Sciencemadness Discussion Board

Exotic thermites & analogs

 Pages:  1  2    4    6

Mardec - 5-2-2008 at 06:22


I fucking hate thermite. It is sort of a urban myth, must be.

I specially got new Fe2O3 and half a kg of german dark to make good thermite.

Make 200 grams, lit perfectly and burned pretty fast. But then I went to look.. It even didn't go trough a coocky can.. pathethic..

What am I doing wrong?!

Bert - 5-2-2008 at 08:22

For a start, you should use coarser Aluminum to slow the reaction enough so it doesn't just blow the molten Iron around the area, rather letting it settle to the bottom of the reaction. You also need to surround the mix with a sand dam or clay flower pot or some other refractory to channel the molten Iron to where you wish to perform your work- Be it welding or melting a hole.

Thermite works for some purposes. It's nothing like the movies, however.

[Edited on by Bert]

Neil - 19-3-2008 at 16:40

Originally posted by Mardec


I hate when something doesn't work...

I made TH-3 thermite today, the same shit the military uses. I lit perfectly like it should. Burned very intense. Left a big pile of molten metal (Iron probably).

But it didn't get through a small (empty) gas container for my bunsen burner.. The iron stayed on top...

I don't get it, why didn't it go through it? Everything in the facinaty was charred.

I used this comp:

-12 g Fe2O3
-4 g Al
-6 g Ba(NO3)2
-0,2 g sulpher.

I lit it with 50/50 kno3/mg and the whole charge was packed in a 1,5 cm ID carbboard tube with the burning side directly aimed against the bottom of the gas container.

And it didn't even made a small hole.

The container was emty yes, I am not a noob/stupid.

It was one like this:

What are these things?! there was an aluminum can holding the tube in place, that completly vanished.

[Edited on 19-9-2007 by Mardec]

The idea of barium and sulfur is to cool the slag/metal and produce flame damage in military thermate. straight thermite produces a hotter slag/metal and confined, directed thermite produces the most effective cut.

if you had a refractory and pressure resistant container themate can be packed in and vented out of a single hole in a jet so that it produces a cutting jet which has moderate penetrating powers.

the card board tube would've sucked out alotta heat too...

Try CaSO4 or better MgSO4... both need to be dehydrated... MgSO4 puts any thermite I've run across to shame.

Neil - 26-3-2008 at 04:32

I'm wondering, has anyone here ever tried to use thermite to reduce zirconium silicate ?

3ZrSiO4 + 8Al + Boost ---> 3SiZr + 4Al2O3 + Crude from boost

Thermite issues

Epew23 - 26-3-2008 at 19:23

Nothing like a good physics class on sat. lol.

I offered to show my physics teacher what thermite can do this up coming sat. I was looking through this book that i bought about all sorts of chemicals and how to make them. Now i have Al And FeO2 but they are both 300 mesh, and the book says that i need 2 g of Iron oxide for every 1 g of aluminum powder that are 400 mesh? will that change the density enough to make it so that i will have a different reachtion???

and how much should i make?? enought to scare the CRAP! out of him?!?!?! lol :D

crazyboy - 26-3-2008 at 19:43

300 mesh will work fine just mix them well. You will also need something to light them like Mg ribbon or potassium permanganate and glycerine.

About how much to make its up to you I would stay around 10-15g or less

The_Davster - 26-3-2008 at 19:50

1g of 400 mesh is the same as 1g of 200 mesh. The only difference will be the burn rate, and volume of the mixture.
Small mesh metal powders sometimes do not give the classic thermite display as they burn too fast leaving no slag.

Also it is Fe2O3

egloskerry - 24-5-2008 at 12:27

I just want to make sure I understand everything correctly. I bought a couple oxides which I hoped would work, but didn't. I wasn't surprised by MgO not working, but CaO didn't, and neither did La2O3 or CeO2. What is the reason they didn't work? What are the calculations needed to determine if a thermite reaction will proceed?


Jor - 28-5-2008 at 16:05

So whos gonna try a HgO thermite? :D:cool:

Zinc - 29-5-2008 at 01:30

I tried it around 2 years ago. It was a small amount of red HgO mixed with Mg. It burned quite fas. All the Hg was vaporised.

[Edited on 29-5-2008 by Zinc]

egloskerry - 9-6-2008 at 18:18

No one's gonna help me out here?

indigofuzzy - 10-6-2008 at 08:24

egloskerry, aren't all of those metals whose oxides you've tried more reactive than aluminum? I'm pretty sure the oxide has to be an oxide of a metal that's less reactive than aluminum.

natriumperoxid - 11-6-2008 at 07:25

Imho all the oxides listed above will not react at room temperature (with Al, no matter how high the activation energy).
An Ellingham diagram can quickly tell you if the oxide involved will be reduced at room temperature (by Al or any other metal) / what temperatures are required in order to make the reaction occur. 2 Mg + O2 --> 2 MgO generally lies above 4 Al + 3 O2 --> 2 Al2O3, however they cross somewhere around 1500K (not too sure about that one, the point is: very hot).
But careful, don't confuse that with the activation energy! The above would mean that Al is capable of reducing MgO at temperatures above 1500 K. Aside from that, a certain activation energy is needed.

[Edited on 11-6-2008 by natriumperoxid]

egloskerry - 11-6-2008 at 20:05

Where would the activation energy be found? Is that what determines if the reaction will proceed? I'm assuming what needs to be found is the temperature at which the reaction will occur.

natriumperoxid - 12-6-2008 at 03:22

My post should give you exactly this information,
for example an Ellingham diagram of the involved compounds will tell you that the following reaction: 3 MgO + 2 Al --> Al2O3 + 3 Mg will only occur at temperatures of roughly 2000 K and above. On top of that, a certain activation energy is needed in order to initiate the reaction.

Ellingham diagram for dummies: if the line of the oxide involved (e.g. 2 Cu + O2 --> CuO) lies ABOVE the line of the reducing metal involved (e.g. 4 Al + 3 O2 --> 2 Al2O3), a reation will occur! This is the case for the above, therefore it is a suitable thermite reaction. However, a certain activation energy is needed, this is the ignition mix such as a sparkler or permanganate + glycerin.

egloskerry - 12-6-2008 at 04:53

Alright, I think I can figure it out. I'll return if I need more help.

That's odd, though. The diagram says ZnO should react, but I've never been able to get it to do anything.

[Edited on 12-6-2008 by egloskerry]

natriumperoxid - 12-6-2008 at 10:29

ZnO should work without a problem, you probably did not provide the necessary activation energy. Some thermites are very hard to ignite and may even stop reacting again (e.g. SiO2 - thermite).

What ignition method did you use? It's probobably worth trying a more "brutal" one.

It is important to note that the distance between the lines does not indicate how exothermic the reaction will be or how rapidly it will progress.

egloskerry - 13-6-2008 at 11:33

I remember using a sparkler as well as KMnO4/Glycerin. I'll try it again, using some Fe3O4 thermite to initiate it.

Useful thermodynamic data on thermites and intermetallics

chemoleo - 16-6-2008 at 15:22

It's a 'must' resource, and it answers several questions above.

This has been posted before, but I think these posts were deleted.

Attachment: A survey of combustible metals, thermites, and intermetallic.pdf (1.2MB)
This file has been downloaded 931 times

497 - 19-6-2008 at 00:14

That is a great reference chemoleo!

The only thing i think it is lacking is information on the speed of the thermite reactions, m/s or something.

I have to say I like the Mg + B2O3, at 2134 cal/g that is pretty amazing. I wonder how fast it would be? I like how easy it is to make though, borax + acid + alot of heat should do it.

Also I wonder how useful some of these might be for producing metals from oxides, like B, Ti, Si, Mn, etc.

not_important - 19-6-2008 at 00:40

Originally posted by 497
I have to say I like the Mg + B2O3, at 2134 cal/g that is pretty amazing. I wonder how fast it would be? I like how easy it is to make though, borax + acid + alot of heat should do it.

Also I wonder how useful some of these might be for producing metals from oxides, like B, Ti, Si, Mn, etc.

I posted this a few dats ago on the MW NaOAc thread
In Inorganic Synthesis V2 is a procedure for producing porous boron oxide. The normal method is to fuse boric acid, at the tepid temperature of 600 to 1000 C; then chilling it to a hard, difficult to powder glass. The alternative method in I.S.v2 is to pull a vacuum on the boric acid, then slowly raise the temperature to 200 C; this results in a lightly sintered porous product that rather energetically rehydrates. On a large scale a multi-stage fluid bed dryer can be used.

Note that powdered B2O3 picks up water from the air fairly readily.

As for other metals, search for "Goldschmidt process" or "Goldschmidt reaction".

497 - 20-6-2008 at 15:27

You are right, the high powered hygroscopicity(?) of B2O3 would pose a major problem for the practical use of a thermite that contains it. Its almost as strong as concentrated sulfuric IIRC. Still I think it would be worth the trouble for a whole 2134 calories per gram! Some of the I2O5 compositions looked pretty amazing too, but I'm not sure I'd want to mess with clouds of iodine and such...

chloric1 - 20-6-2008 at 16:19

NOt to mention iodine pentoxide is a REAL oxidizer compared to the other oxides so I would shy away from this too. Unless it made up less than 20% of a composition and you increase reaction rate and get a purple cloud as a bonus!:P:D

ducksan - 30-6-2008 at 05:14

The most powerful thermite I could think of would involve a noble metal oxide or fluoride...Gold(III) oxide would certainly be nasty. Expensive, but explosive.
Au(3+) is an extremely powerful oxidizer. Standard reduction potential (to Au metal) is about +1.6 V, if I remember correctly.

Manganese thermite using Mn2O3/MnO blends

blogfast25 - 23-7-2008 at 08:01

Making manganese metal by thermite reduction of manganese oxides is one of the more frustrating oxide reductions I've ever carried out.

To get a better understanding of what the issues are I'd firstly recommended reading my blog post on this subject.

The issue with manganese thermite, especially from the higher oxides MnO2 and Mn3O4, is that these reactions generate so much heat that in adiabatic conditions the temperature of the reaction products will exceed the boiling point (BP) of Mn (2,061 C, 2,334 K). It's this characteristic that causes MnO2 thermites to deflagrate often violently, almost 'explosively'. The end-temperature of a thermite reaction (in adiabatic conditions) can be estimated quite accurately from the molar reaction heat (ΔH, reaction enthalpy), and the molar heat capacities and molar heats of fusion of the reaction products, basically by applying [url='s_Law]Hess's Law[/url]. If of interest to readers I can give an example of such a calculation (on request).

