"Aqueous caustic soda (NaOH) 50 wt% solution was prepared to using as the decomposition agent of zircon sand. The prepared solution and Brazilian
zircon sand were mixed in wt/wt ratio of 1.0; 1.2 and 1.5 NaOH/ZrSiO4. The moisture was carefully charged into an iron crucible and set in an electric
furnace. The alkaline fusion reaction was conducted following thermal treatment: first heating at rate 20oC/min up to 100oC, remained at this
temperature for 1h and then heating at rate 20oC/min up to 200oC, remained for 1h and finally the temperature was increased to 600oC for 2h, at that
time the alkali fusion reaction takes place and caustic frit is formed, a mixture of sodium zirconate and sodium silicate. This frit was washed with
water, removing the soluble sodium silicate and hydrolyzing the zirconate to an impure hydrous zirconium oxide. The water-leached frit was treated in
cold 8M HCl at 80oC for obtaining an impure zirconyl chloride solution."
Here the reaction that's supposed to take place is:
(1) ZrSiO4 + 4 NaOH --> Na2ZrO3 + Na2SiO3 + 2 H2O
On leaching with water the Na2ZrO3 is then supposed to hydrolise acc.:
(2) Na2ZrO3 + 3 H2O ---> Zr(OH)4 + 2 NaOH
with the sodium silicate remaining in alkaline solution.
I chose the 1.5 ratio NaOH to ZrSiO4 (so 6 mol of NaOH per mol of zircon) and thus 40.0 g of very finely ground pottery grade zircon (Bath Potteries)
and 28.6 g of NaOH pre-dissolved in the same amount of water were charged (together all ingredients form a fairly thin sludge) into a 'steel
crucible'. In reality this was a former sugar bowl, with an outer diameter about the same as a standard 1/2 L soup can. The wall thickness of the
crucible was about 0.5 mm and it was lodged inside said soup can, using the can lid as a lid for the inner crucible.
The assembly was then heated in a preheated gas oven at 150 DC (my oven's lowest setting) for an hour and then for another hour at 200 DC.
The fusion was then carried out by transferring the hot assembly to a 'paint can furnace', BBQ charcoal fired. This Dinky Toy furnace is very capable
of melting aluminium, at a push even copper coins. Drawbacks are the lack of temperature regulation and the need to refuel with charcoal at least
every 10 minutes.
After 2 hours of heating the assembly was allowed to cool. Inside it was a hard mass, clearly to my mind some fusion had taken place. There were also
some 'canals' formed by the last escaping water. I added some 50 mL of water to it to allow soaking overnight.
The next day the fusion product was prized out of the inner crucible with some difficulty and crushed in a granite mortar. No signs of zirconate
hydrolysis were evident.
To the crushed sludge more water was added and it was gently simmered for about 1/2 h. The strongly alkaline supernatant was decanted off and
neutralised. To my disappointment, very little silica appeared, almost none in fact. At this point I more or less concluded that the fusion had
failed.
I had also added some more water to the empty crucible to soak the few bits I hadn't been able to extract and there small but discernable amounts of
gelatinous Zr(OH)4.nH2O had formed...
In a last ditch attempt, several hundred mL of 32 % HCl were added to the crushed sludge and simmered for about 45 min. It simmered very erratically
(quite differently from the attempt just using water), with loud bangs every so often. It was left to cool and sediment (the zircon is so fine it runs
right through my coffee filter).
Decanting off the clear supernatant liquid and neutralising it with about 30 g of dissolved NaOH, large amounts of what I presume is crude hydrated
zirconia dropped out, resulting in some 300 - 400 mL of a thick white gelatinous sludge.
An attempt at calculating a very semi-quantitative yield will be made by drying the remaining solid that didn't dissolve in the HCl and assuming it's
mostly silica. Some of the residue seems quite clunked together, as if some reaction had taken place...
So fusion had taken place after all but it appears reaction (1) wasn't the one that prevailed. Google for 'zircon alkali fusion' and you'll
find references to the formation at higher temperatures (800 DC and higher) of other zirconates than the one in (1), as well as sodium zirconium
silicates, all of which are claimed to less prone to hydrolysis (and thus industrially less desirable). 600 DC seems to be roughly optimal and the
likelihood is that the fusion here took place at higher temperature.
