Sciencemadness Discussion Board

Alkali fusion of SnO2

blogfast25 - 4-10-2010 at 12:06

... for the synth of M2SnCl6 (hexachlorostannate).

(NH4)2SnCl6 (for instance) can be synthesised via various routes:

1) Direct union of ice cold SnCl4 with saturated ice cold NH4Cl. Crystallise.
2) Oxidation of SnCl2/HCl solutions with air oxygen (to basically H2SnCl6), combine with NH4Cl. Crystallise.

Having neither SnCl4, nor SnCl2 at hand, but having a rather large quantity of 'good grade' SnO2, I chose that as a starting material. SnO2 is perhaps best used for reduction to Sn with C but I've already done that in the past. Dissolving the resulting Sn in HCl is rather a slow process and for the moment I'm not really equipped to attempt directly chlorinating the Sn...

So here's what I did.

75 g of SnO2 were mixed with 48 g of NaOH prills (20 % excess to stoichio) and 48 ml of water and loaded into a small steel crucible. Heated to about 200 - 250 C al forno for about 1 1/2 hours, the mix was still moist and was further dried slowly on the hob. Heat was then cranked up to maximum, which I suspect was around 500 C, for three hours. It was the allowed to cool down and 50 ml of water was added as a cold 'pre-soak' overnight.

The resulting fusion product (2 NaOH + SnO2 --> Na2SnO3 + H2O) is hard and needed to chiselled away with plenty of boiling water, nothing a little elbow grease can't solve. Clearly not all the stuff dissolved to water soluble Na2Sn(OH)6 (Na2SnO3 + H2O ---> Na2Sn(OH)6) but from early on the leachate tended to get covered in a transparent, crusty crystallite: Na2Sn(OH)6 is reportedly less soluble in hot water than cold...

The leachate tested positive for soluble Sn +IV with HCl, which caused rather a lot of 'tin acid' to precipitate. That redissolved remarkably easily in excess HCl (forming basically H2SnCl6), from which the tin acid can then be re-precipitated with NaOH or NH4OH. I would recommend precipitating from the alkaline solution which yields a more flocculant precipitate, from HCl the obtained precipitate is very hard too filter.

The precipitate, filtered and washed, was then dissolved into a minimum of HCl 22%, plus a dash for good measure. That volume by then was about 150 ml of solution, to which 7 g of NH4Cl (being low on KCl) were added and about 50 g of water was boiled away. Immediately upon cooling well-formed crystals of (NH4)2SnCl6 appeared. The solution has now been fridged and the crystals will be recrystallised from a minimum of hot water (plus a dash of HCl).

When I tested the alkaline filtrate (about pH 12), it too still tested positive for water soluble Sn +IV, so the pH was clearly a bit too high for the amphoteric, freshly precipitated 'tin acid'.

So much for the alkaline leachate...

*/*/*/*/*/*/*/*/*

The water insoluble part of the fused mix was then treated with about 300 ml of 22% HCl, simmered for about 30 mins. That reaction went on was unmistakable, it behaved very much like the fusion product of NaOH + zircon that I carried out a few months ago. Testing for Sn +IV showed the leachate to contain quite a lot of SnCl4. This suspension is now standing overnight to decant. From the amount of fine, heavy white powder left at the bottom, presumably unreacted SnO2, I'd say roughly 50 % of the reagent SnO2 did not fuse-react. All the 'hard but crumbly' material had either dissolved or left behind as a fine powder...

It appears that in the relatively mild fusion conditions here used, part of the SnO2 does react to Na2SnO3, while another part forms an acid soluble Sn +IV species, while a third part doesn't react at all...

To be continued...

Interesting reading on (NH4)2SnCl6 and (NH4)2PbCl6 synthesis and characterisation here:

http://cscdic.pbworks.com/f/MacKinnon-Spivak+LakeU+2.pdf

blogfast25 - 13-10-2010 at 06:22

Another test, this time 25 g SnO2, a 50 % excess of NaOH and the same SnO2/water ratio as before, heated in the same way but 2 h at 'high heat' instead of 3 h.

After the fusion, 50 ml of water was added to the fusion product and left to soak overnight. The crucible was then put on a low heat until the water began to boil and simmered for a few minutes. This treatment allowed recovering the fusion product from the crucible remarkably easily.