The obvious solution to the problem would be to create conditions in which the reaction enthalpy is either lower or partly dissipates away (non-adiabatic conditions) but there's another problem complicating this approach.

For a thermite reaction to yield liquid metal, separated from the liquid slag (both solidify on cooling of course), the melting point (MP) of both (whichever is highest) the produced metal and the by-product alumina (Al2O3) has to be reached at the end of the reaction. For alumina the MP is 2,327 K (2,054 C), for Mn, 1519 K. But as stated before, the BP of Mn is also only 2,061 C, perilously close to the MP of alumina.

This creates a real lose-lose situation: to achieve metal-slag separation of Mn/Al2O3 this mixture has to reach a temperature that's really close to the BP of manganese metal. Even higher temperatures will cause much of the Mn metal to simply boil off. At temperatures somewhat below the MP of alumina, the vapour pressure of the Mn will be less but metal/slag separation will not be able to occur, resulting in powdered, sintered metal frozen in the slag.

Accurate control of the end-temperature in near-adiabatic conditions is therefore essential to obtain any lump metal from such a reduction.

I therefore started out on a series of experiments designed to cool the MnO2 thermite by co-reducing MnO2 and MnO. The reaction enthalpies per mol of oxide for both reductions are respectively - 597 kJ per mol of MnO2 and - 173 kJ per mol of MnO. Initial results with a blend of 1 mol MnO and 0.4 mol MnO2 and a high level of CaF2 (calcium fluoride, fluorite, fluorspar) as a heat sink and slag fluidiser, showed that this kind of mix with a stoichiometric amount of Al powder is capable of making nice blobs of clean Mn metal, albeit at low yields.

In the mean time I've switched from MnO2 to Mn2O3 because thermochemical calculations show that an Mn2O3 runs a little cooler than the corresponding MnO2 reaction, mainly because the Mn2O3 reaction generates more moles of reaction products (per mol of Mn2O3, 2 mol of Mn and 1 mol Al2O3, against 1 mol Mn and 2/3 mol Al2O3 for MnO2).

I've run several small (20 g and 50 g batches) thermites using blends of Mn2O3 and MnO, always using a high level of CaF2. I set the level of CaF2 as a constant molar ratio of CaF2/Al = 0.225, the alumina slag therefore contains always the same molar fraction of CaF2.

On the whole the results indicate that obtaining yields (recovered metal/metal present in the oxide x 100 %) is hard to push much beyond 35 % or so. The last two batches, both with 50 g stoichiometric (and CaF2 = 0.225 molar ratio) mixes, gave the following yields:



Yield....................37 %...................19 %

The 1/0.5 blend gave the highest yield of Mn metal I've ever achieved (out of probably over 20 or so reactions) and the metal is clean skinned and solid. The largest regulus was 7.3 g. Both reactions ran well-contained, leaving a molten slag/metal puddle at the bottom of the crucible. The 1/0 batch yielded metal that was significantly more oxidised, yet on the whole of passable quality.

It's clear though although these results constitute a great improvement to the usual 'explosive' MnO2 thermite, there is only so much cooling the Mn2O3 thermite by blending it with the much cooler MnO can actually achieve in terms of yield improvement.

The only real solution to reducing Mn oxides (or halides) with higher yields is by using a reductant with a much lower melting oxide (or halide).

One such reaction that springs to mind is the reduction of anhydrous MnCl2 with Mg. The reaction enthalpy of MnCl2 + Mg ---> Mn + MgCl2 is unfortunately only a measly - 161 kJ/mol of MnCl2, about a 100 kJ short of success. Thermocalcs show that in adiabatic conditions the reaction products would be heated to about 1,200 K, well above the MP of MgCl2 (987 K) but about 300 K short of the MP of Mn (1,519 K). Pre-heating the mixture by about 300 K or simply heating it to spontaneous ignition could work to obtain liquid Mn and liquid MgCl2.

As regards using a reductant with an oxide of lower MP, that excludes both Mg and Ca, as both have oxides with insanely high MPs. That then really only leaves the alkali metals, in particular Li and Na.

Thermochemical calculation for the reaction Mn2O3 + 6 Li ---> 2 Mn + 3 Li2O (ΔH = - 838 kJ per mol Mn2O3) shows that in adiabatic conditions the estimated end-temperature would be 2,320 K which is still too high and too close to the BP of Mn.

But here we could blend with MnO again. Setting a target end-temperature of 2,000 K (well above the MP of Mn, yet well below its BP as well), the blend composition of a stoichiometric mix has been estimated to be about 1 mol Mn2O3 + 2.6 mol MnO.

One small problem: I haven't got any Li... :(

Sodium, with a ΔH = - 295 kJ per mol Mn2O3 for Mn2O3 + 6 Na ---> 2 Mn + 3 Na2O could also be a good candidate... Haven't got any Na either...:mad:

[Edited on 23-7-2008 by blogfast25]

[Edited on 23-7-2008 by blogfast25]

[Edited on 24-7-2008 by blogfast25]

chloric1 - 23-7-2008 at 23:12

Blogfast-read some of your blogs and I noticed you have done alloy thermites. Have you tried ferromanganese? The red iron oxide is probably more tame and the resulting iron could absorb the manganese. This would be a neat way to produce ferromanganese anodes for electrolytic permanganate production.

blogfast25 - 24-7-2008 at 03:56

No, I haven't but that would work without a question of doubt, for just about any Fe/Mn ratio desired. Since as both the Fe2O3 and the higher oxides of manganese produce such energetic thermites, it might be recommendable to use Mn2O3 or even MnO when co-reducing with iron oxide, to keep the heat generated to a manageable level and prevent too much Mn from evaporating off...

I did once run a mixed Cu2O/MnO2/SiO2 thermite, aiming at an alloy of 70 w% Cu/15 w% Mn and 15 w% of Si. Analysing the resulting bronze I found back these proportions more or less.

[Edited on 24-7-2008 by blogfast25]

-jeffB - 24-7-2008 at 12:41

Originally posted by blogfast25I did once run a mixed Cu2O/MnO2/SiO2 thermite, aiming at an alloy of 70 w% Cu/15 w% Mn and 15 w% of Si. Analysing the resulting bronze I found back these proportions more or less.

I've been wondering if I could blend some Cr2O3 with Fe3O4 and get a polishable "stainless steel". Anybody tried it?

chloric1 - 24-7-2008 at 17:25

Why? If you are going to spend chemicals, make something not found at your hardware store. Like ferromanganese anodes, silicon bronze, titanium bronze etc.

-jeffB - 25-7-2008 at 05:11

Originally posted by chloric1
Why? If you are going to spend chemicals, make something not found at your hardware store. Like ferromanganese anodes, silicon bronze, titanium bronze etc.

You know, it's never occurred to me to ask my hardware store for WHITE-HOT BOILING stainless steel. I'll ring them right up and ask. :)

I'm not trying to develop a production process, just an interesting demonstration. Shopping at the hardware store can be interesting, of course, but not in the same way.

blogfast25 - 25-7-2008 at 05:12


I've been wondering if I could blend some Cr2O3 with Fe3O4 and get a polishable "stainless steel". Anybody tried it?

As far as my experience shows (and theory predicts) mixed thermites are no problem. This is certainly true if both individual thermites yield the reaction products in molten form. I've done FeSi, FeCr, FeV, FeTi, CuTi and a few other exotic ones. These can be used as masteralloys, to dope any other steel (or Fe based alloy) with small amounts of these metals.

In the case of FeSi and FeTi the formulations have to be designed bearing in mind that the Ti and Si thermites are hardly exothermic at all and that most heat will have to come from the Fe2O3 (or Fe3O4) reduction. Aiming too high in Si or Ti would lead to the slag being obtained in solid or "slushy" form, the alloy will then not be able to mix and separate out form the liquid reaction products.

Masteralloys comprised of more than two components are also perfectly possible. There are quite a few thermite based patents pertaining to quite complicated mixtures, including those containing Mo, Mn, W, Co, Cr, Nb and other (high MP) transition metals...

chloric1 - 26-7-2008 at 03:57

Originally posted by -jeffB
You know, it's never occurred to me to ask my hardware store for WHITE-HOT BOILING stainless steel. I'll ring them right up and ask. :)


You put it like that, after leaving my neighborhood hardware store with my white hot stainless, I'll head off to the local pub and have myself a molotov cocktail.

Cr2O3 + Mg

ssdd - 31-7-2008 at 06:50

I recently tried a small batch of Cr2O3 + Mg thermite.

I ignited it using United Nuclear's "Thermite Ignition Mix", which I have a can of laying around. The results were a brilliant 2 foot fountain of sparks that was a very bright yellow color. No residue other than some un-reacted mix was left behind. The rest seems to have gone up in a cloud of white ash.

The issue is that when I made this mix it was in a place where I did not have a scale so the ratios were guessed, I think it may have been a bit Magnesium rich which may have contributed to the reaction speed.

Sorry, no pics it was too fast.


[Edited on 31-7-2008 by ssdd]

chloric1 - 31-7-2008 at 13:08

Ouch! Would be nice to have a regulas of chromium:(:(

Actually, I am sticking to using aluminum here and I wanted to use potassium dichromate as a booster. I have both 300 mesh and 400 mesh Al, so I think 15% dichromate is all that should be needed. I worked out the dichromate reduction in thiis equation:

K2Cr2O7 + 2Al >2KAlO2 + Cr2O3
Cr2O3 +2Al > 2Cr +Al2O3
NET: K2Cr2O7 + 4Al >2KAlO2 +2Cr +2Al2O3

Yep, that seems right.

Ritter - 2-8-2008 at 08:38

I came upon a patent to the U.K. MoD for a thermite-like incendiary formulation here

They use soluble Pb(II) salts (either acetate or nitrate) in water solution mixed with powdered boron to create intimate mixtures that can be ignited when dry. Here is my guess at the processes involved:

2Pb(NO3)2 > 2PbO + 4NO2 + O2

This reaction takes place in hot aqueous NaOH. No mention is made of the gas production.

And the thermite-like reaction:

3PbO + 2B > B2O3 + 3Pb

Here is Example I via an OCR scan of the patent pdf:


6.0g. of amorphous boron (around 1.0 micron particle size) are 1.2 liters of aqueous sodium hydroxide solution containing 0.32 mole sodium hydroxide. The suspension is stirred and heated to 90°C.

