric acid to the hydrolyzed product. At the same time another, larger magnet fragment was added to the mix. This time, the magnet dissolved in a huge puff of NO₂ and no trace of it remained after 2 minutes. I still need to add more nitric acid to the mix to redissolve the rest of the hydrolyzed product.

blogfast25 - 6-2-2014 at 10:01

Quote: Originally posted by Brain&Force

I evaporated the magnet nitrate to dryness, but I fear I have caused everything to hydrolyze.

Not sure why you wanted to do that?

Brain&Force - 6-2-2014 at 10:31

Actually, I don't know either.

I was attempting to reduce the volume of the solution, and for some reason I just let it evaporate.

blogfast25 - 6-2-2014 at 10:34

It should be fully recoverable with strong acid. All Fe will now probably be as Fe(III) so even closer control of pH will be needed to remove all of it, assuming you're going the double sulphates route.

Brain&Force - 7-2-2014 at 10:11

I added an excess of nitric acid but it's still not dissolving! Any tips for making this dissolve faster?
Also, will sodium or ammonium sulfate work to precipitate the Nd? Or is potassium sulfate the best option? I don't have much potassium sulfate.

blogfast25 - 7-2-2014 at 10:24

Quote: Originally posted by Brain&Force

I added an excess of nitric acid but it's still not dissolving! Any tips for making this dissolve faster?
Also, will sodium or ammonium sulfate work to precipitate the Nd? Or is potassium sulfate the best option? I don't have much potassium sulfate.

What acid strength are you using? Prolonged heat may be needed to get everything back into solution.

The Na/REE double sulphates are also quasi-insoluble. I believe also the ammonium ones are, IIRW. To be on the safe side, use Na2SO4, this has been reported to have been used for this type of separation, in several references.

MrHomeScientist - 7-2-2014 at 10:41

I used sodium sulfate in my separation attempt on the previous page, and it appears to work nicely. I seem to recall somewhere else in this thread it was mentioned that the potassium double salt has the lowest solubility of all.

I actually just got some sodium thiocyanate in the mail yesterday, so I'm finally able to test for iron contamination. I've been out with a stomach virus all week, though, so experimentation has been put on hold for a while.

Brain&Force - 7-2-2014 at 12:51

My experience with terbium shows that simple rinsing is all that is needed to remove Fe. Fe appears to quench the fluorescence of terbium potassium sulfate, but simple rinsing is all that is needed to remove the iron and restore fluorescence.

I remember hearing about ammonium sulfate being used to remove rare earths; I'll try to dig up the reference.

blogfast25 - 8-2-2014 at 06:45

Quote: Originally posted by Brain&Force

My experience with terbium shows that simple rinsing is all that is needed to remove Fe.

I remember hearing about ammonium sulfate being used to remove rare earths; I'll try to dig up the reference.

If the metals are present as sulphates/double sulphates and the rinse is of low pH then yes, the ferric sulphate is simply washed out.

MrHomeScientist - 2-5-2014 at 11:37

Not very applicable to home chemistry, but I came across an interesting article today relating to neodymium magnet recycling: http://phys.org/news/2012-12-magnetic-idea-rare-earth-recycl...

Try it if you have the capability to boil magnesium!

Unfortunately no progress on my own endeavor. I plan to finish cleaning up my lab this weekend, then get back to regularly-scheduled chemistry. Apologies if anyone got overly excited by the thread bump!

aga - 6-5-2014 at 09:51

Knowing next to nothing, i thought i'd give Neodymium extraction a go, seeing as i have a few hard drives lying about.

About 125g of head actuator magnets were dissolved in 500ml of OTC 20% HCl solution. This took ages - about 1 week.
Filtering removed black/grey sludge and bits of the magnet jackets (presumably Nickel).
Nothing fell out of the resulting liquid after being left to stand for a few days, and it looks the same colour as weak ferric chloride to me.

Several reagents were added to 5ml samples of the resulting liquid, based on what was handy (see photos).
Left to right :
Liquid with nothing added
+Elemental copper
+Copper sulphate solution
+Sodium Hydroxide (s)
+Magnesium (s)
+Sodium MetaBisulphite.

Interesting and fun to do, but i read that the Nd³⁺ ion is a violet colour, so i thought it all wrong and started cleaning the test tubes.

The Copper Sulphate tube showed Violet 'crystals' while being washed with water ! (before that, the deposit appeared a grey colour).

So, i promptly added copper sulphate solution to the remaining liquid, and got a pale violet precipitate (last photo).
The precipitate is only slightly soluble in water, and the paste is rather like diluted toothpaste in consistency.

Assuming that the black sludge was Boron, which weighed 1.20g, that means about 16g of Nd is present somewhere (if assumptions are correct).

After looking at these links :-
http://en.wikipedia.org/wiki/Neodymium%28III%29_chloride
http://en.wikipedia.org/wiki/Iron%28III%29_sulfate
http://en.wikipedia.org/wiki/Iron%28II%29_sulfate

It's Iron (III) Sulphate isn't it ?

[Edited on 6-5-2014 by aga]

blogfast25 - 6-5-2014 at 11:51

aga:

It appears the addition of cupper sulphate has caused neodymium sulphate octahydrate to precipitate (middle photo). Nd2(SO4)3.8H2O is very poorly soluble in water and does look a bit like that.

