[fiorafa]

Hi, first time posting in this forum so sorry if this is the wrong section, i also apologize for any gramatical error as english is not my first language.
My dad owns a mining company where we mine sand, on the process of washing the sand we get clay as the biproduct, some months ago we sent a sample of the clay for analysis to see if the clay was fit for manufacturing ceramics, bellow are the results of the analysis and the composition of the clay:

SiO2	Al2O3	K2O	NA2O	Fe2O3	TiO2	CaO	MgO	Mn2O3	P2O5	SO3
44.88%	35.00%	1.31%	0.02%	4.37%	0.34%	<0.02%	0.12%	0.02%	0.09%	<0.05%

     
From the analysis we learned that our clay is fit for manufacturing ceramics, but after seeing the amount of aluminium oxide present in the clay i started thinking about the viability of producing aluminum, i first wanted to try it at a small scale as a proof of concept (and personal curiosity) and later try aplying it in an industrial scale. I already had a look at the Hall–Héroult process and have some ideas of how to implement it in a hobbyist style, but i am stuck at finding a way of separating the aluminium oxide from the clay. Can someone tell me a process (both on a industrial and hobbyist scale) of achieving that or pointing me some books about it? At any rate thanks everyone for taking the time reading my question.

[billnotgatez]

Cost / benefit analysis for your material versus ores typically used for aluminum production may be appropriate.
Make a silk purse of a sow's ear


It is interesting that your percentages add up to 86.22 not 100

[Enthalpy]

Welcome, Fiorafa! And no worries with your English.

https://en.wikipedia.org/wiki/Bayer_process

35% Al₂O₃ isn't brilliant. Earth's crust contains as a mean 8% Al by mass, equivalent to 15% Al₂O₃, while bauxite contains 30-60% Al₂O₃. But you have a bauxite equivalent without mining it. Then, electricity cost matters. Aluminium producers operate where electricity is cheap, say in Québec, and prefer to transport aluminium ore over the oceans.

To my understanding, most minerals are silicates of aluminium and iron. That is, the oxides of Si, Al and Fe are not separated in them. The analysis result is just a conventional form of expressing element abundances in the form of oxides. So the separation, say by the Bayer process, is a chemical transformation.

The Hall-Héroult process must be seriously difficult to try in hobby style. Interesting challenge. On a small scale, a computer power supply can provide some intensity for little money, at 3.3V, 5V and (cells in series, most power) 12V.

[Corribus]

Many clays are layered in structure and the alumina layer is fused (bonded covalently, with shared oxygens) to the silica layer. You'd have to get rid of the silica layers, and probably reconstitute the alumina. That would be expensive and difficult.