After a bit of crushing and stirring, the equivalent amount of 22 w% HCl (equivalent to the amount of NaOH) was then added, the pH carefully adjusted to about 7 and the solution simmered for a bit. It was clear that much of the SnO2 had been fused to Na2SnO3, which dissolved as Na2Sn(OH)6 and upon neutralising hydrolysed to 'Sn(OH)4'. At the bottom of the beaker was some fusion product that clearly hadn't been affected by the treatment with water and the newly formed Sn(OH)4 was decanted off for filtration.

What I forgot to mention in the post above is that I met the enemy of all filtrations... peptisation! Namely when I was washing the first Sn(OH)4 precipitate, it ran through the filter after several washings.

To avoid this, the new precipitate was washed using a concentrated solution of NH4Cl. It filters very badly, vacuum is really needed (but I haven't got any yet).

The precipitate was then dissolved in a hot excess of HCl 22 w% and the solution was almost entirely clear. After adding NH4Cl and reducing the volume by simmering, on cooling crystals of (NH4)2SnCl6 formed, more so upon overnight refrigerating.

These crystals dissolve relatively easily in cold water, nice and clear. But heating the solution causes Sn(OH)4 to drop out, which on adding HCl dissolves on standing. Dissolving some (NH4)2SnCl6 in 22 w% HCl renders the solution immune to boiling. A neutral solution of (NH4)2SnCl6, when 1.5 M NH3 was added produces Sn(OH)4 too, which disappears when strong HCl is added.

The bottom line is that although H2SnCl6 is a strong acid and its salts behave neutrally, the SnCl6 (2-) anion is far from impervious to hydrolysis. As witnessed also by a failed solubility determination. For future recrystallisations I wanted to determine the approximate solubility of the salt in water at 100 C. About 10 g of the dry salt were combined with a few ml of 22 % HCl and about 15 ml of water but on heating to 100 C much Sn(OH)4 dropped out: clearly the acid reserve of the solution was too low to prevent hydrolysis.

All this points to the fact that the salt can be recrystallised from hot concentrated HCl but not from hot water.

This then left treating the fraction of the fusion product that hadn't reacted with water to form Na2Sn(OH)6. It was treated with hot strong HCl in which most of it clearly does dissolve... except that the remainder (which must be unreacted SnO2) is extremely fine and not separable with my modest filtering media. It was possible to crystallise the hexachlorostannate but of course it's contaminated with ultra fine SnO2. This fineness manifests itself also in extreme clinginess to glass ware: unless you scrub all glass ware with fine steel wool and some elbow grease, you end up with everything coated with an opaque coating of SnO2...

For all this trouble I still have only about 30 g or so of the ammonium hexachlorostannate, so this method is far from ideal.

I'm considering an alternative method which is to oxidise a solution of SnCl2 in strong HCl with air oxygen to SnCl4 and saturating the solution with NH4Cl: cooling to near zero should then make the ammonium hexachlorostannate crystallise out...

Fe2(SnCl6)3?

blogfast25 - 14-10-2010 at 05:50

Today I noticed something unusual and slightly exciting...

Some left over precipitated Sn(OH)4 was decanted off yesterday and dissolved in a minimum of hot HCl 22 w%, then left to stand overnight at RT (I forgot to put it in the fridge). Today the solution showed, apart from a small amount of (NH4)2SnCl6 (there would have been a lot more after cooling to about 5 C), some very well formed amber-red crystals, 1 - 2 mm, with apparently two axes of symmetry perpendicular to each other. Total amount definitely less than 1 g, though.

The supernatant solution is a urine yellow, due to the presence of Fe (III), which is sourced from the technical HCl and maybe some pickup from the SS crucible. The raw material may also contain some Fe as it is off white. So the solution definitely contains H3O+, NH4+, Cl-, Fe3+ and SnCl6 (2-)... There are also some contaminants, in all likelihood small amounts of Ca2+, some Na+ and some organics coming from the NH4Cl (don't ask!)

Could it be that the crystals are Fe2(SnCl6)3? This will need a bit of crafty 'micro' chemistry to resolve... 'In theory' Fe2(SnCl6)3 could synthesised by dissolving fresh Fe(OH)3.n H2O in H2SnCl6.2 H2O (which can be isolated acc. Holleman).