Try doing the same using larger sample of 'magnet chloride' soup but using H2SO4 as source of sulphate (NOT CuSO4).

Dissolving Nd magnets in 20 % HCl at RT is slow, especially if you didn't remove all the nickel protective coating. My magnets (coatings partially removed) dissolve in simmering 37 % HCl in about 1 - 2 h...

[Edited on 6-5-2014 by blogfast25]

aga - 6-5-2014 at 12:03

DOH !

No Magnet Chloride left over.

All got added to the copper sulphate solution (got about a litre left of that mixture).

5ml in test tubes again of the Magnet Chloride + Copper Sulphate solution :
Ethanol precipitated *something*
Ammonia made a large portion of brown stuff immediately settle out, plus a well separated layer of something the same colour as the tetraammine complex (this is the top layer).

What Test is applicable to Neodymium Sulphate ?

It looks exactly like the iron (III) sulphate photo to me, texture and all.

[Edited on 6-5-2014 by aga]

blogfast25 - 6-5-2014 at 12:15

Huh?

How did you manage to 'lose' 500 ml magnet soup so quickly?

What do you mean by 'What Test is applicable to Neodymium Sulphate ?'

aga - 6-5-2014 at 12:20

I added about 500ml copper sulphate solution (mol unknown) to the roughly 500ml Magnet Soup = about 1 litre

The Test thing is basically How To Tell What The Pinky/Violet Stuff is.

There's more precipitated out, but i need to dry it.
I'll send you some to look at if you like.
It's a violet powder.

[Edited on 6-5-2014 by aga]

blogfast25 - 6-5-2014 at 12:40

Try this.

Firstly wash it several times with small amounts of hot. dilute (10 % or so) H2SO4. Nd2(SO4)3.8H2O has the peculiar property of being less soluble in hot than cold water. This procedure washes out any remaining iron, which later would be a nuisance factor.

Then treat it with strong ammonia: you'll see the texture change. Allow this some time by keeping it warm. The ammonia treatment converts the sulphate to the hydroxide Nd(OH)3. This insoluble hydroxide can now be filtered and washed plentifully with hot water to remove the ammonia and any sulphates. Look at the wet hydroxide under sunlight and under saver bulb light: it should look slightly different in colour, depending on light source.

Dry it and you've got basically fairly pure Nd(OH)3.

More detailed separation techniques for Nd are plenty, with several experimental examples in this long thread.

aga - 6-5-2014 at 12:44

many thanks blogfast25.
i will give it a go tomorrow.

Töilet Plünger - 6-5-2014 at 12:50

Does anyone have an IR camera? I'd like to confirm if neodymium salts fluoresce in the IR range. And, blogfast, I thought you told me in the terbium thread that the potassium double sulfate is less soluble in water than neodymium sulfate alone. (Though, according to an old reference I found, neodymium also forms an acid sulfate. I'll have to dig it up - it's on Google Books somewhere.)

Happy 5000th post blogfast25!

aga - 6-5-2014 at 12:52

Yay !

Happy 5000 to yooo. Happy 5000 to yoooo

[Edited on 6-5-2014 by aga]

aga - 6-5-2014 at 12:55

Quote:

Does anyone have an IR camera?

You probably do.
Most cameras are IR sensitive anyway, so a Filter is inline with the lens to correct the colour response of even a $5 usb camera.

Hmm. that gives me an idea.

[Edited on 6-5-2014 by aga]

Töilet Plünger - 6-5-2014 at 13:02

Well, it has to be able to collect light at 1064 nanometers - that's one of the transitions visible. (A neodymium yttrium orthovanadate laser uses second harmonic generation to get green light @ 532 nm.) If it can pick up the IR light from a neodymium laser with the KTP frequency doubler removed or the green light selectively filtered, the camera works.

aga - 6-5-2014 at 13:13

Just try it.

Maybe a cheap camera will react to 1064nm with the IR filter removed.

Seems to me that far too much calculation is going on and not enough experimentation, on which ALL known science is based.

Once upon a time some madman distilled a large quantity of ants.
I seriously doubt that he had calculated the result before just doing it.

Time to put some madness back.

Does it Blend ?

blogfast25 - 6-5-2014 at 13:25

Quote: Originally posted by Töilet Plünger

And, blogfast, I thought you told me in the terbium thread that the potassium double sulfate is less soluble in water than neodymium sulfate alone.

It is. But the Nd sulphate is insoluble enough for a crude separation. You'll lose more of it though. I was keeping things simple for aga. The K double sulphate method is my preferred method.

A hydrogen sulphate? Possibly. We know for sure there are two variations of the sulphate, it's been discussed extensively in the various Nd threads. One is pinkish and sandy and doesn't seem to crystallise easily, another is reddish and crystallises into beautiful crystals. Whether one of them is a bisulphate I don't know.

[Edited on 6-5-2014 by blogfast25]

aga - 6-5-2014 at 14:15

Quote:

I was keeping things simple for aga

Very necessary.
I truly know next to nothing.
Thanks again blogfast25 for keeping it simple for a noob to understand.

You are a Guiding Light.

Finding things (out) is a lot easier when there is Light.

[Edited on 6-5-2014 by aga]

Brain&Force - 6-5-2014 at 15:22

I was going to propose the existence of a bisulfate as a possible explanation for the two different varieties in the thread, "two times neodymium sulfate, different colors."

aga - 7-5-2014 at 08:35

All i did was put some hard drive magnets (no pre-processing) into 20% OTC HCl,
waited for the magnets to dissolve, filtered, added copper sulphate solution,
decanted, washed with water (repeat a few times).