1.06 liters of lead nitrate solution containing 0.8 mole
lead nitrate and 1.06 liter of sodium hydroxide solution
containing 1.60 moles sodium hydroxide are added to
the stirred suspension simultaneously and dropwise
during 42 minutes, the temperature being maintained
at 90°C during the precipitation. A further 15 minutes
stirring after addition is given, the precipitated product
settles quickly when the stirring is discontinued and the
supernatant liquor is decanted hot. The boron is com
pletely incorporated in the lead oxide orthorhombic
crystals and the supernatant liquor is clear and free
from elemental boron. The product is washed twice in
the precipitating pan by decantation.

The product is transferred on to cambric cloth on a Buchner funnel, sucked free of excess water, washed with methylated sprits and dried by passage of dry air
or on a hot table at 60°C. The yield obtained is 180g. The product has a low bulk density of 0.4 g/mI and contains about 3 percent boron by weight. In appearance it ressembles aluminium flake and it readily burns when ignited in an open train.

FYI, PbO & the Pb(II) salts are toxic. Other than that, this looks like a workable procedure for a home lab.

A related boron-fueled thermite-like system is used in a pyrotechnic formulation in to the Swedish arms firm Bofors. In their formulations they use the following mixtures:

Zirconium, titanium or nickel/zinc alloy

[Edited on 2-8-2008 by Ritter]

blogfast25 - 2-8-2008 at 11:58

Originally posted by chloric1
Ouch! Would be nice to have a regulas of chromium:(:(

Actually, I am sticking to using aluminum here and I wanted to use potassium dichromate as a booster. I have both 300 mesh and 400 mesh Al, so I think 15% dichromate is all that should be needed. I worked out the dichromate reduction in thiis equation:

K2Cr2O7 + 2Al >2KAlO2 + Cr2O3
Cr2O3 +2Al > 2Cr +Al2O3
NET: K2Cr2O7 + 4Al >2KAlO2 +2Cr +2Al2O3

Yep, that seems right.

With dichromate alone this is likely to be near-explosive. Small amounts of dichromate could be used to increase reaction temperature.

Personally for Chromium I prefer the classic KClO3 + 2 Al ---> KCl + Al2O3 as a booster system for Cr2O3 + 2Al ---> 2 Cr + Al2O3

I have a recipe (with KClO3 and 400 mesh Al) that gives good Chromium reguli, if you're interested...

chloric1 - 2-8-2008 at 16:02

Originally posted by blogfast25

Personally for Chromium I prefer the classic KClO3 + 2 Al ---> KCl + Al2O3 as a booster system for Cr2O3 + 2Al ---> 2 Cr + Al2O3

I have a recipe (with KClO3 and 400 mesh Al) that gives good Chromium reguli, if you're interested...

Well, in that case I am interested in your formula. I wish to try both and do a comparison. If I have time tonight I want to try a thermite based on this formula:

Chromium Oxide 240 grams
Aluminum 300 mesh 99 grams
Potassium Dichromate 36 grams

I believe this what I am after. After you give me your formula then I can try some other time and compare both.

BTW I know thermite usually burns 2400 to 2500 degrees C. What about aluminum with STRONG oxiders like dichromate,perchlorate, permanganate etc? Possibly 3000 degrees?

[Edited on 8/2/2008 by chloric1]

ShadowWarrior4444 - 2-8-2008 at 16:56

Originally posted by chloric1
BTW I know thermite usually burns 2400 to 2500 degrees C. What about aluminum with STRONG oxiders like dichromate,perchlorate, permanganate etc? Possibly 3000 degrees?

[Edited on 8/2/2008 by chloric1]


(Permanganate and Al is flash powder.)

Small amounts added as a booster will likely increase burn rate and ease of ignition.

[Edited on 8-2-2008 by ShadowWarrior4444]

blogfast25 - 3-8-2008 at 05:28


Firstly, the end-temperature to which a thermite reaction runs depends on various factors, the thermochemistry itself being a prime factor. Thermites based on TiO2, SiO2 and some other very, very stable oxides either don't burn at all or burn slowly and not very hot. This is certainly also true of a few Period 5 and Period 6 transition metals, where reduction of halides is usually preferred.

The Gibbs Free Energy released during the reaction (ΔG = ΔH - TΔS) indicates whether the reaction can take place or not and the reaction enthalpy (heat) ΔH, together with how much reaction products are formed, their heat capacities and heats of fusion (if applicable), as well as to what extent the reaction vessel is thermally insulated from the rest of the world, determines the end-temperature precisely and unequivocally.

Regards the Chromium thermite, remember that to obtain good quality metal reguli, usually fluorite is needed to obtain good metal/slag separation (but it isn't always indispensable). Most of my thermites, designed to produce good quality metal, contain fluorite (CaF<sub>2</sub>;). Here's my Cr formula (this is a stoichiometrically balanced mix):

in mol: Cr<sub>2</sub>O<sub>3</sub> = 1 mol; Al = 2.82 mol; KClO<sub>3</sub> = 0.41 mol, CaF<sub>2</sub> = 0.6345 mol

in gram: Cr<sub>2</sub>O<sub>3</sub> = 152 g; Al = 75.8 g; KClO<sub>3</sub> = 50.2 g, CaF<sub>2</sub> = 49.5 g

The fluorite level could be reduced or it could be eliminated altogether but bear in mind that fluorite is a slag fluidiser as well as a heat sink: without fluorite the reaction will run hotter (and that's not necessarily an advantage, if metal production and not pyrotechnics is your goal).

Let me know how you get on... :cool:


Also, the reaction between K2Cr2O7 and Al is likely to run to elemental K, not KAlO2, because K2O + 2/3 Al ---> 2 K + 1/3 Al2O3 is exothermic by -195 kJ/mol of K2O.

The overall booster reactions can then be balanced as:

K2Cr2O7 ---> K2O + 2 CrO3
K2O + 2/3 Al ---> 2 K + 1/3 Al2O3
2 { CrO3 +2 Al ---> Cr + Al2O3 }

K2Cr2O7 + 14/3 Al ---> 2 Cr + 2 K + 7/3 Al2O3

The third part is the one likely to produce enormous amounts of heat.

If you want to use 240 g (1.58 mol) of Cr2O3 and 36 g (0.122 mol) of K2Cr2O7, the amount of Al should then be increased slightly to about 107.4 g to respect stoichiometry.

The level of heat booster in your dichromate boosted formulation is quite low: 0.077 mol of K2Cr2O7 per mol of Cr2O3. For my chlorate based formulation it is 0.41 mol of chlorate per mol of Cr2O3.

But since as I have no thermochemistry data for K2Cr2O7 or Cr (+VI) oxide at hand, it's difficult to say whether the chromate boost will be enough to obtain molten slag and metal. The advantage of using dichromate is of course that it also produces Cr and not just alumina and KCl. I think it's the safest way for now to test the low level of dichromate first.

[Edited on 3-8-2008 by blogfast25]

chloric1 - 3-8-2008 at 17:27

Thanks blogfast! I appreciate that and I am taking notes. I got the dichromate idea from a couple of older digitized inorganic chemistry books. One uses Chromium trioxide and the more modern one uses potassium dichromate! The hexavalent chrome makes up significant portions simular to your chlorate formula but they are using 30 mesh Al!! So I toned it down to accommodate for finer powders.

OUCH! I did not factor in the fact Potassium metal is formed. Certain that it will be a purple vapor flame!:o:o
I really need to get a grasp of theorical thermochemistry. I am planning on studying in the University soon.

I do not have fluorspar at this moment but plan to get some. For this, and the fact I have spent the weekend entertaining guest I have yet to do my trial. I will let you know when I get around to it(possibly Monday or Tuesday evening). My next big order will include some fluxes so I can
compare the run with/without flux.

I would think slightly less fluorspar could be used because the potassium vapor should lower temperature somewhat but how much I don't know.

[Edited on 8/3/2008 by chloric1]

blogfast25 - 4-8-2008 at 08:27


Thermochemistry of thermites (and assorted reactions) isn't very complicated at all: it requires only basic algebra and invoking a couple of laws.

Here's an example of a thermochemical calculation for a boosted TiO2 thermite (the post is mine)

Another one on ferrotitanium, as well as a general calculation of the estimated end-temperature for a generic thermite (the post is mine too)

The cooling effect of the potassium will be small but could be accurately estimated using the heat capacities (Cp,l and Cp,s), heat of fusion and heat of evaporation of K. In your proposed formulation the cooling effect is likely to be negligible because the amount of K formed is small.

But in the KClO3 boosted formulation, where 0.41 mol of KCl are formed, heated, fused and evaporated (following Hess), the cooling effect has to be taken into account even though compared to the heat generated by KClO3 + 2 Al ---> KCl + Al2O3 is enormous. It all depends how accurate you want the calculation to, of course... :)

[Edited on 4-8-2008 by blogfast25]

[Edited on 4-8-2008 by blogfast25]

[Edited on 4-8-2008 by blogfast25]

chloric1 - 4-8-2008 at 15:06

Well the more I thought about superheated potassium vapor, the more I realized how little of an effect it will have. Not only that but I am considering the chemistry here. You will have molten Aluminum Oxide not to mention heated air surrounding reaction. So any losses of potassium would result in immediate combustion releasing white clouds of potassium oxide(don't breath;)),or immediate combination of said oxide with the molten alumina hence my original deduction of potassium aluminate as a product intermixed in a matrix of the alumina slag. I appreciate the post on thermochemistry of thermites. With your permission, I would like to copy them and save them as a Word doc.

blogfast25 - 5-8-2008 at 03:23


Feel free to copy.

I'm quite interested in testing K2Cr2O7 (I have some) in a Cr2O3 formulation myself. The dichromate should be incredibly powerful, because of its oxygen (molar) content (7 mol O per mol), slightly higher than twice the molar oxygen content of chlorate (3 mol O per mol). Very roughly speaking it will generate about twice the amount of energy of potassium chlorate (compared on a mol to mol basis). That's why it's definitely recommendable to start with low quantities: it should be very, very energetic.


I've just estimated the dichromate booster reaction K2Cr2O7 + 14/3 Al ---> 2 Cr + 2 K + 7/3 Al2O3 to have a reaction enthalpy of about - 2800 kJ/mol of K2Cr2O7, more than twice the value of - 1255 kJ/mol of KClO3 (for KClO3 + 2 Al ---> KCl + Al2O3).

