Sciencemadness Discussion Board - Thermite from Sand : make Silicon

Thermite from Sand : make Silicon

metalresearcher - 30-10-2011 at 10:14

I found this video http://www.youtube.com/watch?v=73YmP_JSrlU on which the guy showed fine sand (SiO2) , aluminum poweder and *sulfur*. The S was added for reason of lowering the ignition temp.

Is the sulfur really needed ? I have it but I think the Al2S3 and Al2O3 will contaminate and mess up the resulting Si 'metal' regulus.

He also shows making manganese from Al + MnO2 .

Does somebody have experience with this ?

hkparker - 30-10-2011 at 11:37

Everything needs to be extremely fine and the Sulfur is a must. The reactions will not get hot enough nor will it sustain without it. The impurities can be removed later.

Neil - 30-10-2011 at 13:06

The sulphides are easy to remove with a soak in acid or water. Getting Mn is very hard, most of it vaporises and you end up with punitive yields.

chlorates and sulphates can be used in place of sulphur.

blogfast25 - 30-10-2011 at 13:07

Quote: Originally posted by metalresearcher

The S was added for reason of lowering the ignition temp.

Is the sulfur really needed ? I have it but I think the Al2S3 and Al2O3 will contaminate and mess up the resulting Si 'metal' regulus.

He also shows making manganese from Al + MnO2 .

Does somebody have experience with this ?

Yes, the sulphur is indispensable but has nothing to do with your stated reason.

The reaction SiO2 + 4/3 Al === > Si + 2/3 Al2O3 (1)

… does by itself generate almost no heat. To obtain something useful the reaction needs to generate enough heat so that at the end of the reaction both the alumina and the Si metal are in the molten state and the two can then neatly separate from each other.

In the preparation of ‘technical Si’, this is achieved by adding extra aluminium and sulphur so that 2 Al + 3 S === > Al2S3 (2),

... which generates loads of heat. Reactions (1) and (2) run simultaneously from a homogeneous mixture and the result is a mixture of molten Si, molten Al2O3 and molten Al2S3, from which the metal then coalesces out. I’ve done this a lot. It stinks because the Al2S3 readily hydrolyses: Al2S3 + 6 H2O === > 2 Al(OH)3 + 3 H2S.

Manganese thermites are hard because of Mn’s low boiling point, which causes Mn metal to become evaporated, thus giving quite low yields. MnO2/Al mixtures can behave as flash powders…

metalresearcher - 30-10-2011 at 13:34

In the meantime I already tried the silicon thermite successfully just a small amount (3 grams) in the fumehood. I could find the Si particles in the (H2S stinking) remains.

http://www.youtube.com/watch?v=1pyaRTc2bKM

blogfast25 - 31-10-2011 at 03:05

@metalresearcher:

Fairly decent result for a first go and nice video. Try and isolate some of the suspected Si metal(loid) and dunk it in strong NaOH or KOH. It should dissolve quite easiliy with hydrogen evolution, evidence of Si...

Larger batches tend to give better metal/slag segregation. Your mix also needs more sulfur/Al I think, because your slag was porous (not really fully molten). The mix should really reach white heat for good results.

Edit: actually you seem to have a bit too much sulfur in there, for the amount of excess Al; 0.9 g SiO2 requires 0.54 g Al (acc. (1)) and the excess Al (0.9 - 0.54 = 0.36 g) requires 0.64 g S, not 1.2 g.

[Edited on 31-10-2011 by blogfast25]

[Edited on 31-10-2011 by blogfast25]

MrHomeScientist - 31-10-2011 at 08:00

That was actually my video

All the above comments are correct - the sulfur is there just to react with extra aluminum to provide the heat necessary to drive the reaction and produce molten products. Larger batches tend to do better when trying to recover the product, because the whole thing can heat up to the right temperature and stays hot for long enough to give good separation of the slag.

Another thing I should have done was treat my sand to purify it. Washing with water and then acid a couple times would help remove water-solubles and get rid of seashell fragments (CaCO3) and other detritus. Looks like you took care of that with a purer source of sand.

I really like your reaction 'vessel' too! Every time I do a thermite I destroy a ceramic flower pot, so it'd be very nice to have something reusable. I'll have to look into making one of those for myself.

As for Mn, yes it's very hard to get any metal from this type of reaction. Mn thermites are pretty energetic in my experience so it throws particles out of the container, plus much of the metal produced boils off. I've only ever gotten a few small nuggets of Mn from any of my tries. My source of MnO2 isn't the greatest though (battery paste and eBay garbage that was cut with sand).

blogfast25 - 31-10-2011 at 08:16

Well, good vid, MrHomeScientist. Now I know where you live (in a place that speaks my mother tongue!)