When washing, the upper liquid layer has a blue tint, very much like the Copper Sulphate solution, rather than any hint of green.

Result from 125g magets is 10.12g of pale pink talc-like powder.

The nearest wiki photo suggests it's Iron (III) Sulphate,
but it looks identical to the photos in this thread,
which suggests that it's Nd₂(SO₄)₃.8H₂O

Am i missing something here ?

[Edited on 7-5-2014 by aga]

blogfast25 - 7-5-2014 at 08:56

Quote: Originally posted by aga

Am i missing something here ?

[Edited on 7-5-2014 by aga]

Nope. That's neodymium sulphate in one of its two forms. A bit silly to waste CuSO4 as a source of sulphate ions. Instead use sulphuric acid or Na2SO4 (Glauber salt) or K2SO4 or (NH4)2SO4. The latter two can be found in garden centres but need recrystallizing to use as reagents.

Ferric sulphate is so soluble it's hard to crystallise. The wiki photo is probably of the monohydrate.

Object lesson: appearances are deceptive.

aga - 7-5-2014 at 09:06

Quote:

silly to waste CuSO4 as a source of sulphate ions

I can get copper sulphate for 4 euros a kilo, but have not found sulphuric acid yet.

Isn't the fact that on washing the precipitate you can see the blue copper sulphate a useful indicator ?

As in, keep washing until no blue colour seen = copper sulphate mostly gone.

elementcollector1 - 7-5-2014 at 09:09

What about Epsom salts (magnesium sulfate)?

MrHomeScientist - 7-5-2014 at 09:42

Blogfast is right, appearances are deceiving especially in the case of rare earth salts like neodymium sulfate. When describing what it looks like, you have to specify under what lighting conditions because it can look very different depending on the light source. My neodymium sulfate (pictured in this thread somewhere) was white to tan under small fluorescent tube lights, pink under incandescent bulbs, purple under long tube fluorescent lights, and a very richly colored pink under sunlight. A fun past-time is to take a vial of your sulfate and walk from one room to another and then go outside, watching the color change smoothly.

blogfast25 - 7-5-2014 at 10:20

Quote: Originally posted by elementcollector1

What about Epsom salts (magnesium sulfate)?

Yep, why not, indeed?

elementcollector1 - 7-5-2014 at 11:00

In recent news, my neodymium oxide is proving immune to sulfuric acid (diluted with a little bit of water). I wonder if this requires prolonged boiling in order to react?

IrC - 7-5-2014 at 11:19

Quote: Originally posted by MrHomeScientist

Blogfast is right, appearances are deceiving especially in the case of rare earth salts like neodymium sulfate. When describing what it looks like, you have to specify under what lighting conditions because it can look very different depending on the light source. My neodymium sulfate (pictured in this thread somewhere) was white to tan under small fluorescent tube lights, pink under incandescent bulbs, purple under long tube fluorescent lights, and a very richly colored pink under sunlight. A fun past-time is to take a vial of your sulfate and walk from one room to another and then go outside, watching the color change smoothly.

Is this not due to the amount and frequency of uV light in the source? The 'appearance' of the material ends up being a combination of both the reflected light and the light being produced by uV induced fluorescence in the specimen?

aga - 7-5-2014 at 11:30

OK. Point gratefully received.

It's going dark here, so i'll trog down to the chem shed to see what it looks like under florescent light, LED light and a halogen light.

Under tri-colour LED light it appears pale violet
White Florescent light makes it grey/pink
Halogen light makes it appear violet/pink again.

[Edited on 7-5-2014 by aga]

Brain&Force - 7-5-2014 at 12:09

elementcollector1, why are you redissolving the oxide in sulfuric acid? Have you already separated the Nd?

elementcollector1 - 7-5-2014 at 15:29

Yes. Now that it's separated, I need to convert it to a more volatile compound like chloride or fluoride to get the metal - this is most easily accomplished, in my view, with sulfuric acid, followed by sodium hydroxide, followed by hydrochloric acid.

Seriously, though, I must have done a really good job calcining it. No reaction whatsoever, and it's been a whole day.

blogfast25 - 8-5-2014 at 05:15

Quote: Originally posted by elementcollector1

Seriously, though, I must have done a really good job calcining it. No reaction whatsoever, and it's been a whole day.

Object lesson: never calcine anything you might later want to redissolve.

What colour is your oxide?

[Edited on 8-5-2014 by blogfast25]

Brain&Force - 8-5-2014 at 06:10

If it's already seperated, then why not dissolve it in hydrochloric acid now?

When you add hydrochloric acid, advise me on whether excess hydrochloric acid causes the solution to become more lavender in color. I saw this in woelen's site, and I want to confirm its occurence.

http://woelen.homescience.net/science/chem/exps/neodymium/in...

elementcollector1 - 8-5-2014 at 15:39

Color: White, white, and more white.

Given that sulfuric acid is doing precisely nothing to it, I would assume hydrochloric will be equally ineffective.

aga - 8-5-2014 at 15:44

Sulphuric Acid !

So, if i electrolyse CuSO₄ back into elemental copper and sulphuric acid, would that not work with Nd₂(SO₄)₃, precipitating Nd out of the solution ?