That is a phenomenally high value for a booster system. A formulation with 0.15 to 0.2 mol of K2Cr2O7 per mol of Cr2O3 will almost certainly give liquid Al2O3 and liquid Cr. That level of booster will yield an extra 420 to 560 kJ, to supplement the heat coming from the Cr2O3 reduction.

But for now, for safety reasons, I'd stick to your proposed lower level. Crank it up gradually, if need be...

[Edited on 5-8-2008 by blogfast25]

[Edited on 5-8-2008 by blogfast25]

chloric1 - 5-8-2008 at 15:07

Well, it definately ignited I will post a link to video when I get a chance to upload it. I use a flower pot half fill with sand. Truthfully, I was worried the pot would fly apart but it just made a "ping" when after the permanganate/glycerol hypergolic subsided and the thermite ignited. I really am getting tired of this ignition system because the smoke obscures my view and it smells like burnt sugar. Anyhow on to the mix. It gave off green smoke:o Obviously some chrome oxide was mechanically lifted by the vigor of the reaction. The slag/mess looks porous so I am not too optimistic. I will break her open around 9 PM GMT-5 to look for metal.

Since I prepared like 383 grams of thermite I still have 70-80% left:cool: so we can optimize together then compare to chlorate bosster and I will order fluorspar this week.

Update: Just uploaded my videohere

I did not find a distinctive globule or regulas. More like a broccoli sprout:( It is very hard but very brittle. I think it is chromium in a alumina matrix. I also seen some unreacted chrome oxide so I think it was not hot enough. I will post more findings later.

[Edited on 8/5/2008 by chloric1]

blogfast25 - 5-8-2008 at 22:51

You're very close to success. I can tell just by looking at the vid that the whole thing is running just a little too cool. Increase the dichromate/Al slightly (stoichiometrically).

For ignition I use a small amount of KClO3/Al (stoichiom.) mix, about 1 ml, lit with a piece of magnesium ribbon. Alternatively, light the chlorate with a pen sized blow torch. KClO3/Al burns incredibly hot, yet is very safe: it isn't a flashpowder at all.

The slag is made up mainly of fused alumina (and that IS HARD!) and powdered chromium, a kind of cermet. CaF2 will help the alumina/Cr flow to the bottom of the crucible before it solidifies on cooling.

chloric1 - 6-8-2008 at 02:11

Well I have magnesium ribbon coming this wek I hope. I ordered 225 feet! I plan on selling most of it to finance research:cool: Next I do this I will measure out a 100 gram charge. I will figure out the dichromate addition based on your numbers of .15 tp .2 moles. Today I have to decide who to order fluorspar from. Not all pottery suppliers are created equal. I also need to order black iron oxide and possibly something else. This all helps to justify the shipping costs.

blogfast25 - 6-8-2008 at 07:17

Why not get some vanadium pentoxide (pottery grade)? That's a real nice thermite, very hot and nice metal too. Cobalt's nice too, from CoO... The one I'm most interested in is niobium pentoxide: no boost needed, burns straight to about 2740 K! But I can't get any Nb2O5 here... :(

Instead of CaF2, cryolite (Na3AlF6) should also do the trick but it's more expensive of course.

I'll probably be moving away from oxides and move to halides (mostly chlorides). I've a MnCl2 + Mg ---> Mn + MgCl2 on the drawing board.

chloric1 - 6-8-2008 at 13:16

Thats just it, I got numerous suppliers for CaF2 and cryolite but I need the supplier with reasonably priced vanadium pentoxide. So I can thermite it and sell some too. I have one supplier that wants $45 per pound!! I have a couple others that range $13-15 per pound! There is definately no set standard on these type of item hence money can easily be made. The niobium pentoxide will probably have to be purchased directly from China or India in 25Kg quantities or more. I am definately unable to swing that as of now:(

Why do you want to move away from oxides? Want a new challenge? The problem with halides is that many are VERY soluble and hydroscopic. Many Fluorides are insoluble and would be easy to obtain anhydrous.

Personally, I wanted to make up some lead bromide and make a thermite with that and try to catch the aluminum bromide smoke in a bell jar or?:o

[Edited on 8/6/2008 by chloric1]

blogfast25 - 7-8-2008 at 01:16

Yes, a new challenge. Most chlorides are indeed hygroscopic but drying them is usually possible. Anhydrous fluorides are hard to make: usually hydrofluoric acid or anhydrous HF and in some cases fluorine is needed. They're also much more expensive and more restricted for sale to private individuals (at least in Europe).

High grade Nb2O5 is available from ChemSavers but needless to say only at ridiculous prices.

The reduction PbBr<sub>2</sub> + 2/3 Al ---> Pb + 2/3 AlBr<sub>3</sub> is only just barely exothermic: ΔH = - 64 kJ/mol of PbBr2 (at 298 K) (for chlorides and bromides, Mg is much preferred to Al but for fluorides Al also works. This is mainly because of the fact that in the series of AlX3, AlF3 has by far the highest heat of formation).

I'm not sure (but it can be calculated) whether the temperature would be high enough to evaporate the AlBr3 and in any case it would be heavily contaminated with lead, due to mechanical entrainment of the nascent lead. AlBr3 must also be very prone to hydrolysis. Like for AlCl3, reaction of Al with dry HBr in a glass labware set up with condenser is to be much preferred here.

I just took receipt of my Mg powder, so the MnCl2 reduction project can now go ahead. :)

chloric1 - 7-8-2008 at 15:03

Well, I had to order cryolite because the supplier was out of fluorspar. I also ordered nickel and black iron oxides to justify shipping. I ordered the vanadium pentoxide from another supplier at hopefully a low price along with red copper oxide. They will confirm with me later.

Good luck with your manganese chloride project. You wouldn't have the thermal date for cryolite would you? Have not had been able to visit NIST yet.

blogfast25 - 8-8-2008 at 05:07

Na<sub>3</sub>AlF<sub>6</sub> thermodata:

here they are. MP ≈ 1,000 C.

For my own experiments I would simply substitute CaF2 by Na3AlF6 mol per mol, then reduce the amount by 20 - 30 %: because of cryolite's lower MP compared to CaF2 its slag fluidising effect should be stronger, as liquid Na3AlF6 at thermite end-temperatures should be even less viscous than CaF2.

Out of curiosity, what's your interest in AlBr<sub>3</sub>? :)

chloric1 - 8-8-2008 at 14:13

No REAL interest in aluminum bromide since I prefer inorganic chemistry over organic. I guess its I want to take a video and say hey I made this and I did so unconventionally. Besides aluminum bromide vapors will look cool on video:cool:

I received my magnesium ribbon today so I can use this to ignite my thermites now. I will try the chlorate/aluminum comp although I will be extra cautious. At what mesh does this mix be come flash?

Thanks for the data!

P.S. Please keep me informedon your halide explorations. If you can shoot a video I would like to see it.

[Edited on 8/8/2008 by chloric1]

blogfast25 - 9-8-2008 at 06:28

Originally posted by chloric1
Besides aluminum bromide vapors will look cool on video:cool:

I received my magnesium ribbon today so I can use this to ignite my thermites now. I will try the chlorate/aluminum comp although I will be extra cautious. At what mesh does this mix be come flash?

Thanks for the data!

P.S. Please keep me informedon your halide explorations. If you can shoot a video I would like to see it.

[Edited on 8/8/2008 by chloric1]

The problem with thermite reactions (and analogues) that produce gaseous reaction products in significant quantities is that it's precisely that that can cause the reaction mix to deflagrate ("explode" as some would call it). MnO2 thermites are an example: with very fine ingredients such termites simply go "poooofff!" because the heat generated vaporises the formed manganese. Other thermites with quite low HoF oxides/relatively low boiling metals tend to do that too.

In the case of the PbBr2/Al reaction, reaction speed should be low and temperature accordingly low because of the small heat of reaction.

Chlorate/Al is in my experience very safe. I use fine potassium (reagent grade) with 400 mesh Al powder: never had a problem. Not sure whether there is a mesh size at which this mixture tends to flash. I did read somewhere that including sulfur in the mix can lead to a shock-sensitive mixture. Yet for some time (but not anymore) I did used to spike the chlorate/Al mix with small amounts of S, to increase heat output. I had no problems doing this.

If you're a little jittery about using chlorate I suggest to mix the chlorate with the oxide first, then add the Al powder. But I usually simply mix everything together w/o problems. Chlorate/Al booster systems are used to spike thermites (for metal production) in industry all the time.

With Mg powder it becomes a different story: chlorate/Mg is supposed to be a flasher. I've never tried it but I did once try an SiO2 thermire with magnalium (50 Al/50 Mg), boosted with chlorate. The thermite did flash (no explosion though), leaving an empty crucible....

As regards vids, I'll try. But my experience is that most successful reactions (success being defined here as 'obtaining good, clean metal') look very similar. They run fast, to white heat (to well above the MP of alumina and possibly up to 2,500 C), well contained and at the end most of the slag and metal are found as a flat, white hot, molten puddle at the bottom of the crucible. Below are some vids of the TiO2 experiments, unboosted and boosted with KClO3. Scroll down a bit to the "530 grams of KClO3-boosted..." and watch the video: this is what thermite heaven looks like: this one is shot by Jeffrey using my TiO<sub>2</sub> formulation (slightly adapted). See the white-hot, flat puddle of molten alumina and molten metal gathered neatly at the bottom. That's my baby!

All my successful thermites with transition metals look strikingly similar, even though reaction times may vary a little. Temperature is the great leveler here: if the reactions are designed to achieve more or less the same end-temperatures (as mine are) they will tend to run quite similarly.

The MnCl2/Mg reaction will be the exception to the rule because here I need to keep the end-temperature purposely low (considerably lower than the BP of Mn: that's why the chloride/Mg route was chosen in the first place).

