clearly_not_atara
International Hazard
Posts: 2926
Registered: 3-11-2013
Member Is Offline
Mood: Big
|
|
Novel titanium refining process
Due to the extreme conditions used in this process (hot hydrogen fluoride) and several gas-phase steps requiring a complex apparatus, I don't expect
to try this at home, but it's a puzzle I've thought about for several years.
Titanium refining has some problems:
- direct electrolytic reduction of titanium oxide usually gives metal with some dissolved oxygen, which tends to be brittle and basically useless (the
FFC Cambridge process has yet to produce industrial titanium for this reason)
- titanium (IV) chloride is a covalent, volatile, toxic compound
So the proposed process has two steps:
- conversion of titanium ores to titanium (III) fluoride by precipitation of NH4TiF4 and its pyrolysis (basically known reactions)
- electrolytic reduction of TiF3 with a sacrificial tungsten anode (new)
The reduction of TiO2 to Ti2O3 with hydrogen (inter alia, [1]) or by carbothermal methods [2] has been documented many times. Unfortunately, it's hard
to find good examples because they're all complaining about the failure to produce Ti metal directly this way. Alternatively, waste titanium can be
used to reduce TiO2, itself being hard to use directly due to dissolved oxygen.
But owing to the solubility of titania in fluoride solution, the direct leaching of titanium ores with hydrogen fluoride has been proposed as well
[3]. The direct reduction of the solution with hydrogen is not described, but it is energetically favorable if we suppose that the relevant ionic
species are TiO(2+), H2/H+ and Ti(3+). Usually titania is solubilized using ammonium fluoride or bifluoride, due to the difficulty of working with
hydrofluoric acid.
Once titanium is in the form of titanium (III), there is a known process for producing anhydrous NH4TiF4 by precipitation from hydrofluoric acid [4].
This is the key step, because oxygen and iron impurities significantly degrade the properties of the product, and they are removed in this step.
NH4TiF4 decomposes under heating to give TiF3.
Once TiF3 is produced, it can be dissolved in molten fluoride salts and electrolysed. According to the standard electrode potentials, the oxidation
Ti3+ >> Ti4+ requires more energy per electron than the oxidation of W >> W6+. So at a tungsten anode, WF6 is produced and under these
conditions is extremely volatile. While not a particularly friendly compound, WF6 is still preferable to fluorine gas, which will corrode any anode
and is generally avoided whenever possible. The cathode is probably a piece of titanium wire, which accumulates oxygen-free titanium from the melt.
WF6 in turn is used (today) primarily for tungsten plating by the decomposition of a mixture of WF6 and H2 gases onto a hot surface, a very old
process [5], which is proposed to regenerate the anode.
One possibility is that the anode could comprise most of the reaction vessel for electrolysis, which would reduce the needed rate of deposition. If
the tank has a surface area of one square meter, several pounds of tungsten are deposited per mm of thickness. But this might not be practical.
Alternatively, the anode is an inert wire, which is coated in tungsten, then used for electrolysis, and then recycled by tungsten vapor deposition.
So effectively, all of the intermediates are recycled and the net reaction is just TiO2 + 2 H2 >> Ti + 2 H2O. Oxygen is rigorously excluded from
the product.
The downside is the use of lots of hot corrosive gas. But the competition is all based on TiCl4, using a carbothermal reduction in the presence of
chlorine. So we're not much worse. But I wonder if Sciencemadness could suggest some improvements.
1: https://link.springer.com/article/10.1023/A:1015378931111
2: https://www.sciencedirect.com/science/article/pii/S027288421...
3: https://www.scientific.net/SSP.265.542
4: https://www.jstage.jst.go.jp/article/nikkashi1972/1973/1/197...
5: https://iopscience.iop.org/article/10.1149/1.2426649/meta
|
|
|
j_sum1
Administrator
Posts: 6472
Registered: 4-10-2014
Location: Limbo
Member Is Offline
Mood: Just got through yet another "take this job and shove it" moment.
|
|
You just remjnded me of why we don't use titanium more often.
It is such a lovely metal frkm an engjneering standpoint. But acquisition is another matter.
|
|
|
nux vomica
Hazard to Others
Posts: 268
Registered: 18-7-2013
Member Is Offline
Mood: No Mood
|
|
You should look at the info about the TiRO™ Process .
https://www.coogeetitanium.com.au/
[Edited on 10-6-2025 by nux vomica]
|
|
|
Radiums Lab
Hazard to Others
Posts: 299
Registered: 18-3-2025
Location: India
Member Is Offline
Mood: Experiencing the elegance of science.
|
|
Good write up, and the process seems doable on industrial scale.
[Edited on 10-6-2025 by Radiums Lab]
Water is dangerous if you don't know how to handle it, elemental fluorine (F₂) on the other hand is pretty tame if you know what you are doing.
|
|
|
Radiums Lab
Hazard to Others
Posts: 299
Registered: 18-3-2025
Location: India
Member Is Offline
Mood: Experiencing the elegance of science.
|
|
But HF is pretty bad (even industrally).
Water is dangerous if you don't know how to handle it, elemental fluorine (F₂) on the other hand is pretty tame if you know what you are doing.
|
|
|
grabau
Harmless
Posts: 1
Registered: 17-10-2025
Member Is Offline
|
|
Would it work? Probably. It is worth noting that the idea of a volatile sacrificial anode is only something I have seen a few times in the literature.
I imagine you suggested a titanium (III) fluoride route instead of titanium (IV) fluoride one due to the higher volatility of TiF4 compared to TiF3,
and the ability to use a higher temperature for molten salt electrolysis? (Aside: the latter was attempted at Albany Research Center, and called the
'Albany Titanium Process', but using metallothermic reduction instead of electrolysis; see the article "Low Cost Titanium - Myth or Reality?" which is
widely available)
That said, there are some fundamental challenges with the proposed approach.
First, TiF3 has a b.p. well below the Ti m.p. This doesn't kill the process, but it lowers the throughput if you cannot tap molten. If you instead
collect solid Ti, and separate it from residual electrolyte by washing, you're going to oxidize some Ti to TiO2 (refer to Hunter's 1910 paper, where
he replicates the sodiothermic reduction of Na2TiF4 or K2TiF6. I suspect Ti(III) will oxidize here, introducing losses). The Kroll process gets
around this via the volatility of MgCl2. (Another aside: if you're looking at metals that can both reduce titanium fluorides, and the metals and their
fluorides themselves have b.p. at or near Ti m.p., you're much limited to the rare earths, refer to, e.g., the research coming out of Okabe's group,
or calcium [b.p. is lower but difference in stability allows its use]).
Second, lots of operations compared to exiting processes, and, moreover, two separate electrolysis operations. Electrolysis is expensive from a
capital perspective, and electricity is also expensive.
I do think that any future titanium process is going to be metallothermic reduction of the oxide or metallothermic reduction of the oxide at
temperatures that allow tapping of molten titanium.
|
|
|
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
