Talk:FFC Cambridge process

Even though I wrote the section on calcium solubility in the melt as opposed to sodium or magnesium, I'm not quite so sure now - does anybody have data? It may be that both sodium and magnesium are quite soluble in their own chloride melts, except they are not strong enough reducers of oxides such as titanium of magnesium to do anything useful. Sillybilly 16:51, 20 June 2006 (UTC)

Disambig of "insulator oxide"

The article says:
"...the electrolysis can be carried out on an insulator oxide, such as zirconia and silica..."
Does this mean oxide materials that are electrically insulating (of which most, if not all, are, including titanium dioxide), or does it mean electrical insulator components used as commercial insulator products?

--Mikiemike (talk) 12:15, 23 February 2008 (UTC)

It means that the oxide which is the feed or precursor in the process is an insulator. It forms a part of the cathode in the form of, for example, a porous pellet attached to a metal current collector. —Preceding unsigned comment added by 86.9.167.205 (talk) 10:38, 4 October 2009 (UTC)

Calcium cannot reduce Yttrium

"There are other difficulties following the “calciothermic reduction” hypothesis. One example is that according to thermodynamic data, Ca cannot reduce Y2O3. However, Y metal was produced by electrolysis of Y2O3 in molten CaCl2 (Wang et al, Angew. Chem. Int. Edit., 45 (2006) 2384-2388.)"

Though I don't have lab measurements or literature references of thermodynamic data, I remember reading that Lanthanum oxide cannot be reduced by metallic calcium, but lanthanum chloride can. If the Y2O3 is slightly dissolved in the chloride into Y ions (whether Y³⁺, or YO⁺, or whatever Y₂O₂²⁺, or whatever highly complex solvated ion forms), the original solid oxide crystalline lattice binding energy will no longer be present.

Still, one issue in general with this method is the oxygen accumulation in the melt, because at the anode chlorine gas evolves. Fluoride melts, as used for aluminum production, do not build up oxygen. One needs to find a way to drop out oxides, such as cooling the molten calcium chloride to a certain temperature and collecting the oxide precipitate, if it forms. Another method might be carbochlorinating a portion of the melt in a separate container, whereby CaO+CaCl2+C+Cl2 --> CaCl2+Cl2(gas)+CO2(gas). The CO2 can then be separated from the Cl2 and the Cl2 recycled. Note that this method uses stoichiometric amount of C for each oxygen, and renders the process into a carbochlorination one, and one might as well carbochlorinate the initial metal oxide, without dissolving it into a molten salt. This is so unless the carbochlorination reaction rate of the original oxide is extremely slow, and justify using a molten salt, which can attack and react with the oxide at a lower temperature. If dropping out calcium oxide from the melt works, then it can simply be cooled, treated with HCl and dried back to the electrolyte material, CaO+CaCl2+HCl--> CaCl2+H2O. Sillybilly (talk) 00:17, 4 October 2009 (UTC)

Actually, the simplest thing to do is to cool a portion of the melt fully to room temperature (without trying to separate out solid CaO at high temperatures), react it HCl, then dry/reheat the whole thing to reintroduce into the melt.Sillybilly (talk) 18:42, 4 October 2009 (UTC)

There is no accumulation of oxygen in the melt

In the FFC Cambridge Process, the overall cathode reaction is that the solid oxide is electro-reduced to the metal and oxygen ion according to the following reaction, using the reduction of SiO₂ as an example.

Cathode: SiO₂ + 4 e = Si + 2 O^2-

The use of molten CaCl₂ is important because this molten salt can dissolve and transport the O^2- ions to the anode to be discharged. The anode reaction depends on the material of the anode. If an anode of graphite is used, a mixture of CO and CO₂ evolves. However, if an inert anode is used, such as that of high-density SnO₂, the discharge of the O^2- ions leads to the evolution of oxygen gas. The anode reactions can thus be expressed as follow.

Carbon anode: x O^2- + C = CO_x + 2x e (x = 1,2)

Inert anode: 2 O^2- = O₂ + 4 e

Because of the continuous anodic discharge of the O^2- ions, no accumulation of oxygen will occur in the melt.

It is important to point out that, due to the hygroscopic nature of CaCl₂, moisture is always present in the salt, which leads to hydrolysis when the salt is heated in air as shown by the following reaction,

CaCl₂ + H₂O = CaO + 2HCl(g)

This hydrolysis reaction means that there is always a small amount of CaO present in the molten salt if it is prepared using the usual thermal treatment in air. This little amount of CaO, likely in the level of much lower than 1 mol%, is beneficial in terms of avoidance of chloride ion discharge in the beginning of the electrolysis. —Preceding unsigned comment added by 86.9.167.205 (talk) 10:30, 4 October 2009 (UTC)

I agree that because of the hydrolysis there is always a small amount of CaO present, but this is minor. Unlike hydrated AlCl3, hydrated CaCl2 can be dried without much loss of chlorine. I was talking about the buildup of oxygen content on prolonged melt use, such as 50% or 90% oxide, and no chloride left. At the anode chlorine gas evolves in large quantitites, not oxygen. If the anode material is selectively reactive with oxygen, such as a graphite anode would be, then it is possible that CO and CO2 evolves, and no chlorine. That is another way to keep the oxygen content low, but then the reaction is MOx + C + electric --> M + CO, and equivalent amount of carbon is consumed for each mole of oxygen.