One way to find out i guess.

elementcollector1 - 8-5-2014 at 15:44

Heh. Nope. Look up the reactivity of neodymium - it's not friendly with water, to put it one way.

blogfast25 - 9-5-2014 at 04:41

Quote: Originally posted by aga

Sulphuric Acid !

So, if i electrolyse CuSO₄ back into elemental copper and sulphuric acid, would that not work with Nd₂(SO₄)₃, precipitating Nd out of the solution ?

One way to find out i guess.

Relatively few metals can be successfully plated out of watery solution, Nd is DEFINITELY not one of them: it's very electropositive and reacts with water.

MrHomeScientist - 9-5-2014 at 10:31

Yeah if it were that easy this entire thread wouldn't exist...

I plan on processing more of my magnet soup via the double sulfate method this weekend. Re-reading this thread has got me excited again. I want to try both the lithium reduction and deep eutectic solvent electrolysis. For lithium, I'm thinking harvest the foil fresh from a new battery and use it immediately. That way I avoid having to deal with getting rid of the oil on my current sample.

aga - 9-5-2014 at 10:48

Well, i tried it, but couldn't even dissolve the Nd₂(SO₄)₃ even using an ice bath.

The aim is to recover elemental Neodymium right ?

MrHomeScientist - 9-5-2014 at 11:27

That's my aim, at least.

Dissolving Nd-sulfate takes forever, even in cold solution where it is most soluble. Just crush it up as finely as possible and leave it stirring for an hour or so, if I remember right.

blogfast25 - 9-5-2014 at 12:07

It's much more soluble on ice but the rate of dissolution is likely to be slower as well.

Aga: recovering elemental Nd is very difficult. So far we know from literature that electrolysis of the molten chloride works. Little else seems to be possible, going by what's been discussed in the various Nd threads.

Töilet Plünger - 9-5-2014 at 12:42

I still attempting reduction of a dehydrated neodymium salt in something like pyridine, if at all possible. The problem is that a lot of neodymium salts are hydrated or hygroscopic. (And how would you produce the molten chloride if it tends to decompose when heated? Would it be possible to electrolyze the oxide?)

blogfast25 - 9-5-2014 at 13:25

If it wasn't clear, I meant of course the electrolysis of anh. NdCl3, in a eutectic mixture.

Acc. wiki that's how mischmetal is (was?) produced, from the mixed anh. LnCl3.

There's little reason to believe NdCl3 isn't stable enough: if it wasn't we'd be able to reduce it by chemical means.

Electrolysis in non-watery electrolytes remains on the table. Think deep eutectics?

Oxides: find a suitable solvent. Cryolite won't work here, I think

[Edited on 9-5-2014 by blogfast25]

Brain&Force - 9-5-2014 at 15:43

The problem with cryolite is that if you reduce it, you would probably get an Al-Nd alloy. (Which may be suitable if I study yttrium-scandium-aluminum alloys at some point.)

What about dissolving neodymium oxide in neodymium fluoride? I'm sure that's beyond the scope of our equipment, but it may be possible (I don't know if a sodium fluoroneodymate [that is an AWESOME chemical name] could exist).

blogfast25 - 10-5-2014 at 05:54

I don't think the Cryolite would be reduced, if you use the right voltage. But do we know it forms a eutectic with it, like alumina does? I don't think so.

NdF3/Nd2O3? Well, firstly look for eutectics of that system. The MP of NdF3 is about 1400 C (acc. http://www.phelly.com/ndf3/) Don't they use NdF3 in special glass? There may be some clues there...

The Ln don't seem very prone to forming complex anions.

[Edited on 10-5-2014 by blogfast25]

MrHomeScientist - 22-5-2014 at 19:32

Sort of thinking out loud here, so forgive me if this sounds rambling.

I'm processing more of my "magnet soup" neodymium and iron sulfate solution via the potassium double salt method, and initially got similar results as my test at the top of page 12. This time I combined 180mL of magnet soup with about 200mL of saturated potassium sulfate solution - after several minutes, a sandy pink precipitate slowly formed, and after letting it sit overnight a thin layer of tan gunk settled on top. I again attempted to dissolve this by adding sulfuric acid (thinking it must be iron hydrolysis products), but the tan precipitate remains. Maybe I just need more acid? The solution is quite acidic already, though.

I also recently got some NaSCN, so I tried my hand at testing for iron in the filtrate from rinsing the precipitate with acidified K₂SO₄ solution. Initially this showed no color change at all, but I added some hydrogen peroxide and the test tube immediately turned deep red. So amazingly enough it looks like iron(II) survived in solution for over a year and a half (!) and in fact there is barely any iron(III) present at all. Acidity doesn't interfere with the test for iron, does it? Maybe adding hydrogen peroxide to the precipitate will make the tan stuff soluble? This is going somewhat less smoothly than the sodium sulfate test I did, unfortunately.

blogfast25 - 23-5-2014 at 04:42

Quote: Originally posted by MrHomeScientist

This time I combined 180mL of magnet soup with about 200mL of saturated potassium sulfate solution - after several minutes, a sandy pink precipitate slowly formed, and after letting it sit overnight a thin layer of tan gunk settled on top. I again attempted to dissolve this by adding sulfuric acid (thinking it must be iron hydrolysis products), but the tan precipitate remains. Maybe I just need more acid? The solution is quite acidic already, though.