Crucibles: nah, stick with pottery pots, old ceramic teacups, egg cups and such like ; their shape is much better suited to collecting the molten slag in a nice puddle. Shallow cavities don't do that: V-shape always trumps.

Yes to desalinating/decarbonating any beach sand.

For MnO2, try adding a good dollop of either fluorite or cryolite, about 10 - 20 % of the total mix. That cools things down a bit (they act as a heat sink) and improve flowability of the alumina. Also, avoid the finest grades of ingredients: slightly slower burning will give better yield...

metalresearcher - 31-10-2011 at 09:39

Next time I drill a hole in the refractory, put the thermite inro it and cover it with a lid with a hole with an ignition through it, e.g. Mg ribbon or KClO3 with sugar wrapped in cigarette paper as a fuse. Of course, the lid lies loosely on the reaction vessel so the violent reaction can find a way out. This keeps the heat inside.

Can Mg drillings also be used ? Here a sacrifical anode in which I drilled holes with an 8mm bit and saved the drillings. I just ordered fine Mg powder.
Can the drillings be used with SiO2 or MnO2 (possibly with sulfur) ?

[Edited on 2011-10-31 by metalresearcher]

blogfast25 - 31-10-2011 at 12:08

Mg isn't suitable here because of the insanely high melting point of MgO (2852 C).

Never combine sulfur with any energetic thermite like MnO2: several are near-flashpowders as it is!

metalresearcher - 31-10-2011 at 12:15

Here I tried 3.1g MnO2 with 1.2g Al. It ignited with Mg ribbon: KClO3 + sugar wrapped in a cigarette paper did not ignite it. Strange.
It did react but how to extract the Mn metal is unclear to me.
BTW the 'low boiling point' of Mn is actually 2060oC, way above the reaction temperature of the Mn thermite.

<iframe sandbox width="420" height="315" src="http://www.youtube.com/embed/Nzr2JIpdF4M" frameborder="0" allowfullscreen></iframe>

EDIT: I took a pic of the reacted stuff: did it really react ??
It is still powdered but colored brown.

[Edited on 2011-10-31 by metalresearcher]

Neil - 31-10-2011 at 13:03

That powder is your Mn, it burned above the thermite producing the blinding flame you surly noticed.

Consider that Al2O3 melts in a MnO2 thermite, last time I checked alumina melted at a high temperature then Mn boiled.

Using Mg in a thermite gets an additional strike against it, in addition to Blogfasts point on the melting point of MgO. The low boiling point of Mg means it can vaporize from a thermite reaction while aluminium has a much higher boiling point and is unable to exit the reaction.

blogfast25 - 1-11-2011 at 03:15

Quote: Originally posted by metalresearcher

It did react but how to extract the Mn metal is unclear to me.
BTW the 'low boiling point' of Mn is actually 2060oC, way above the reaction temperature of the Mn thermite.

Well… NOOO!!!

If your thermite mixture is well designed and does what it says on the tin then the reaction runs MUCH HOTTER than 2060 C. That is actually a REQUIREMENT if you want to obtain good quality metal: the reaction mix has to develop so much energy, say ΔH, that the end temperature of the reaction product mix exceeds the melting point of alumina (2072 C), because if the mixture doesn’t end up completely molten the metal CAN NEVER separate from the slag and you end up with a half-sintered mass of alumina/metal particles which is totally useless. Good thermite mixtures produce a neat slag puddle of completely molten alumina with the molten metal sitting at the bottom of your crucible, nicely protected from air by the lsag. The whole thing then cools down.

In the case specifically of manganese that creates a contradiction: you need Tend > 2072 C (to obtain molten Al2O3) but that’s above the boiling point of Mn!!!

BTW, these mini mixes you’re doing will never lead to good metal either: you need at least 20 g of mixture to be able to get anything and you need to use small cups, crucibles or plant pots, smooth surfaced and V-shaped. Covering the reaction like you do will not save considerable energy (some of it simply gets absorbed by the weighty assembly which acts as a giant heat sink!)

What grade of MnO2 are you using?

Judging by the last photo you may well have some mini-nuggets of Mn in there: react any suspected metal with very dilute HCl: Mn is extremely reactive, then test the solution for typical Mn reactions.

Hard to judge but I think a can see some unreacted MnO2 in there: if I’m correct about that it points to the mixture deflagrating during reaction (NOT GOOD).

[Edited on 1-11-2011 by blogfast25]

Neil - 1-11-2011 at 04:00

To expand a bit,

Adding a flux such as CaF2 raises the burn rate and decreases the melting point of Al2O3 which allows for a better burn. Used in small amounts it seems to be able to raise the upper temperature limit but in larger amounts it decreases the maximum temperature by heat sinking the reaction.