[Edited on 9-8-2008 by blogfast25]

chloric1 - 9-8-2008 at 08:31

amazingrust is indeed your site!:D I like to visit this alot. Just want to let you know that for some reason your vid links are down. I get Internet Explorer cannot load page. Let me know if you can get the vids back up.

blogfast25 - 9-8-2008 at 09:40

No, no. isn't my website, it belongs to a guy called Jeffrey with whom I'm friendly, that's all. When I checked a few vids less than an hour ago they were fine. Temp glitch, I feel...

chloric1 - 9-8-2008 at 17:40

Oh OK my mistake I wonder if there is something wrong on my end. I mistook your enthusiasm for the titanium vid as a praise of your own creation. None the less you have been a tremendous help to me and I sincerely thank you. I am starting to really look at the thermochemistry of this to get a better feel for this. One thermite I have been thinking about and it might need coarser aluminum becuase of the activity, is barium sulfate. One product barium sulfide is easily soluble in hot water and can be reacted with excess KOH or NaOH to precipitate the barium hydroxide on cooling to near zero celsius.

blogfast25 - 10-8-2008 at 05:22

Yes, the video is his but the development of the TiO2 thermite formulation is mine. At the time I had emailed Jeffrey about it almost immediately and the same day he enthusiastically reproduced my initial results and made the video. The initial find was reported on this SienceMadness thread, where Jeffrey also commented as mrjeffy321 and reported confirming my results (looking through the thread I see that you were there too).

The barium sulfate thermite will almost certainly work, no question about it: I've run countless such reactions with dried plaster of Paris according to:

CaSO4 + 8/3 Al ---> CaS + 4/3 Al2O3

That reaction releases a hell of a lot of energy and I've even used it as a booster reaction for TiO2 thermites, successfully. In terms of ΔH of the overall reaction it's more or less the same as the chlorate reaction and thermochemically speaking both can be used more or less interchangeably.

But to get access to the reaction product BaS may prove difficult, mixed in with alumina as it is. The slag from the CaSO4 reactions positively reeks of H2S because of hydrolysis of the sulfide, so if you treated it with HCl you'd leach out some CaCl2, no doubt about that. The question is how much? It may well be worth looking into that, using CaSO4 as a model, before you start sacrificing the more expensive BaSO4... The yields may be surprisingly disappointing, not sure though... Size-reducing the slag for the purpose of extraction will also be difficult: fused alumina is incredibly hard.

Perhaps unusually high levels of CaF2 (or Na3AlF6) may cool the reaction enough, so that a porous, rather than compact slag heap results. CaF2 (and Na3AlF6) is also much softer than Al2O3 which would make a softer slag, easier to size reduce, essential to successfully extract as much BaCl2 as possible...

An approximate thermocalc will allow you to estimate the amount of heat-sink (CaF2 or Na3AlF6) needed to cool the reaction, say to just below the MP of alumina, thereby obtaining a porous 'slag muffin', much easier to break up, size reduce and extract the BaS from. The level of heat sink needed may however be so high that the mixture proves difficult (or impossible) to ignite because the heat balance equation makes no pronouncements about the kinetics of the reaction.

All in all a rather smelly method to access soluble Ba salts but well worth contemplating I feel... It will generate a mol of H2S per mol of soluble Ba, that's a helluvalot of rotten eggs!

And thanks for your appreciation. Frankly I love nothing more than to expand a little about my experiences with thermites and assorted reduction thermochemistry because I feel far too many backyard experimenters (but not all, of course) take an approach of "let's stick a bit of magnesium ribbon in it and keep our fingers crossed". But as with all reactions, there's a more than a bit science needed to get the results wanted. This is as much true if you want to create spectacular pyrotechnics or if producing small amounts of relatively pure metal is your purpose...

[Edited on 10-8-2008 by blogfast25]

First reduction of MnCl2 with Mg powder

blogfast25 - 10-8-2008 at 07:30

I ran the first 20 g batch (stoichiometric mix 1:1 moles) of MnCl<sub>2</sub> + Mg today.

Lit with a few g of KClO3/Al/Mg ribbon fuse, it burned right through, slowly but nevertheless quite hot and with a continuous yellow flame, a few cm high. There was no unreacted mix left, at least going by visual inspection.

The slag/metal mixture was very predictably of the 'porous slag muffin' type, so typical of heat-starved thermites. The slag looked like what I would imagine sintered anhydrous MgCl<sub>2</sub> to look like. Predictably, no Mn reguli to be found.

I recovered all the slag and ground it down to powder in my granite mortar and pestle. I transferred the powder into a Pyrex measuring jug and added dionised water. Considerable heat evolved, presumably the heat of hydration of MgCl<sub>2</sub>. There is also gas evolution which I presume to be hydrogen, from Mn + 2 H<sub>2</sub>O ---> Mn(OH)<sub>2</sub> + H<sub>2</sub> (Mn is known to react with hot water). Right know this thin suspension is grayish. If the gas evolution stops, I'll filter and test for Mg and Mn. The black-gray solid matter, assuming it is indeed powdered Mn, should react strongly with HCl.

All in all, this mixture, without any heat boosting measures, does what it says on the thermochemical tin.
Much work now lies ahead to try and heat boost it to obtain lump Mn metal.

The production of anhydrous MnCl2 will also have to be stepped up and refined: I know the product contains residual NH4Cl (from the drying process) and that it contains small amounts of oxide (going by the colour).

Nick F - 11-8-2008 at 08:16

I have about a hundred grams of UO3, and out of curiosity would like to try making some U metal, but only a gram or two at a time (to reduce wastage if it doesn't work).
Obviously though the charge will have to be above a minimum size to allow it to stay hot enough for long enough for the metal droplets to sink and coalesce.
I was thinking of putting UO3, Mg and some sort of flux at the bottom of a large graphite crucible, with a charge of KClO3, Al and Al2O3 (to reduce burn rate) on top.
What I hope would happen would be that the charge on top would burn down and produce lots of very hot liquid, and then ignite the thermite mix at the bottom. The Mg would then reduce the UO3, and due to all the alumina slag on top everything would stay hot for long enough to get a nice blob of U. I also hope that the slag on top would reduce the amount of U that escapes in the fumes.
What do you think of this plan? Also, any ideas for a suitable flux? My initial thought was KF, but then I worried about the volatility of UF6.
Thanks for any advice!

blogfast25 - 11-8-2008 at 09:02

Nick F:

I'm a little pressed for time right now so I'll have to keep this short.

IMHO and at first glance, for the reduction of UO<sub>3</sub>, Al would be a much better reductant: Mg is likely to be far too energetic, yet the insanely high MP of MgO makes it difficult to obtain both the nascent U metal and the MgO in liquid form. This is strictly necessary if lump metal is what you're looking to make. With Al, the experimental set-up would be much simpler that what you propose here...

Again and as always, only a thermochemical calculation can tell you what the expected end-temperature of such a reaction would be. If I find the time, I'll have a look later on or tomorrow.

The heat of formation (at 298 K) for UO3 is - 1224 kJ/mol (of UO3), that makes the reaction enthalpy for UO<sub>3</sub> + 2 Al ---> U + Al<sub>2</sub>O<sub>3</sub> (at 298 K) about ΔH = - 452 kJ/mol of UO<sub>3</sub>. That's a respectable value and it might be enough to heat the one mol of U and the one mol of alumina to above the MP of alumina. I have the heat capacities and heats of fusion of both U and alumina in my files so I can calculate the estimated end-temperature for such a reaction (stoichiometric mix) in adiabatic conditions.


A quick thermocalc shows that in adiabatic conditions, a stoichiometric mix of UO<sub>3</sub> and Al would burn to an estimated end-temperature of about 2,300 K (2,030 C), close to the melting point of alumina (2,054 C) (and well above the melting point of U i.e. 1,132 C). To obtain a liquid metal/slag mixture and to allow the U to separate out and settle at the bottom of the crucible, small amounts of KClO3/Al may be needed to boost overall heat generated.

For safety reasons, such a reaction should IMHO, always be carried out in a bomb-type reactor, such as a bomb calorimeter or similar, because thermites will always emit some smoke, which in the case of a uranium based reaction could contain quite lethal microparticles of U or its oxides.

Alternatively a much tamer reaction can be obtained with UCl<sub>4</sub> and Mg (or Li, Na or K). Some basic thermocalcs starting with UCl4 can be found here.

The safety concerns would remain the same but halide reductions can be run at lower temperatures...

[Edited on 11-8-2008 by blogfast25]

[Edited on 11-8-2008 by blogfast25]

[Edited on 11-8-2008 by blogfast25]

Nick F - 14-8-2008 at 07:02

I'll let you know if I ever get round to trying it, but it might have to wait until I visit my parents because they have a bigger test area (garden :)).

Dichromate boosted chromium thermite

blogfast25 - 14-8-2008 at 07:29

I ran two K2Cr2O7 boosted Cr2O3 thermites today, both stoichiometric mixes, one with 0.2 mol K2Cr2O7 per mol of Cr2O3, one with 0.4 mol K2Cr2O7 per mol of Cr2O3.

Somewhat to my surprise, neither yielded good quality Cr reguli.

The first one ran fast but clearly not hot enough because a porous slag/metal muffin was obtained. Inside glistening Cr could be seen but all metal was sub mm.

Increasing the dichromate level to 0.4 mol made the thermite run decidedly faster and the slag/metal was largely molten and collected (more or less) at the bottom of the crucible. Breaking open, clear areas of Cr metal could be seen but poorly separated from the alumina slag. Poor metal coalescence due to slightly too low end-temperature (premature freezing of the alumina) is clearly the cause here (I've seen many such cases before).

Success should be possible at 0.5 - 0.6 mol of dichromate but that's much higher than the estimated - 2,800 kJ of booster heat (per mol of dichromate) would suggest...

To be continured...

[Edited on 14-8-2008 by blogfast25]

Dichromate Boosted Thermite

chloric1 - 14-8-2008 at 16:01

Well this is a better result but my metal yield was paultry. I was going to throw out the the byproduct but I have not and will not. I am going to put small amounts of it where ever chromium may be usefull. Hoping the heat from another charge will help separate more chromium out of its cermet. You can see my improved video here

@blogfast-did you add flux? I am not using flux because it has not been delivered yet. I will post pictures of my metal later. Only one nugget was significant size at just under 1cm.
Also, what kind of metal yeilds do you get from chlorate boosting?
[Edited on 8/14/2008 by chloric1]

[Edited on 8/14/2008 by chloric1]

blogfast25 - 15-8-2008 at 04:15


Yes, all these formulations are fluxed. I set the CaF<sub>2</sub> to a level that is equal to (in mol) CaF<sub>2</sub> = 0.225 Al. This ensures the slag (Al2O3 + CaF2) contains a constant molar fraction of CaF2, for purposes of comparison. Sometimes I use CaF2 = 0.1125 Al, it depends...