[Edited on 14-10-2010 by blogfast25]

woelen - 14-10-2010 at 22:52

This synthesis looks quite interesting. I also have SnO2 (a white powder) and I never did anything with it, after I discovered that it is so inert. Do you really have to heat for such a long time as you mentioned in your first post before the stannate(IV) can be leached out of the solution? My powder is very fine.

I have a nickel crucible and I could try the reaction in that. I do not, however, want to destroy the crucible (does it withstand oxidation by SnO2?).

The red crystals you have might certainly be an iron(III) salt of hexachlorostannate(IV). I have seen quite a few iron(III) salts which have a deep red blood-like color.

Could you make some pictures of your ammonium salt and of the supposed iron-salt?

watson.fawkes - 15-10-2010 at 04:43

Quote: Originally posted by woelen  
This synthesis looks quite interesting. I also have SnO2 (a white powder) and I never did anything with it, after I discovered that it is so inert.
In mineral form, this is cassiterite, the classical, as in > 4000 years old classical, ore for smelting tin. It smelts easily with carbon.

blogfast25 - 15-10-2010 at 06:23

Quote: Originally posted by woelen  
This synthesis looks quite interesting. I also have SnO2 (a white powder) and I never did anything with it, after I discovered that it is so inert. Do you really have to heat for such a long time as you mentioned in your first post before the stannate(IV) can be leached out of the solution? My powder is very fine.

I have a nickel crucible and I could try the reaction in that. I do not, however, want to destroy the crucible (does it withstand oxidation by SnO2?).

The red crystals you have might certainly be an iron(III) salt of hexachlorostannate(IV). I have seen quite a few iron(III) salts which have a deep red blood-like color.

Could you make some pictures of your ammonium salt and of the supposed iron-salt?


@Woelen:

I have no idea whether these long heating times are necessary but some test tube tests, using for instance 50 w% NaOH or KOH, using prolonged boiling/heating could probably show how quickly a significant water (or acid) soluble leachate can be obtained. It probably depends a lot on the product: fineness and degree of calcination. I modelled this fusion on the only other alkali fusion I've ever carried out and that was on zircon (ZrSiO4), which is reported on this forum. For SnO2 it may well be slight 'overkill'...

As regards nickel in these circumstances, I can see no problem: I don't think it will be attacked significantly... Note that there is no change of oxidation state of the Sn, so there can be no 'oxidation by SnO2' (Ni isn't capable of reducing SnO2), but air combined with strong alkali may still have some effect on the crucible... 'Lucky' you to have such a crucible (green with envy!)

As regards pics, I'll see what I can do this WE but there really isn't much to look at: the best crop of (NH4)2SnCl6 was the very first one (about 10 g) and looks almost completely white, sheet like structures that are clearly crystalline. The actual crystals (that make up the sheets) are quite small because there very clearly is a very big drop in solubility between 100 C and RT and the crystals thus don't have much time to grow. Subsequent crops have tended to be more off-white, I suspect some residual fine SnO2 which has proved hard to remove...

As regards the suspected Fe (III) hexachlorostannate, I've yet to collect the crystals. I think it would be highly interesting to convert the NH4 salt to H2SnCl6.2 H2O (ac. Holleman by saturating a saturated solution of the salt with HCl: sheet like crystals of the stable and strong acid are then supposed to appear) and combine it directly with fresh Fe(OH)3.n H2O...

As regards watson.fawkes' mention of reducing SnO2 with carbon, that too is a very interesting bench experiment which I carried out a few years ago. Combine an excess of fine C with SnO2 (I believe the reduction runs to CO, not CO2) and heat it to about red heat (test tube + strong bunsen flame might actually do it...), the mixture then starts behaving like a so-called 'fluid bed', with bubbles of the CO escaping the powdered mix, giving it the appearance of a boiling liquid. At the same time small glistening pearls of liquid Sn form which over time coalesce and gather at the bottom of the reactor. After completion you end up with a molten bath of Sn, covered with unreacted C which protects it from oxidation. Neat...

blogfast25 - 15-10-2010 at 08:20

Here's a picture of about 10 g of (NH4)2SnCl6. Sorry about the poor focus: in reality the sheets glisten quite brightly with crystalline surfaces. I also noted quite a strong smell of HCl coming off it when I got it out of the CaCl2 desiccators, perhaps not surprising since as it was cold re-crystallised from strong HCl...