It's better (and also mentioned in the lit) to add the potassium sulphate directly to the solution, as a fine powder. That's what I do. Calculate the amount to achieve saturation at RT. Then heat and stir for about 1/2 hour and allow to cool. This way you're not affecting the pH of the solution, which risks precipitating ferric thingymejibs.

You may have to start over: alkalise with strong ammonia to convert double salt to Nd(OH)3 + K sulphate, filter off and dissolve solids in HCl, add new K suphate.

Remember that the concentration of iron in your soup is quite high: that makes it even more prone to hydrolysis, very simply put:

Fe3+ + 3 OH- < === > Fe(OH)3

... is pushed to the right by high [Fe3+].

[Edited on 23-5-2014 by blogfast25]

Brain&Force - 23-5-2014 at 14:00

In my work with terbium I add potassium sulfate directly to the solution as crystals. It works fine and a needle-like coating of the double sulfate grows on the crystals.

When I did this with terbium I had a clear coating of the gunk on top of the solution. It may be the neodymium that is forming this layer.

blogfast25 - 24-5-2014 at 04:19

Quote: Originally posted by Brain&Force

It may be the neodymium that is forming this layer.

I doubt it: the Nd double salt is white to pink too.

aga - 25-5-2014 at 13:32

Neodymium from magnets is simple.

All you need is a some Spin.

Magnet, acid ... Nd Sulphate.

You got the Nd didn't you ?
The Fe got lost yeah ?
The SO₄ is a Free Bonus !

Vote Aga at the next election. Vote Spin.

MrHomeScientist - 27-5-2014 at 06:41

I'm still working on this, but in the meantime I wanted to say that wow, adding solid potassium sulfate to the magnet solution is definitely the way to go. I figured people that added the solid were only concerned with keeping the volume as low as possible, but blogfast is right about the need to not change the pH!

I took 50mL of magnet soup (Nd and Fe sulfates) and added in just over 5g of solid K₂SO₄ with stirring. Within just a few minutes, the solution had gone totally opaque with pink precipitate of the neodymium double salt. I continued stirring for 40 minutes, then filtered. I washed the precipitate with 3 portions of acidified, saturated K₂SO₄ solution, followed by 2 portions of cold distilled water. I captured the last water wash and tested for iron by adding NaSCN and a small amount of H₂O₂ to oxidize the iron(II) to iron(III). This test came back positive, giving an orange color. I did several more water washes and tested again, with the same result. I'm now letting the precipitate dry on the filter paper, after which I'll scrape it off and vigorously stir in some distilled water to hopefully dissolve the remaining iron. I might need to acidify this a bit if I end up with iron hydroxides, etc. after drying.

TLDR: The precipitate from adding solid K₂SO₄ to the magnet solution appears much, much cleaner than my previous trial of adding a saturated solution instead.

blogfast25 - 27-5-2014 at 13:17

MrHS:

Strange how hard you find it to get rid of the last bits of Fe. I would try washing the dry precipitate with small amounts of hot 20 % H2SO4. Distilled water won't get rid of ferric iron: it'll be highly insoluble Fe(OH)3 by then...

"followed by 2 portions of cold distilled water" is pointless: all you do is precipitate the last bit of iron as Fe(OH)3. Think about it: at the stage where the Nd is still as double sulphate a bit of free potassium sulphate doesn't hurt at all. So there's no need to wash with pure water.

[Edited on 27-5-2014 by blogfast25]

MrHomeScientist - 28-5-2014 at 06:08

blogfast: I see, thanks for the input. I thought I was following the procedure you used way back earlier in the thread, but I could be mistaken. I could also be mistaken on how I'm doing the thiocyanate test. I take my sample and add a few drops of sodium thiocyanate solution (conc. not measured, just dissolved a few crystals in several mL of water). No reaction. I then add a small amount of 3% hydrogen peroxide to oxidize iron(II) to iron(III), and the solution turns a dark-ish orange. I believe this is a positive result, but everywhere I've read states the color of this complex should be very dark, blood red.

I found a procedure (attached) that uses potassium permangante to oxidize the iron instead. I imagine they use this so the end of oxidiation is clear ("Stop adding potassium permanganate drops when the purple colour persists for several seconds after addition"). Contrary to what I just said above, there's also a picture (Fig. 4) of several standard solutions of the complex - mine is usually about the color of the tube second from the right. Even they use "blood-red" as a descriptor, which isn't at all how I would label those colors.

Attachment: Determination of Iron by Thiocyanate Colorimetry.pdf (168kB)
This file has been downloaded 1220 times

blogfast25 - 28-5-2014 at 11:43

It's slightly unusual you're getting orange. Try this. Add peroxide to the iron but no thiocyanate yet. Heat a bit, allow to cool again or tap cool, then add thiocyanate.

When extracting Nd from magnet chloride with potassium sulphate, it helps to work kind of fast: don't allow the FeCl2 to oxidise to Fe3+. The Fe2+ is easier to get out because FeCl2 doesn't hydrolyse so much.

So, dissolve magnets in HCl, filter while still hot, add potassium sulphate in the right amount to filtrate, filter off double sulphate and take it from there.

MrHomeScientist - 8-6-2014 at 06:19

My magnet solution has been sitting on the shelf for over a year now, so I think I'm beyond the point of working fast

I took your advice from earlier and started over. Unfortunately I realized after doing all of the following that I forgot to take pictures!