The generic formulation of the dichromate-boosted chromium formulation then becomes:

Species ............................................. mol

Cr<sub>2</sub>O<sub>3</sub> ................................................. 1
K<sub>2</sub>Cr<sub>2</sub>O<sub>7</sub> ............................................. x
Al ........................................................ 2 + (14/3) x
CaF<sub>2</sub> .................................................. 0.225 [2 + (14/3) x]

Yields (RECOVERED metal/stoichio metal times 100 %) from chlorate boosting are about 50 - 60 %. It has to be borne in mind that considerable amounts of the metal get locked into the slag that freezes when it hits the cold walls of the crucible. That helps explain why larger thermites tend to yield more recoverable metal: the fraction of metal/slag mix prematurely frozen on the crucible walls tends to be smaller. Also, closed reactors with some degree of thermal insulation (alumina or magnesia lining for instance) should give higher yields.

Your latest test seems to run much hotter: which variable did you change?

I would strongly advise against piling up one charge on the remains of the previous one: thermodynamics being what it is one can show very easily that unless the succeeding charge is much larger or much hotter than the preceding one, the remains of the preceding one will simply not melt (heat up, yes, melt? NO!!)

Now I've got an appointment with an MnCl2/Mg reaction... :)


Two more tests at x = 0.6 and x = 0.5 were both disappointing. At this level the slag/metal mass was really completely molten, yet metal coalescence was piss-poor. No real Cr reguli whatsoever. I'll try one with a much reduced level of CaF<sub>2</sub>...


A test trying to use Mg + I<sub>2</sub> ---> MgI<sub>2</sub> as a heat booster for MnCl2 + Mg ---> Mn + MgCl2 failed: the iodine simply fumes off in a spectacular purple cloud, but the main reaction still proceeded.

Ca + I2 ---> CaI2 is sometimes used as a booster in calciothermic reductions but obviously this must only work in closed reactor conditions, where the iodine has nowhere to go but react with free Ca.

Another test using permanganate as a heat booster at 0.04 mol (per mol of MnCl2) was inconclusive. I know, I know: Danger! Flash powder! But at this very small level this is w/o danger, IMHO.

The mixture was for some unknown reason difficult to light and when it finally did start it went on quite irregularly, sputteringly. At one point it seemed to go to white heat but by then the reaction mix had been used up. The slag metal/mix clearly shows a higher end-temperature was achieved: more fusion of the MgCl2 was noticable.

At such low levels of KMnO<sub>4</sub> it's actually quite difficult to mix it in homogeneously. Another attempt at a slightly higher level and with better mixing, soon. Assuming the overall booster reaction is KMnO<sub>4</sub> + 4 Mg ---> K + Mn + 4 MgO, the estimated reaction enthalpy should be in the order of about - 2,000 kJ/mol (of permanganate).

Other tests planned are pre-heating the mix (no booster) by about 300 K, prior to ignition and heating the mixture (no booster) to auto-ignition.

[Edited on 15-8-2008 by blogfast25]

[Edited on 15-8-2008 by blogfast25]

chloric1 - 15-8-2008 at 15:03

@blogfast- Point taken on the thermodynamics of heating previous runs. I may someday try to make a pot of molten boric acid or borax and add the slag to see if it will help me separate more metal.

As far as run #2 goes:

As opposed to the 0.07mol of potassium dichromate I simply doubled the dichromate and added the additional aluminum to compensate. So I took 100grams of my previous mix and added 9 grams of potassium dichromate and 4 grams of aluminum for a total mass of 113 grams

Also, my ignition system changed as I only used magnesium. This was done in two staged a ribbon burning into partially imbedded shavings in the mix itself.

I have not measured the yield of my chromium yet but I know its not 50 or 60%.

I will do one last trial with remainder of the dichromate/chrome thermite adding the same amounts of dichromate and aluminum but submerging the reaction vessle in a much larger pot filled with sand to hold in heat a little longer. I wish I had some vermiculite around.

After this I will use my other 220+ grams to make chlorate boosted Chrome thermite. I do not know why chlorate would yield more chromium. Unless molten potassium chloride is a really good flux or insulator.

BTW my cryolite arrived yesterday after I posted to this thread:)

[Edited on 8/15/2008 by chloric1]

blogfast25 - 16-8-2008 at 05:14


So you were at 0.14 mol of dichromate.

I'm at a loss as to why this booster system, which clearly works in terms of boosting heat output and thus end-temperature, seems to impede metal coalescence. I've noticed the slag seems a little more glassy compared to what is being obtained with KClO<sub>3</sub>.

I also thought the elemental K might be causing problems, but how??? One suspect here could be interference of the hot, free K with the fluorite, according CaF<sub>2</sub> + 2 K ---> Ca + 2 KF. This has a positive heat of reaction (is endothermic) at 298 K of + 88 kJ/mol (of CaF2), so at first glance cannot proceed. But in many cases such endothermic reactions (at 298 K) do proceed at sufficiently elevated temperatures. Hell, the most used reduction reaction on Earth, the reduction of iron ore with CO in blast furnaces, proceeds only at the cauldron conditions of the furnace!

What's more, at thermite temperatures, KF, with a BP of a mere 1505 C, would be volatile, pushing the equilibrium to the right by removal of the reaction products.

It sounds far fetched but right now it's the only thing I can come up with (LOL). Right now it's merely a hypothesis, nothing more

If I'm remotely right about this then cryolite should suffer even more from 'potassium attack' because AlF<sub>3</sub> + 3 K ---> Al + 3 KF is exothermic at 298 K by about - 200 kJ (per mol of AlF3).

And on KCl in chlorate boosted mixes: KCl has a boiling point of about 1500 C and boils off. Chlorate boosted thermites do indeed tend to be quite smoky.


On the MnCl2 reduction side of things, I'm thinking of using good ole' MnO<sub>2</sub> (wild horses couldn't drag me away!) according MnO<sub>2</sub> + 2 Mg ---> Mn + 2 MgO, ΔH = - 684 kJ/mol (of MnO2, @ 298 K).

I'll be thermocalcing that this afternoon. If favourable that's me back to making some MnO2. Some things never change, I guess...


For those interested, here's another example of a thermochemical calculation, applied to an MnCl2/MnO2/Mg stoichiometric mix (in adiabatic conditions).

Main reaction: MnCl<sub>2</sub> + Mg ---> Mn + MgCl<sub>2</sub>, ΔH<sub>298 K</sub> = - 161 kJ/mol (of MnCl<sub>2</sub>;)

Booster reaction: MnO<sub>2</sub> + 2 Mg ---> Mn + 2 MgO, ΔH<sub>298 K</sub> = - 684 kJ/mol (of MnO<sub>2</sub>;)

I set the target end-temperature at 1,800 K, well above the MPs of both Mn and MnCl<sub>2</sub>, but well below the BP of Mn (and below the MP of MgO).

Now we need to find out how many mol of MnO<sub>2</sub> is needed per mol of MnCl<sub>2</sub> for the burn to reach this target temperature.

Assume the amount of dioxide needed is x mol per mol of chloride. The reaction products will then be: (1 + x) mol of Mn, 1 mol of MgCl<sub>2</sub> and 2x mol of MgO.

NIST thermochemical data allows to calculate how much enthalpy is needed to heat these reaction products to 1,800 K, using the relevant Shomate equations.

For Mn we obtain ΔH<sub>Mn</sub> = 58.75 kJ/mol, for MgCl<sub>2</sub> ΔH<sub>MgCl2</sub> = 133 kJ/mol, for MgO ΔH<sub>MgO</sub> = 74.4 kJ/mol.

The total heat of reaction ΔH<sub>R</sub> = - 161 - 684 x and in adiabatic conditions:

ΔH<sub>R</sub> + (1 + x) ΔH<sub>Mn</sub> + ΔH<sub>MgCl2</sub> + 2x ΔH<sub>MgO</sub> = 0


-161 - 684 x + (1 + x) 58.75 + 133 + 2x 74.4 = 0

and x = 0.06 mol of MnO<sub>2</sub>.

Since as we're not in adiabatic conditions and some heat losses are inevitable, a good starting point formulation (in mol) would be MnCl<sub>2</sub> = 1; MnO<sub>2</sub> = 0.1; Mg = 1.2

[Edited on 16-8-2008 by blogfast25]

Fluoride delema

chloric1 - 16-8-2008 at 17:36

Blogfast- The idea you proposed does not seem far fetched to me. In fact it seems like a logical reason. The reason I say this is that after the magnesium was consumed I was puzzle by the fact white smoke was forming. I did not care to breath it so I avoided it. Of coarse, as noted before, I had no fluoride at my disposal so I am wondering if the smoke was actually a cloud of potassium hydroxide formed by potassium oxide and water vapor. The physical properties of both potassium compounds are:

Potassium Hydroxde mp 380C bp 1324
Potassium oxide mp 350C with decomposition

I seriously doubt that the smoke is potassium oxide. Next and last run with dichromate booster I will try to condense some the smoke into an inverted beaker in hopes I can do a pH test to see if it is alkaline.

You make very convincing thermodyanic statements regarding possible outcomes of this system. I, now caught up in the whole thermodynamic calculations aspect of it now realize things on a new level. I have some conclusions of my own based on my limited research.

1) Alkaline earth oxides have higher heat of formation than their respective group one cousins.
2) oxidizers of group1 elements that may result in alkaline residues may not function well in thermite boosting compositions (sulfates, permanganates, chromates, persulfates, nitrates etc)

My proposal is if it is desirable to use an oxidizer with one of the above mentioned anions, maybe it should be of the group 2 ilk. Such as: Barium Chromate, calcium nitrate, calcium permanganate etc. Not sure about the sulfates though , will have to calculate that. to explain what I mean here, I have not yet looked at the heat of formation of calcium or barium sulfide. I know aluminum sulfide has lower heat of formation than the oxide but still is high enough that most sulfides can be reduced by aluminum exothermically.

One such composition I want to evaluate it molybdenite(MoS2) with aluminum:D

Of coarse for simplicity and cost one may use the more available sodium/potassium chlorate/perchlorate.

[Edited on 8/16/2008 by chloric1]

12AX7 - 16-8-2008 at 18:29

BTW, you should be getting potassium vapor. Alkaline oxides (and hydroxides) react with magnesium and aluminum, yielding alkaline metal vapor, which obviously will promptly burn in air. The thermite may simply be too bright to see the purple flame as it burns.


blogfast25 - 17-8-2008 at 06:03


I'm pretty convinced the smoke you see is K<sub>2</sub>O/KOH because the K vapour would oxidise immediately after leaving the crucible. The reaction isn't noticeable because the amounts are small. That's my take on it.