Hopefully a picture of the amber-red crystals tomorrow...


[Edited on 15-10-2010 by blogfast25]

watson.fawkes - 15-10-2010 at 10:26

Quote: Originally posted by blogfast25  
Combine an excess of fine C with SnO2 (I believe the reduction runs to CO, not CO2) and heat it to about red heat
Cupola melting of iron has a large fraction of CO in the exhaust. As I understand it, this is generic, given the reaction CO2 + C --> 2 CO at high temperature.

Shortly after I posted this, I got to wondering whether you could reduce SnO2 with CO for a higher-purity smelt operation. I'm certain that H2 works, but that's rather overkill. In the case of tin, you've got stable oxides at both +4 and +2, yield the following reactions: SnO2 + CO --> SnO + CO2; SnO + CO --> Sn + CO2. I imagine this is already happening to some extent in ordinary carbon smelting.

Practically speaking, I had a thought about a way to replace a CO gas generator for trying this. I was thinking that perhaps a double-crucible system might work: an outer crucible with the carbon source and an inner one with the SnO2, with a loose-fitting lid on the whole thing allowing gas to escape. Initial atmospheric oxygen creates the initial CO charge. Eventually the nitrogen would be driven out and the atmosphere inside would be a CO + CO2 mixture. It's conceivable that there's a need to bootstrap nitrogen exclusion, in which I'd use a carbonate. Put a layer of sodium or calcium carbonate on the bottom and a layer of carbon over it.

blogfast25 - 15-10-2010 at 12:12

Quote: Originally posted by watson.fawkes  
Quote: Originally posted by blogfast25  
Combine an excess of fine C with SnO2 (I believe the reduction runs to CO, not CO2) and heat it to about red heat
Cupola melting of iron has a large fraction of CO in the exhaust. As I understand it, this is generic, given the reaction CO2 + C --> 2 CO at high temperature.

Shortly after I posted this, I got to wondering whether you could reduce SnO2 with CO for a higher-purity smelt operation. I'm certain that H2 works, but that's rather overkill. In the case of tin, you've got stable oxides at both +4 and +2, yield the following reactions: SnO2 + CO --> SnO + CO2; SnO + CO --> Sn + CO2. I imagine this is already happening to some extent in ordinary carbon smelting.

Practically speaking, I had a thought about a way to replace a CO gas generator for trying this. I was thinking that perhaps a double-crucible system might work: an outer crucible with the carbon source and an inner one with the SnO2, with a loose-fitting lid on the whole thing allowing gas to escape. Initial atmospheric oxygen creates the initial CO charge. Eventually the nitrogen would be driven out and the atmosphere inside would be a CO + CO2 mixture. It's conceivable that there's a need to bootstrap nitrogen exclusion, in which I'd use a carbonate. Put a layer of sodium or calcium carbonate on the bottom and a layer of carbon over it.


Reduction of SnO2 with CO probably happens at higher temps. I'm not sure what the temperature inside the reactor was: I heated the (steel) crucible on a hot charcoal fire but the contents of the crucible didn't emit any light yet...

12AX7 - 15-10-2010 at 17:00

I wouldn't be surprised if SnO2 + C doesn't proceed at all without catalytic oxygen (i.e., in a vacuum). I mean, yes obviously it'll still work, but being contact and diffusion limited means a way higher temperature.

Carbon burns to CO2 at a fairly low temperature, and CO at red heat. Normally, the boundary layer of burning CO is invisible, since CO has a pale flame. It's quite possible that you simply can't see the CO layer at low temperatures and it always burns to CO as an intermediate. That would work for a pile of SnO2 + C where airflow is weak.

Tim

blogfast25 - 16-10-2010 at 04:15

Hmm... interesting points, Tim...

arsphenamine - 16-10-2010 at 14:10

Water gas is an equimolar mix of CO and H2,
formed by passing steam through burning carbon.
Might be useful here.

blogfast25 - 24-10-2010 at 07:01

Finally a (not so good) shot of the amber-red crystals but still mixed into a matrix of (NH4)2SnCl6 which is coloured far more yellowish than it normally is:



The colour contrast does not do reality justice.