Recovered 20.2g of iron-contaminated double sulfate. To break the double salt, I used 18.2g KOH in 200mL water. Stoichiometrically, only 8.6g is required, but I wanted a large excess to ensure full conversion to the hydroxides. The amount of water used was calculated to ensure all potassium sulfate was able to dissolve. This solution was added directly to the double sulfate powder, and stirred for ~1 hour.
Immediately, the solution turned coffee brown, indicating severe iron contamination. After stirring, this was filtered off to yield a chocolate brown, sandy solid of iron and neodymium hydroxides. The solution after filtering was water clear.

I then convered these hydroxides back into sulfates using 6.5mL conc. H₂SO₄ in 280mL H₂O. This was calculated starting from the 20.2g of double salt, rather than drying and weighing the mixed hydroxides themselves. After ~15 minutes of stirring, there was a small amount of solid that stubbornly refused to dissolve. This was filtered off to obtain 283mL of clear, tan colored (under lab fluorescents) solution.

To saturate this again with potassium sulfate, I weighed out 29.0g K₂SO₄ and added the solid directly to the solution with stirring. Stirring continued for ~1 hour. The precipitate that formed was light pink and very clean-looking (compared to the last contaminated batch). This was filtered and washed with several portions of acidified K₂SO₄, and the filtrate captured for testing.

I used a different strategy for testing this time. I still tried NaCSN to test for iron(III), but also took a second test tube and used potassium ferricyanide to test for iron(II). The iron(III) test came back negative, and over time this solution went milky from colloidal sulfur. I'd guess the acidity decomposed the NaCSN. The iron(II) test came back weakly positive, with the yellow solution slowly turning somewhat green. I washed twice more with acidified potassium sulfate, then tested this again. It was an even weaker positive, taking all night to eventually turn a light green. I think this is an acceptable level of iron contamination, since I believe this test is supposed to be pretty sensitive.

It looks like I've finally been able to achieve effective separation of Fe and Nd!

The Volatile Chemist - 8-6-2014 at 10:46

Quote: Originally posted by MrHomeScientist

It looks like I've finally been able to achieve effective separation of Fe and Nd!

Congradulations! Are you going to 'blog' about it? Or are there no pictures to show?

Brain&Force - 8-6-2014 at 12:04

It took you several tries? That's strange, because I was able to seperate terbium from the contaminated mixture in one go. But I had far less iron in solution than you had.

I have magnet soup of my own that I need to process.

blogfast25 - 8-6-2014 at 12:12

Personally I think it's possible to get the iron level down to almost 0 in one single purification, provided most of the iron is still in the ferrous state and you maintain high acidity all throughout the process.

But that's not really based on experience. When you precipitate the RE double sulphates from a sea of ferrous sulphate (Nd magnet soup), some co-precipitation of the FeSO4 may well be unavoidable...

Brain&Force - 8-6-2014 at 12:17

blogfast25, the differences in solubility between neodymium and terbium sulfates (terbium sulfate is far less soluble) may account for the observed differences. However, I need to find out exactly how much iron is needed to quench Tb fluorescence in order to set an upper limit on Fe contamination.

blogfast25 - 8-6-2014 at 13:04

Quote: Originally posted by Brain&Force

blogfast25, the differences in solubility between neodymium and terbium sulfates (terbium sulfate is far less soluble) may account for the observed differences.

B&F, we're talking potassium Ln double sulphates here, not simple Ln (III) sulphates. I don't see how the solubility of the latter could influence the degree of separation from iron.

MrHomeScientist - 8-6-2014 at 13:45

Quote: Originally posted by The Volatile Chemist

Quote: Originally posted by MrHomeScientist

It looks like I've finally been able to achieve effective separation of Fe and Nd!

Congradulations! Are you going to 'blog' about it? Or are there no pictures to show?

No photos of the process, sad to say. If I ever make it all the way to neodymium metal, there's absolutely a video in store when the process is perfected. It may be worthwhile to post one just getting the Nd separated (my progress up to this point).

The Volatile Chemist - 8-6-2014 at 18:31

Quote: Originally posted by MrHomeScientist

Quote: Originally posted by The Volatile Chemist

Quote: Originally posted by MrHomeScientist

It looks like I've finally been able to achieve effective separation of Fe and Nd!

Congradulations! Are you going to 'blog' about it? Or are there no pictures to show?

Definitely would be a worthwhile post!

Brain&Force - 10-6-2014 at 14:43

I tried adding some potassium sulfate to my magnet nitrate mix, but barely any neodymium sulfate precipitated. After adding in some sulfuric acid to keep the pH up and halving the volume of the solution, still nothing precipitated. I then added some more potassium sulfate, and something did precipitate - and it redissolved after a few minutes! This neodymium does NOT want to leave solution, even after being subjected to yet another round of heating.

The concentration of all the iron and neodymium must be nearing 5 molar now, and that's a very conservative estimate. I don't have much time, nor can I keep the solution so I really might just have to dispose of the solution and use another magnet.

[edit] wait, there may be something I can do to save the solution...

[Edited on 10.6.2014 by Brain&Force]

MrHomeScientist - 11-6-2014 at 06:42

Just an off the top of my head hypothesis, but mixing anoins might be "confusing" the precipitation. You could try dropping out everything as hydroxides, redissolving in sulfuric acid, and trying again to make the double sulfate. Perhaps in your case there just isn't enough sulfate around?