This time there was also iron pick up, from the crucible. Presumably molten alkali, iron and air can form FeO3(2-), (Fe [+IV]) as described by
Holleman, which then hydrolyses to Fe [+III]. The crude zirconyl chloride solution was yellow.
The crude zirconia hydrate has now been filtered and washed with boiling tap water, for further purification. JohnWW - 21-5-2010 at 15:01
A plain cast iron crucible is liable to lose a substantial amount of Fe into solution with a hot concentrated alkali, although it can be used as a
cheap alternative in situations where the Fe is either not critical or can easily be removed afterwards. A much better crucible material to use would
be pure Ni, or high-alloy austenitic stainless steels particularly Inconel (high Cr+Ni steel), or cupro-nickel alloys particularly the Monel alloys.
See chapt 23 of Perrys Chemical Engineers Handbook, for which I have posted links in the References (chemical engineering) section.
Zr is much more electropositive than Si, so ZrO3-- can be expected to precipitate out as Zr(OH)4 much more easily when the pH is increased, than would
SiO3-- as Si(OH)4. Apparently, mere dilution of the solution, without adding any acid, is sufficient according to the article.
Anther point is that naturally occurring Zr minerals such as ZrSiO4 also contain some Hf (which would also go into solution on alkaline fusion); and
sometimes Th, although the Th will not go into solution on alkaline fusion, ThO2 not being amphoteric. There is not usually Ti present in significant
amounts, however, unless the zircon sands also contain Ti minerals such as rutile or ilmenite, which may be separable before alkaline fusion. To
separate Hf from Zr, if desired, would require either finely-controled precipitation by dilution, or fractional crystallization after conversion into
a more easily soluble form, particularly as salts of ZfF6-- and HfF6-- or ZrF7--- and HfF7---.
Also, what do you want to use the zirconia for? It is used, as is, to make zirconia chemical crucibles, and also components for nuclear reactors on
account of the high nuclear cross-section of Zr and high melting-point of ZrO2; and as a pigment and opacifier in paints and vitreous enamels as an
alternative to TiO2. It is also used in a water-soluble analytical reagent for the SPADNS colorimetric/spectrophotometric method of analysis of F in
water. Zr metal is used as a "scavenger" for oxygen in certain steels.blogfast25 - 22-5-2010 at 06:43
Hi John,
I'm not looking to produce zirconia but zirconium metal itself, by magnesiothermic reduction of ZrF4. For this I need suitable quantities of a water
soluble Zr salt, zirconyl chloride being the most obvious one. I've no access to HF, so I'm going via the (NH4)2ZrF6 route, assuming I get that far!
I kept the reaction product of a previous attempt at fusion (a couple of years back, gathering dust in a filter on a shelf somewhere) and yesterday
treating a small amount of if with 32 % HCl, I'm getting soluble Zr too, so presumably that fusion hadn't failed either, although at the time the
water leachate didn't show any silica either.
For now the main bother is Fe contamination which I need to get on top of. Even with Fe-free crucibles and reagents, Fe2O3 is usually present in
zircon anyway.
Another problem is not having any pure zirconyl chloride to get familiarised with. It's a bit of a kooky compound, with this quasi-polymeric
structure...
As regards separating the Hafnium out, I've contemplated that but to obtain a bit of Hf compound, rather large amounts of natural Zr would have to be
processed, prohibitively expensive for me I think...
[Edited on 22-5-2010 by blogfast25]blogfast25 - 6-6-2010 at 10:37
Electroplating out the Fe3+ from zirconyl solutions appears possible. I first tested the idea with some ferric ammonium solution buffered to pH = 5
with acetate buffer, using Zn and also Al. Both work but the Al needs de-passivating (soaking in HCl until reaction starts).
Then a ZrOCl2 solution was carefully neutralised and buffered with pH = 5 acetate buffer. Some white, gelatinous precipitation took place but not
much. It was filtered off and set aside (I presumed it contains no Fe) and the filtrate subjected to de-passivated Al: the Fe disappeared (going by
colour and SCN- test). Re-alkalising the solution made large amounts of hydrated zirconia precipitate.
Zn would have the advantage that upon the final precipitation with ammonia, Zn2+ would stay in solution as Zn(NH3)2 2+.
Venables' monography of zirconium and its compounds mentions 'basic acetates', so maybe that's what the initial precipitate was (or maybe it's just
plain old Zr(OH)4.n H2O?). I will run another test with an ammonium based buffer, possibly at slightly lower pH still.