Oxides of metals of higher valence tend to have higher HoF all round, as a general and quite reliable rule. I suspect the lattice energy (Coulomb attraction - energy released when M<sup>m+</sup> and O<sup>2-</sup> form a crystal lattice M<sub>2</sub>O<sub>m</sub>;) is the main cause here: it's much higher for higher valence ionic lattices than for lower ones, leading to higher HoFs, despite the fact that the ionisation energy to obtain M<sup>m+</sup> is much higher than for M<sup>+</sup>.

Ca(NO<sub>3</sub>;)<sub>2</sub> is super deliquescent, so count that one out.

MoS<sub>2</sub>/Al is a very interesting proposition. The heat of reaction ΔH<sub>R</sub> for MoS<sub>2</sub> + 4/3 Al ---> Mo + 2/3 Al<sub>2</sub>S<sub>3</sub> is (at 298 K) a whopping - 841 kJ/mol of MoS2 (MoS2 HoF = - 276 kJ/mol, Al2S3 HoF = - 651 kJ/mol)!!

This might even be enough to reach the MP of Mo (2,896 K). If not, boosting with extra Al and S should do the trick. I have no C<sub>p</sub> or ΔH<sub>fusion</sub> for The Smelly One, so a little, simple experimentation would be needed. Caution not to get H<sub>2</sub>S poisoning. Bleach is a good way to 'neutralise' the rotten eggs gas because hypochlorite oxidises the S (-II) back to elemental S (0), so treating the aluminium sulfide slag with copious amounts of thin bleach is the way to get to the metal, make some sulfur flour in the process and avoid H2S poisoning. Win-win-win, I say...



Dichromate boosted Cr thermite:

Well, well, well: if at first you don't succeed, try, try, and try again...

I repeated the formulation with 0.4 mol of K2Cr2O7 but with the level of CaF2 cut to one tenth (CaF2 = 0.0225 Al), The formulation was thus: Cr2O3 = 1 mol; Al = 3.8666... mol; K2Cr2O7 = 0.4 mol; CaF2 = 0.087 mol. A 17.3 g batch of this mixture was mixed and lit.

It ran very fast and smoothly, resulting in a very flat slag puddle at the bottom of the crucible (that's always a very good sign). Breaking open, I found 1 regulus of Cr metal about 1 cm across (well formed), 1 of about 3 mm across and one of about 1 mm across. In the frozen slag even smaller droplets could be seen, too small to recover. The metal is clean-skinned, non-oxidised. It's chromium metal alright...

The mix contained about 6.60 g of Cr and the reguli weighed in at 4.40 g, a 67 % yield. Not bad, not bad at all for such a small reaction.

This would appear to confirm (but not prove) that interference of K with CaF2 might have been the problem here.

I will repeat a much larger batch with this formulation 'soon'.

[Edited on 17-8-2008 by blogfast25]

chloric1 - 17-8-2008 at 16:34

Well something to consider. I might try calcium oxide as a flux since the calcium aluminates have melting points between 1500 and 1600 degrees C. The potassium should not interfere with this. not sure what the melting point of the fluidized slag is but I am sure its comparable since I am sure the alumina raises the melting points of cryolite and fluorite.

I still am considering the molybdenite reduction but molybdenite BEGINS to sublime at 450 C according to Merk. Hopefully the reaction is fast enough before too much ore flies away!:o


Update! Bonus video!

Well, all of this browbeating and calculating has got me to kik back for some down to earth laughs and good times. Watch this video here This was some silly toy that came in a happy meal and my 3 year old HATED it calling it stupid bear.

[Edited on 8/17/2008 by chloric1]

blogfast25 - 18-8-2008 at 03:31

MoS2 begins to sublime at 450 C??? That sounds quite impossible, given its polymeric structure. Remember, synthetic MoS2 is used as a high temperature lubricant, an alternative to graphite no less... No, I can't see sublimation being a problem here...

CaO: definitely something to try...

What was the thermite used in the video? Looks like Classic Thermite (Fe2O3) to me...

chloric1 - 18-8-2008 at 14:02

That is Fe3O4 thermite in that last video. I have yet to try the Fe2O3 thermite. Interesting that in all these videos I am using aluminum powder that is 9 to 10 years old:o I know aluminum powder has been noted to react with water so I figured this stuff wouldn't burn. Well, I was wrong. I tried thermite when I first bought this powder in 1998 or 1999 and failed so put this stuff aside and forgot about it for awhile. Now I obviously know better and I don't have ignition problems.

I forgot to comment on aluminum sulfide oxidation with bleach in my last post so here goes. I don't know how long I will be able to obtain sulfur so this method could be quite important. I hoping the hydrolysis of the sulfide yields a gell like aluminum hydroxide so it can easily be removed from the sulfur residue with 16% HCl. Next run of Chromium thermite will be soon and it will be 0.14 mole potassium dichromate like the last run but will install reaction vessel in larger sand filled one. But a modification will possibly involve a CaO flux. After this it will be a chlorate boosted run. Then I will be almost out of chromic oxide for awhile and any remainder will get mixed with magnetite to make iron/chromium alloys.

I got two pounds of vanadium pentoxide and cuprous oxide in the mail today:D

blogfast25 - 19-8-2008 at 03:40

Aluminium powder is pretty resistant to degradation because it passivates.

Yep, it should be perfectly possible to separate the sulfur from the Al(OH)3 by dissolving the latter in dilute HCl, especially if you get to it fairly quickly... But it'll be a smelly business... You can't get garden grade sulfur in the US???

For chromium, my tests show 0.14 mol dichromate to be too low. Good results with 0.4 mol.

Hmmm, V2O5, one of my favourites: burns very fast and hot because it's a higher oxide (lots of alumina formed) - no booster needed...

This MSDS on MoS2 mentions an MP of 2375 C but says nothing about sublimation...

[Edited on 19-8-2008 by blogfast25]

chloric1 - 19-8-2008 at 08:37

I might try the higher level of dichromate to see. Last night I was shuffling through my CRC handbook and it gave the high melting point but ALSO said that sublimation begins at 450C. I am quite aware of molybdenites high temp lubrication properties. I roasted it a year or so ago to make molybdates and I did not notice any vapors exept SO2. I do not know why I am finding the data unless the amount that sublimes is miniscule.

AFAIK garden sulfur is still available but I am thinking of the future and this forum has been illustrative about how bad things are getting for home chemists outside the US especially in Australia and Europe. It seems the US is about 5 or years behind but we will catch up soon enough. Might next a few pounds of garden sulfur as growing season ends in September. Also would like to add that in desperation, very fine sulfur can be obtained from thiosulfate solutions with a little HCl added. Thiosulfate complexes many heavy metals forming complexes that are unstable to heat. They easily can decompose to there respective sulfides. This might of interest in thermite chemistry especially since one does not need to roast carbonates, hydroxides, or hydrated oxides to obtain a suitable oxidant for Al or Mg. Just filter and dry. Garden sufur needs to be dissolve by boiling toluene then crystallized. I can still order pure sulfur online but I should do the crystallization route just to get used to it. If they take away sulfur from online sources garden sulfur will soon follow.

[Edited on 8/19/2008 by chloric1]

[Edited on 8/19/2008 by chloric1]

blogfast25 - 20-8-2008 at 07:41

I'm not convinced that the US will follow suit on the restriction of chemicals for private usage: in Europe, but in the UK in particular, these restrictions have gradually grown and grown over the past decades, even before terrorism. In the US, AFAIK, there is more emphasis on personal freedom, at least in that respect. Here in the UK, knee-jerking is the standard reaction to even the minutest (or potential) problem. For those unfortunate enough to be interested in pyrotechnics (for instance) life has become almost impossible (and all this to try and prevent a few fools from blowing off their faces): in UK forums on pyro, legal matters have now become standard debating points. Sad but true...

DerAlte - 20-8-2008 at 21:54


A bit OT but...

I ought to have guessed you were from the UK. Things have indeed deteriorated badly. When I was a lad of 16-20, (1953-1957) I could get anything I wanted bar such things as arsenic and antimony oxides, mercuric chloride (but not mercurous, nor mercury metal), cyanide, and other compounds on the Poison Register. I used to get them from the local chemist (=pharmacist in UK) - even oleum. My father got the poisons by merely signing 'for experimental purposes'. Al powder. Mg powder, Na, K metal, chlorate, white P, red P, you name it. I had them all. Also any chemical glassware under the sun.

Yes, I was into pyro technics, too. My best ( worst?) effort was one Guy Fawkes night, when I set off a permangante, sulphur and Mg flash powder about 100 g worth in a cardboard tube reinforced by copper wire. It rattled the windows. Ibid, I tried same on the common in a forked tree and blew one fork off. In misty weather, the flash was spectacular!

I managed to gas myself with chlorine (2 days of lung ache), Bromine (worse!) and the carcinogenic chromyl chloride, to say nothing of phosphine, H2S et al. Happy days, denied to today's youth of all ages by the Nanny State, PC bullshit, and idiotic Greens (but not true environmentalists, of which there are few).

Enough, this thread is no the place to rant. Keep up the good thermite work, you and Chloric...

Best regards,

Der Alte

blogfast25 - 22-8-2008 at 09:43

Thanks, DerAlte, I've lived in several European countries and they're all quite Nanny state-ish when it comes to "better safe than sorry", but this one (UK) really does take the biscuit...

Anyroads, continuing with my "highly dangerous project" of trying to obtain lump metal from the reduction of anhydrous MnCl<sub>2</sub> with powdered magnesium, I ran another 20 g test today, using MnO2 as a heat-booster.

The molar formulation was: MnCl<sub>2</sub> = 1; MnO<sub>2</sub> = 0.2, Mg = 1.4

In terms of burn rate and heat this is starting to resemble what a successful thermite (this is more a hybrid, of course) looks like: very fast and regular, much more so than both previous attempts. Slag/metal had flowed more or less to the bottom of the crucible (an eggcup) but not completely and it was irregularly shaped and somewhat porous. Still, considerable fusion took place but no metal reguli were found.

Crushing the slag/metal mix, glistening metal could be seen in it and the crushed slag reacts with water as before: strong release of heat and quite some gas. Adding 32 w% HCl to the slag/water slurry results in a strong reaction and lots of hydrogen being released. The dissolved manganese formed will be recovered.