As for the earlier comment that it took me several tries, I think that was because of faulty method. Remember that I first started with "magnet sulfate" solution and added a saturated solution of K₂SO₄ to this. I got a good amount of precipitation, but this was heavily contaminated with iron due to, as blogfast pointed out, the addition of extra water changing the pH. I then returned this to solution and re-precipitated the double salt by directly adding solid K₂SO₄ instead of solution. This affords much, much cleaner precipitation. Comparing the leftover solution from the original precipitation with the leftover solution from the second method, it's clear that there's still some Nd left in the first one. So direct addition of the solid to the sulfate solution is definitely the way to go for effective and complete separation.

[Edited on 6-11-2014 by MrHomeScientist]

Brain&Force - 11-6-2014 at 14:22

That's a good idea with the multiple anion mixture, but the only thing I've been using to precipitate the neodymium are sulfates, and the only other anion in solution is nitrate - no hydroxides anywhere.

Also, I don't know if you've observed this, but whenever I heat the magnet nitrate solution, it turns very dark, but returns to its original color as it cools down. Is this because the nitrates are decomposing in solution?

Just wondering, where would be the best place to store a magnet nitrate solution for easy transport? It's quite acidic at this point, and I don't want to end my work now.

blogfast25 - 12-6-2014 at 04:39

Quote: Originally posted by Brain&Force

Also, I don't know if you've observed this, but whenever I heat the magnet nitrate solution, it turns very dark, but returns to its original color as it cools down. Is this because the nitrates are decomposing in solution?

One possible explanation is the following. Using nitric acid as the acid to dissolve the magnets, the iron is oxidised to Fe3+. Ferric ion solutions are thermochromic because higher temperature pushes (simplified):

Fe3+ + 2 H2O < === > Fe(OH)2+ + H3O+ to the right.

Fe(OH)2+ is dark coloured in high concentrations (such as encountered in magnet soup). It appears this cation gives ferric solution it colour ranging from yellow to orange to much darker colours because [Fe(H2O)6]3+ appears to be almost colourless.

[Edited on 12-6-2014 by blogfast25]

MrHomeScientist - 12-6-2014 at 05:28

Quote: Originally posted by Brain&Force

That's a good idea with the multiple anion mixture, but the only thing I've been using to precipitate the neodymium are sulfates, and the only other anion in solution is nitrate - no hydroxides anywhere.

What I meant was that you have a 'magnet nitrate' solution, and are adding sulfate to this. I have a 'magnet sulfate' solution, and am adding sulfate. So in my case there's more sulfate around, perhaps leading to easier precipitation. So my suggestion was to add hydroxide to precipitate everything, then redissolve in sulfuric acid instead.

Although now that I say that, blogfast has been working with 'magnet chloride' with no problems... hm.

As for the reversible color change, I've definitely noticed that too. It looks like some sort of thermochromism, not related to the anion (since ours are different). This experiment is full of surprises!

blogfast25 - 12-6-2014 at 09:56

Quote: Originally posted by MrHomeScientist

Although now that I say that, blogfast has been working with 'magnet chloride' with no problems... hm.

It shouldn't matter what the main anion in the soup was: the moment you saturate it with K2SO4, the least soluble crystallites, i.e. the (Ln(III),K) double sulphates (as hydrates), crystallise out. Should also happen with nitrates present...

Brain&Force - 13-6-2014 at 07:54

I'm just spitballing here, but is it at all possible that a less soluble salt forms with the chloride along with the sulfate?

blogfast25 - 14-6-2014 at 05:14

Quote: Originally posted by Brain&Force

I'm just spitballing here, but is it at all possible that a less soluble salt forms with the chloride along with the sulfate?

It's not impossible but we have no evidence for it. Better to stick with what we know.

Back on the Wagon

MrHomeScientist - 26-10-2014 at 14:48

Once again I've come back to this project - this time attempting the final step: lithium reduction of neodymium fluoride.

3Li + NdF3 == 3LiF + Nd

For the impatient, it appears to have been a failure. The hot lithium most likely reacted with the air. Read on the for details.

For this first test, I used my first batch of the fluoride since it is very obviously contaminated by iron. It's green under fluorescent light, and pink in daylight.

After grinding to a fine powder, this was 1.4g NdF₃. To this, I planned to react it with two slugs of my lithium, which was 0.17g.

Molten lithium is very corrosive to a wide variety of materials, so choosing an appropriate crucible has proven to be difficult. I went with graphite this time, even though it is unsuitable for liquid lithium at 1000C (the target temperature here is 1024C, the melting point of Nd). I chose this because, frankly, it was what I had. Also I figure the lithium won't be around for very long at that temperature anyway, having mostly been reacted away by that time. Here's everything in the crucible.

I also did not use an inert atmosphere because, frankly, I haven't figured out the logistics of providing one to my crucible and furnace. I just wanted to try it out, and see if it did anything at all. Shortly after starting heating, the lithium started glowing red hot. You can see by the crucible that this is well under the temperature where this should be happening, so that must mean it reacted with something.

After heating for several more minutes with no change, I put the crucible in my mini furnace and cranked it up to somewhere in the neighborhood of 1000C. Nothing ever melted.