Venables also touches at length upon the problematic separation of Fe and Zr, with plenty of old references. Today I'd imagine the separation to be
carried out by means of ion exchange columns.
It would also appear the source of the Fe is mainly from my 'drugstore HCl', not from the bleach. I now have a 32 % grade (about 9 M) but it too
contains small amounts of Fe.
One of the biggest problems has been to crystallise the ZrOCl2.8 H2O: it hydrolyses like mad if you're not careful, especially by boiling down
solutions. It also crystallizes rather slowly (a consequence of its alleged polymeric nature?): boil down too far and the solution cools down to a
gummy mass that no longer dissolves in pure water (although it redissolves in strong HCl). From my observations, the trick seems to be to reduce the
solution until fumes of HCl start coming off. Let cool and a white, water soluble material then slowly crystallises over minutes, if not longer. And
there must always be some liquor left, otherwise you potentially end up with "ZrOxCly.z H2O". ZrOCl2 appears highly soluble at boiling point...
On several occasions I've reduced the volume of solution by boiling, only to find that the remaining volume had solidified the next day into a wet,
white precipitate that dissolved back upon dilution extremely easily. I assume this really is ZrOCl2.8 H2O. On one occasion I observed needle like
crystals, according to Holleman characteristic of the compound.
I will try and isolate the last batch by filtration, perhaps washing with a non-solvent and careful drying: it is known to lose HCl by drying at too
high temperature.
The other method of crystallising ZrOCl2.8 H2O by adding conc. HCl still doesn't work with 32 %, presumably it needs 40 % for that to work.
In reality I don't really need solid, pure ZrOCl2.8 H2O but a relatively concentrated (and of well known concentration, about 1 M or stronger)
solution of fairly pure ZrOCl2, at pH not much higher than 4. From that, adding a known amount of NH4F or NH4HF2 should precipitate the almost
insoluble (NH4)2ZrF6, the thermal precursor to ZrF4. For this, a method of analysis is needed and Zr isn't easy to determine with traditional methods.
EDTA titration will probably be needed...
[Edited on 6-6-2010 by blogfast25]blogfast25 - 10-6-2010 at 08:58
The more I work with zirconyl chloride, the more it does my head in.
The difference in solubility at BP and RT must be very large: boiling in solutions of it in strong HCl, up to a point where the first cloudiness
appears and leaving the solution to cool and stand overnight crystallises out what is presumably ZrOCl2.8 H2O. This redissolves upon adding water
effortlessly. I think the crystallisation may be near quantitative, if you get it right.
I tried this and it has one very desirable side effect: acetone extracts the FeCl3 contaminant effortlessly! The acetone phase turns slightly
yellow and the crystals turn snow white.
A small amount of these were dried at 70 C for 30 min but to my disappointment the resulting product is only sparsely soluble in water and it seems
some hydrolysis is taking place. After adding a bit of HCl and warming on a steam bath, the solution was filtered and on addition of strong NaOH a
non-amphoteric hydroxide precipitated, hydrated zirconia. But looking at the filter most of the product hadn't dissolved.
I then washed some more crude ZrOCl2 with acetone, let it drip out on the filter for a bit and mixed a small amount (still wet with acetone) of it
with some cold water: it dissolved effortlessly without any hydrolysis into a completely clear solution which precipitated zirconia hydrate on
addition of NaOH. The paper above did mention drying in dessicator: presumably ZrOCl2.8 H2O is seriously sensitive to drying conditions?
The acetone will be recycled: adding dry Na2CO3 should pull the Fe3+ out as Fe2O3 into the watery phase, filtering and redistilling should do the
rest... turd - 10-6-2010 at 09:59
Of course you are aware that "zirconyl" chlorides, bromides and nitrates do not exist. They're zirconium(IV) hydroxides with bridging H2O or OH
ligands.blogfast25 - 10-6-2010 at 11:40
Of course you are aware that "zirconyl" chlorides, bromides and nitrates do not exist. They're zirconium(IV) hydroxides with bridging H2O or OH
ligands.
Yes. It depolymerises in very high [H3O+] concentrations. 1 to 2 N H2SO4 is
recommended for back titrarions of EDTA (with Fe3+), for instance...
It depolymerises in very high [H3O+] concentrations. 1 to 2 N H2SO4 is
recommended for back titrarions of EDTA (with Fe3+), for instance...