Chlorides have a similar effect though not as pronounced - any alkali-halide salts work with Al based thermite mixtures as catalysts AFAIK.

Here is a nifty PDF that covers this nicely http://www.sciencemadness.org/talk/viewthread.php?tid=10981&...

Going a bit further, a loosely packed thermite will burn much faster then a hard packed thermite. A loose thermite will also have much more oxygen entrained in it, then a hard packed thermite, and be much more likely to throw partials out of the reaction vessel.

MY own experience with getting Mn is flux it well, I used NaCl and CaF2 (the NaCl would boil out during the reaction) and pack it into the crucible hard enough to insure it is one homogeneous nearly air free mass.

Find phase diagrams for Al2O3-CaF2 systems and dream of cryolite.

blogfast25 - 1-11-2011 at 04:42

Neil:

I could be wrong on this but I can’t see CaF2 increase the burn rate. Right from the off, it makes the thermite run slightly cooler, due to heat sink effect. And cooler means slower. It does form a eutectic with alumina (the Al2O3/CaF2 phase diagram should not be hard to find - I studied it a few years ago online) and resulting higher slag fluididty might have that effect. But I’d really like to see conclusive proof of that [increased burn rates].

As regards CaCl2, I’m truly sceptical: with a BP of 1,935 C this most likely gets blown off completely: you really need at least 2,500 C end temp. for good results (except for Mn). I use CaF2 only, quite a high dose and always maintain the same calculated Al2O3/CaF2 molar ratio.

I can’t see a *.pdf, only a *.rar (WTF??).

Cryolite: I can sell you some but shipping UK to US (assuming that’s where you are) maybe prohibitive. I also believe that basically there’s nothing CaF2 can’t do that Na3AlF6 can do.

You are right about the packed/loose thermites. I never ‘ram’ mine.

The Mn thermite is something I flogged to death: I tried various Mn oxides: MnO2, MnO, MnO/MnO2 blends: all largely to no avail, yield never exceeded about 30 %...

Edit:

Aaargh! I spoke too soon about the CaF2/Al2O3 phase diagram, it was the Al2O3 - cryolite phase diagram someone digged up for me some time ago:

http://www.crct.polymtl.ca/fact/documentation/FThall/FThall_...

And:

http://www.crct.polymtl.ca/fact/documentation/FThall/Na3AlF6...

Going through my lab notes I systematically use a ratio of Al/CaF2 of 72/47 (w/w), which obviously means there’s always the same amount of moles of CaF2 per mole of Al2O3.

[Edited on 1-11-2011 by blogfast25]

Neil - 1-11-2011 at 06:59

Hey Gert, the RAR is a zip because the PDF was too large to upload directly.

Quoting from
"Thermite reactions: their utilization in the synthesis and processing of materials "

"Addition of salts of alkali metals (e.g. NaF, KF,
NaC1 and KC1), alkaline earth metals (e.g. A1F3,
MgF2), and cryolite (NaA1F6) can effectively increase
the mass combustion rate of a thermite mixture as shown in Fig. 6 [100]. The effect of salt addition is
most prominent with the addition of small amounts,
and the highest combustion rate was found in the
compositions containing aluminium fluoride (A1F3)
and cryolite (NaA1F). It is proposed that salt ad-
ditives reduce the temperature at which the reaction
between the oxide and the aluminium commences.
The oxide film on the aluminium particle, which acts
as a barrier to the interaction, can be disintegrated by
the alkali metal or alkaline earth metal salts at a tem-
perature significantly lower than the ignition temper-
ature of the thermite, and consequently, the ignition
temperature of the thermite mixture with salt addition
is notably reduced [100]"

Here is a TXT file sans graphs. If you can not un-zip the RAR let me know and I can email you a copy if you U2U me your address.

Attachment: Thermite-Reactions-their-Utilization-in-the-Synthesis-and-Processing-of-Materials.txt (72kB)
This file has been downloaded 1474 times

Edited to remove smiley faces which replaced some of the chemical txt.

[Edited on 1-11-2011 by Neil]

Neil - 1-11-2011 at 07:05

From the mouth of the horse:
Attachment: Thermit Journal.txt (39kB)
This file has been downloaded 500 times

No revelations but lots of good base knowledge for planning aluminothermic reductions

And also:

Attachment: thermite_kinetic_breakdown.pdf (1006kB)
This file has been downloaded 591 times

Thanks for the offer of cryolite but right now I have my sights on carbothermic iron stuff.

metalresearcher - 1-11-2011 at 09:54

According to what I saw yesterday is that the result is incomplete, i.e. not all MnO2 has reacted. Maybe the MnO2 is impure, it is from batteries and might contain KOH (electrolyte).
@Neil:

Quote:

Chlorides have a similar effect though not as pronounced - any alkali-halide salts work with Al based thermite mixtures as catalysts AFAIK.