Tin oxide I did not know about. It's an oxygen deficient semiconductor, that may preferentially absorb oxygen ions compared to chloride ions from the melt, and reduce it inside the electrode, or even the other side of the electrode, similar to high oxygen ion conductive yttria doped zirconia used for high tempereature solid oxide fuel cells. One issue with tin oxide in a chloride environment might be the formation of volatile SnCl4 and maybe SnOCl2, or simply the dissolution of the tin oxide into the melt. You may want to try that yttria doped zirconia from an SOFC(solid oxide fuel cell) that should be more resistant to dissolution than tin oxide. Another idea is that, lacking high chlorine content in the fuel that would corrode/volatilize the tin away, tin dioxide based SOFC membranes might be better/cheaper/more resilient in SOFC applications than zirconia ones. However even automotive oxygen sensors are zirconia based.Sillybilly (talk) 18:42, 4 October 2009 (UTC)

It is impossible for the chloride ions to discharge at the anode if oxygen ions are present in a chloride melt (unless an unreasonably large current or cell voltage is applied). This is because the oxidation potential of chloride ion is more positve than that of oxide ion. In addition, the solubility of oxide ions in the chloride is also limited. For example, in molten CaCl2, the maximum CaO concentration is about 21 mol% in the working temperature range of the FFC Cambridge Process.

As I pointed above, there is no chance for the accummulation of oxygen in the melt during the FFC Cambridge Process, all oxide ions from the reduction of the metal oxide on the cathode will be removed by their discharge at the anode. THIS IS AN ESTABLISHED FACT from all reported experimental results in the past 10 years.

Regarding the selection of suitable anode materials, both SnO2 and yttra stabilised zirconia are currently being investigated, together with others, for application in the FFC Cambridge Process in different laboratories. Satisfactory results have been reported. —Preceding unsigned comment added by 86.9.167.205 (talk) 21:37, 4 October 2009 (UTC)

With an inert nonreactive metallic anode, such as gold or platinum, oxygen should stay in the melt and chlorine gas evolve. Especially when the chloride ion concentration is so high compared to oxygen's. Only fluoride is more electronegative than oxygen. In case the anode is selectively reactive with oxygen, but nonreactive with chlorine, such as the graphite anode is, then carbon monoxide is formed, with anode consumed more per mole of oxygen. Sillybilly (talk) 04:21, 7 October 2009 (UTC)

Sillybilly: Your above assumption has been disproved by a recent paper published by Jiao and Fray (Development of an Inert Anode for Electrowinning in Calcium Chloride-Calcium Oxide Melts, Metall. Mater. Trans. B, 41 (2010)74-79.) In their work, oxygen gas was produced on a ceramic inert anode during electrolysis in molten CaCl2 containing CaO. By the way, your suggestion of using a nonreactive metallic anode is also misleading. All metals will dissovle when used as the anode during electrolysis in molten chloride salts. —Preceding unsigned comment added by 82.5.49.81 (talk) 22:22, 19 May 2010 (UTC)

Was the process known decades ago?

Moved from the article. Comments? Materialscientist (talk) 12:32, 27 March 2011 (UTC)

It seems someone didn't do their homework properly as the same process was already patented in 1904 as german patent 150557. Some more references can be found on page 137 of .

If you read the 1904 German patent carefully, and also the English text in the book "Industrial Electrometallurgy, Including Electrolytic and Electrothermal Processes", you should understand that the 1904 patent did/does not teach you to do the same thing as the FFC Cambridge process. — Preceding unsigned comment added by 94.197.39.193 (talk) 22:25, 17 May 2012 (UTC)

Proposed for use on Moon and Mars

This process has proposed for metal production from regolith on Moon and Mars. Some estimates of the energy requirements, and any consumables like electrodes would be interesting. - Rod57 (talk) 13:49, 12 November 2020 (UTC)

Method for producing Titanium

I made some edits to try to make it clear that the FFC-Cambridge process is specifically a method for producing titanium, but an editor named Smokefoot deleted them. Yes, it has also been proposed that FFC-Cambridge could possibly be used for reduction of other metals (and I did give a reference), but the reason it was invented and pretty much all of the work done on it is as a titanium production method (read the paper: "Direct Electrochemical Reduction of Titanium Dioxide to Titanium in Molten Calcium Chloride").

For some reason the editors of the article want to phrase it as "this is a generic method of reducing metal oxides" rather than "this is a method of reducing titanium that could also be adapted to other metals". I'm not sure why. Nobody would ever use this to make, for example, iron; pretty much the only reason you'd go with a molten-salt reduction route is for hard-to-reduce insoluble refractory metals, and titanium is pretty much the only one that is desired in the quantities needed for implementing this. Skepticalgiraffe (talk) 16:29, 18 April 2021 (UTC)

@Skepticalgiraffe:Most of the content that you contributed is still there, but I apologize if some core info was removed. Some hype was deleted. So far as I can see, the method has been applied only to Ti. It is one of many methods.

Most of the section under "Process" is far too detailed in my opinion. I was going to compress that section drastically, but just lacked the time and energy.

There is nothing wrong with a sober report that just states facts. We are not here to sell something, which is that tone that I was trying to rip out of this thing. --Smokefoot (talk) 19:11, 18 April 2021 (UTC)

I'm not sure what in the world you think was "hype". In removing the "hype" you also removed all reference in the article to the fact that it is a process for producing titanium.

Maybe the easiest soultion is to simply rewrite the lede to remove the statement that it is a generic process for reducing "metal oxides". (I could add a sentence saying that it has been also been proposed for other metals, but since you deleted the only reference in the article that said this, if I did I would have to add a "citation needed" tag.) Skepticalgiraffe (talk) 00:07, 19 April 2021 (UTC)