After cooling, I was left with a rock hard mass of grey crust. This stuff is nearly impossible to remove from the crucible.

No metal of any kind in sight. It looks like the lithium reacted with the air first. I guess I'll need an inert blanket after all. I figured as much, but at least wanted to try this to see what happened.

j_sum1 - 26-10-2014 at 14:59

Great stuff -- well kinda.
Watching you closely since I intend to do much the same with LaCl3
(after buying some fluoride, after converting to LaF3, after crystallising, after purchasing a crucible, maybe after lining it with Mo or Ta, after building a furnace and after working out a way to feed argon into my setup...)
Right behind you.

Tdep - 26-10-2014 at 16:11

This is such an interesting project.

I'm probably missing something, but can't you cover the reaction in sand or something similar? Surely that's a lot easier than building an inert atmosphere around it. Sure, you'll gets some silicon impurities from the lithium/SiO2 reaction but the molten Nd isn't going to reduce the sand is it? And if the NdF3 is layered above the lithium, the sand shouldn't get in the way that much. You're going to have to physically remove it, but you'll have to do that for the lithium slag anyway, even if a inert atmosphere is used.

If not sand, is there some other inert flux that will resist the high temperatures and the super reactive lithium? Aluminium oxide possibly??

j_sum1 - 26-10-2014 at 16:14

I had the same thought. But I was thinking excess NdF3. If the lithium was completely contained within a pile of the reactant then this would resolve some of the containment and side reaction issues.

Pok - 26-10-2014 at 16:57

@MrHomeScientist: don't you think that neodymium itself will burn in air at such high temperatures? If any of it was produced by reduction with lithium, the product would look exactly the same (a mixture of lithium/neodymium oxide/fluoride).

Unfortunately, neodymium reacts with nitrogen when it burns (like lithium does). So if you can detect nitrides by reacting the product with water to produce ammonia, it would not necessarily mean that lithium reacted with air.

[Edited on 27-10-2014 by Pok]

Zephyr - 26-10-2014 at 18:53

I recently made a small sample of neodymium sulfate by heating a magnet in concentrated sulfuric acid and then after letting the hot solution settle decanting off the top layer, and then repeating.
Here is the end product:

MrHomeScientist - 26-10-2014 at 19:25

Tdep: Molten lithium is some extremely corrosive stuff. We've been discussing it in a few other threads. The gist of it is that pretty much any oxide is no good, because the Li will rip the oxygen out of it. It reacts with glass pretty vigorously, so sand would be a no go. I'd definitely prefer inert blanket, if I can figure out how to do it.

Pok: Well the hope was that the LiF produced would form a protective layer over the denser Nd metal, if everything was good and molten. I charged the crucible by putting some of the NdF₃ on the bottom, then the Li ingots, then covered with the rest of the fluoride. It might be better to put the Li on the bottom to better protect it and ensure the reaction goes from the bottom up, keeping the Nd buried and protected.

blogfast25 - 27-10-2014 at 05:55

Mr HS:

I think it could be useful to line the bottom of the crucible with a small excess of NdF3, to prevent the Nd metal sticking to the crucible. Just an idea...

MrHomeScientist - 27-10-2014 at 09:01

Perhaps... I'm torn between using an excess of lithium (precious and reactive!) and using an excess of NdF₃ (takes literally years for me to make!

).
In my other experiences with graphite crucibles, the stuff just pops right out. So I don't know - could be indication of a reaction between C and Li?

I thought a bit more about inert atmosphere today, and this is my preliminary idea for such a setup:

A crucible (iron, graphite, Ta-lined, whatever) is placed inside an iron pipe with space around it. This space is filled with small steel ball bearings for heat transfer. An iron pipe cap is placed on top, with two holes drilled in it. Through these holes are inserted two smaller iron pipes for gas entry/exit.

Argon flows in from the left tube, which extends down into the crucible. It exits from the right tube, which only just comes in past the pipe cap. The entire assembly is heated from the outside via propane torch, and is expected to reach at minimum 1024C. (which is why copper won't work)

I have precisely zero experience working under these conditions, so if I have overlooked anything or if there is a severe safety hazard I'm missing please let me know.

blogfast25 - 27-10-2014 at 09:39

MrHS:

A small excess of NdF3 helps also as follows:

NdF3 + 3 Li < === > Nd + 3 LiF

... is pushed to the right. But use an excess Li instead and where will it end up?

Re. your argon design, perhaps using copper wire instead of ball bearings? But if you're heating from the bottom then neither may actually do all that much...

MrHomeScientist - 27-10-2014 at 11:00

Well I wanted to avoid copper because the target temperature of Nd's melting point (1024C) is pretty close to copper's melting point (1085C). The last thing I want is a non-nuclear meltdown!

It might not be clear from picture 5, but my furnace has the torch come in from the side. Similar to picture 4. I figure I'll need to heat the whole thing up to 1024 to ensure molten products. I don't know if the reaction itself will proceed with heat to spare ala thermite.

Come to think of it, now that I have this furnace setup, is magnesium reduction a possibility again? We scrapped that because it didn't produce enough heat of reaction, but if I heat the whole thing externally that won't be a problem. Then I wouldn't need inert atmosphere either, which would simplify greatly. In fact I think I still have the Mg + NdF₃ mixture that failed from my first attempt...

Pages: 1 2 3 4 5 6