The problem is that virtually all alkali metal salts (including Na2O) have a boiling point below 2000oC so they cool (and slow) down the reaction. Low boiling point blendings have the same effect as water added to reaction mixtures which react well above 100oC.
You say, small amount do not work as efficient. Why ? When the stoiciometric ratio is OK the amount does not matter, if it are tons or milligrams.

EDIT: I found this:

http://developing-your-web-presence.blogspot.com/2008/07/man...

How can I get CaF2 ? On eBay I can only find pills.

[Edited on 2011-11-1 by metalresearcher]

Neil - 1-11-2011 at 10:33

In order for a good slag/metal separation you need your thermite to remain liquid long enough for the alumina and metal to separate.

If you were bringing your crucible to heat and then igniting the thermite as Dr. Goldschmidt did with his Cr thermites, then most likely the size would not matter.

Then again if you heat a crucible of MnO2/Al you may blow your crucible up so that hardly seems like a good solution.

As it is, a thermite in a cold crucible needs to burn hot enough and long enough to heat up the crucible and the reaction products to affect slag coverage of molten metal and separation of oxides from the metallic reaction products. There is a certain critical mass where you can achieve this, think surface area to volume.

In the case of MnO2/Al you not only need to heat the reaction and crucible hot enough for the products to remain molten but you also have to cross your fingers that enough Mn condenses and remains un-oxidised to give you some sort of yield. It follows that the more you react the better your chances.

If you take your brown powder and sprinkle it into a blowtorch flame do you get white streaks of burning Al? If so you had your reaction mixture blow apart before it was fully reacted. If you only get a few sparks you had likely had proper combustion but your Mn vaporized and was re-oxidized.

Did you refine your Mn? If not the KOH is the least of your worries, The graphite in batteries is something that must be removed as C poisons thermite reactions with veracity. The formation of aluminium oxycarbides and carbon dioxide readily interrupt the thermite reaction.

The Mn/O ratio will also be off if the battery was used – there are some threads on this, I can't remember them off the top of my head...

Quote:

The problem is that virtually all alkali metal salts (including Na2O) have a boiling point below 2000oC so they cool (and slow) down the reaction. Low boiling point blendings have the same effect as water added to reaction mixtures which react well above 100oC.

I'm not sure I understand your query, try reading through the papers I linked to. I specified halide salts because they are the only ones I've read about interacting with thermites. My suspicion is that they vaporize and interact with the oxide coating of the Al particles such that they depassivate the Al.

Adding water to a thermite has nothing similar with adding a stable salt, water will oxidise the Al and any formed metal while producing a violent eruption of steam and hydrogen.

Originally I used NaCl in a thermite in an attempt to provide the reaction with a inert gas cover and to cool the reaction while fluxing the Al2O3. Given the MnO2 thermite runs way to hot for the Mn to form a health regulus; cooling the reaction, provided you flux the Al2O3, makes perfect sense. I believe it was Gert who first thought of using CaF2...?

To get CaF2 try searching for fluorite. Find a Ebay pottery shop and get some MnO2 from them while you are at it.

As a cool bonus I ran some MnO2 thermites with SiO4 dust as a heat sink, the sand around the reaction area was left highly hydrophobic by the SiO2 vapours

blogfast25 - 1-11-2011 at 10:34

Neil:

Very interesting but a tad old. Modern texts I’ve consulted no longer mention low boiling fluxes. Will print off nonetheless.

Metalresearcher:

Did you at least purify the MnO2??? Battery crud contains up to 40 % (and even more) graphite (as conducto). Also KOH (for alkaline batteries), ZnCl2 and NH4Cl (for non alkaline batteries), usually also small amounts of ferric oxide. W/o purification battery crud is virtually unusable for thermites.

That article you linked to is mine, BTW…

CaF2 (fluorite or fluorspar): cheap offcuts from mineral shops, some pottery supply businesses sell it as powder, eBay usually sells ball milled CaF2 and other cheap fluorite based ‘jewelry’. Fluorite grinds down easily in a granite mortar and pestle.

Neil - 1-11-2011 at 10:38

@Blogfast- It is old but I haven't seen anything which discounts the data or anything recent which displaces it, if you have any more recent articles I'd be very interested in seeing them.

metalresearcher - 1-11-2011 at 11:24

I found this might be a good and cheap source:

http://www.ebay.co.uk/itm/500g-high-grade-Manganese-Dioxide-...

Two D cell batteries cost more.

blogfast25 - 1-11-2011 at 12:37

Incredible value for money. I know this seller and he's good. Buy forth and multiply!

metalresearcher - 1-11-2011 at 12:41

Just ordered.