The reaction also produced lots of smoke. I'm wondering if residual NH4Cl is the (partly) cause. In previous reactions a lingering smell of ammonia could be observed (but not this time). The starting material anhydrous MnCl<sub>2</sub> also tests positive for NH<sub>4</sub><sup>+</sup> (add alkali and the smell of NH<sub>3</sub> is unmistakable). There is thus certainly residual salmiac present, so I'll have to assay the raw material for MnCl<sub>2</sub> content (gravimetrically via carbonate, I guess). Residual salmiac would also cool the reaction a bit, mainly because of the heat of evaporation...

Weather permitting, another test tomorrow with a higher level of KMnO<sub>4</sub>, which I haven't ruled out as a heat-booster yet... :)

12AX7 - 22-8-2008 at 10:06

How hot can you take MnCl2 before it melts or evaporates to nothingness? Most transition metal chlorides are rather volatile (for that matter, I wouldn't be surprised if a noticable amount co-sublimates with the NH4Cl, something to think about I guess). MnCl2 is an odd one out, perhaps resembling MgCl2 or CaCl2 more than others.


blogfast25 - 22-8-2008 at 10:42

Hi Tim,

MnCl2: MP = 654 C, BP = 1,225 C

This reaction [reduction of the chloride with magnesium] was used by E.Glatzel in 1889 to prepare the metal. I don't have the original article and the precise conditions unfortunately. But I reckon similar successful reductions have been carried out with more volatile oxidants. I'm aiming for yields of > 90 % of lump Mn metal. I'd prefer to work with the fluoride but can't for obvious reasons.

Also, since as I'm still below the MP of Mn (1246 C) at this point, significant blow-off of the chloride doesn't sound plausible...

Generally taking stock of the MPs and BPs of all reagents and all reaction products is of course good practice but perhaps we can push that too far too. Take the example of niobium thermite, the only method in use for industrial production of that metal (albeit followed by refining, of course): the BP of Al is 2,792 K, the MP of Nb 2,477K. To allow the metal to separate from the slag the reaction mix therefore needs to heat well beyond 2,477 K, in all likelihood beyond the BP of Al, yet Al boil-off isn't reported. It's to be assumed that the reduction reaction starts taking place at much, much lower temperatures...

As regards co-sublimation, the recorded weight losses from 4 or 5 fumings do not indicate any loss of MnCl2: if anything the weight losses seem too low compared to theo expectations. I might need to drive up temperature (or simply dry in a stream of dry HCl).

[Edited on 22-8-2008 by blogfast25]

[Edited on 23-8-2008 by blogfast25]

blogfast25 - 24-8-2008 at 08:10

It's also occurred to me that if I assume KMnO4 + 4 Mg ---> K + Mn + 4 MgO to proceed, then MnCl2 + 2 K ---> Mn + 2 KCl has also to be taken into account, as it's highly exothermic too...

12AX7 - 24-8-2008 at 10:44

Ah yes, so in total you might have something like...ahhh that would be...
2 KMnO4 + (n+8) Mg + (n+1) MnCl2 = (n+3) Mn + 2 KCl + 8 MgO + n MgCl2 (for n >= 0)
Tally up the HOF for those, in terms of the proportion n, and you should be sitting pretty accurately...


blogfast25 - 25-8-2008 at 07:13

Yes. Nicely put.

But simpler:

MnCl<sub>2</sub> + n KMnO<sub>4</sub> + (1 + 4n - n/2) Mg ---> (1 + n) Mn + n KCl + 4n MgO + (1 - n/2) MgCl<sub>2</sub> (n >= 0)

Similarly for the thermite of Cr<sub>2</sub>O<sub>3</sub>, boosted with K<sub>2</sub>Cr<sub>2</sub>O<sub>7</sub> because the K from K2Cr2O7 + 14/3 Al --->2 K + 2 Cr + 7/3 Al2O3 can also react as Cr2O3 + 6 K ---> 2 Cr + 3 K2O (the latter is highly exothermic). Here too an adjustment of the stoichiometry is needed to account for the reactivity of the K:

Cr<sub>2</sub>O<sub>3</sub> + n K<sub>2</sub>Cr<sub>2</sub>O<sub>7</sub> + 2 (1 + 2n) Al ---> 2 (1 + n) Cr + (1 + 2n) Al<sub>2</sub>O<sub>3</sub> + n K<sub>2</sub>O (n >= 0)

Aaaah... fun with stoichiometry! :)

blogfast25 - 9-9-2008 at 11:31

Work commitments and bad weather have so far prevented me from carrying out more tests with MnCl<sub>2</sub>, despite having a new batch ready. But I found time for a little theorising, in particular in an attempt to find other anhydrous chlorides suitable for reduction by magnesium, <i>in open reactor conditions</i> (and not bomb reactor conditions).

One drawback of Mg is that MgCl<sub>2</sub> for open crucible reactions is its relatively low boiling point (BP) of 1,685 K (1,412 C). It would therefore be impossible to obtain lump metal without the slag boiling (or "exploding") off if the melting point of the metal pursued was higher than 1,685 K, (unless in bomb conditions).

But even for metals with lower MPs there remains the possibility of a very energetic reaction, exothermic enough to heat the MgCl<sub>2</sub> well past its boiling point.

Take the case of a generic chloride, MCl<sub>n</sub> with a standard (298 K) heat of formation (HoF) of ΔH<sub>f, MCln</sub>. The standard reaction enthalpy ΔH<sub>R</sub> for the reaction MCl<sub>n</sub> + n/2 Mg ---> M + n/2 MgCl<sub>2</sub> is:

ΔH<sub>R</sub> = - ΔH<sub>f, MCln</sub> + n/2 ΔH<sub>f, MgCl2</sub>

with ΔH<sub>f, MgCl2</sub> = -642 kJ

The BP of MgCl<sub>2</sub> is 1,685 K and the enthalpy to heat 1 mol of M and n/2 mol of MgCl<sub>2</sub> to that temperature is ΔH<sub>M</sub> + n/2 . ΔH<sub>MgCl2</sub>, which from the relevant NIST Shomate equations can be calculated to be resp. ≈ 50 kJ (approx. for many metals) and n/2 . 122 kJ.

In adiabatic conditions: <i>∑ΔH = 0</i>, so:

ΔH<sub>R</sub> + ΔH<sub>M</sub> + n/2 ΔH<sub>MgCl2</sub> = 0

or: - ΔH<sub>f, MCln</sub> + n/2 ΔH<sub>f, MgCl2</sub> + ΔH<sub>M</sub> + n/2 ΔH<sub>MgCl2</sub> = 0

Reworked, the standard HoF of MCl<sub>n</sub> should be smaller than:

ΔH<sub>f, MCln</sub> < ΔH<sub>M</sub> + n/2 (ΔH<sub>f, MgCl2</sub> + ΔH<sub>MgCl2</sub>;)

or ΔH<sub>f, MCln</sub> < 50 - n/2 . 526 (kJ/mol)

as otherwise in open reactor (and adiabatic) conditions MgCl<sub>2</sub> will be heated past its boiling point.

With a HoF of - 481 kJ, it becomes immediately clear that MnCl<sub>2</sub> satisfies this condition.

Alas, few relatively easily accessible chlorides do meet it, in fact I've yet to identify a single one!

Here's a few that don't meet the criterion:

CuCl2: HoF = - 206 kJ/mol, n = 2
ZnCl2: HoF = - 415 kJ/mol, n = 2
PbCl2: HoF = - 359 kJ/mol, n = 2
FeCl2: HoF = - 342 kJ/mol, n = 2
FeCl3: HoF = - 399 kJ/mol, n = 3
AlCl3: HoF = - 706 kJ/mol, n = 3
CoCl2: HoF = - 313 kJ/mol, n = 2
SnCl2: HoF = - 333 kJ/mol, n = 2

ZnCl<sub>2</sub> would probably be the best choice, provided th reaction can be cooled a little with a heat sink.

blogfast25 - 17-9-2008 at 11:32

Finally another couple of tests designed to increase end-temperature of the MnCl2/Mg reduction.

One test was a 20 g batch of MnCl2/Mg, w/o any booster but pre-heated to 225 C for about 45 min in a gas fired oven. That burned remarkably fast but the slag was porous and no lump metal was found.

Secondly, a 20 g test with a high level of MnO2 booster. MnCl2 = 1 mol, MnO2 = 0.4 mol, Mg = 1.8 mol. This 20.0 g batch and (steel) crucible were accurately weighed before and after reaction.

The reaction was very fast and hot (comparably to a 'good' thermite) and VERY smoky. Weight loss during reaction was about 18 w%. The reaction products are 1.4 mol Mn, 1 mol MgCl2 and 0.8 mol MgO, or about 46.5 w% of MgCl2. It appears roughly half of the latter was blown off, presumably because MgCl2 has a relatively low BP (1,412 C) and thus a high vapour pressure at these temperatures. Evaporating MgCl2 would also cool the reaction because of the used latent heat of evaporation. (for KCl it's 79.1 kJ/mol but I haven't got a value for MgCl2).

The obtained slag/metal was extremely porous and could be crumbled by hand. Needless to say no lump metal was found.

I'm now convinced the volatility of the slag is the problem here, but I need first to eliminate with certainty volatility from residual ammonium chloride, the drying-aid.

If I'm right then obtaining lump metal from such a reaction would require a closed, pressure-proof reactor, to avoid the slag from volatilising. I'm thinking of a defunct RC plane engine, 10 - 20 cc. Or a lawnmower motor, something like that. With a sparkplug it's possible to engineer RC ignitions.

Ironically, the MnCl2/Mg reduction was meant to cure the problem of high Mn volatility in MnO2 thermites, while the remedy seems to present the "opposite" problem: evaporating slag!

blogfast25 - 20-9-2008 at 08:13

And as it happens, I've just found the abstract of Glatzel's original paper on the reduction of manganous chloride with magnesium:

Preparation of Manganese from Manganese Chloride and

By E. GLATZEL (Ber., 22, 2857-2859).

-Manganese can be prepared by heating a mixture of finely divided, anhydrous
manganese chloride (100 grams) and dry, powdered potassium
chloride (200 grams) in a covered Hessian crucible until it just
melts, and then adding magnesium (15 grams) in portions of
3-4 grams, at intervals of 2-3 minutes; if the fused mass is too
hot a very violent reaction occurs, and the contents of the crucible
are thrown out. The crucible is covered again, heated more strongly,
and then allowed to cool slowly in the furnace. The yield of manganese
is 20-25 grams, the metal containing traces only of silica,
and being quite free from magnesium.
The specific gravity of manganese, as the average of four determinations,
was found to be 7.3921 at 22". F. S. K.

So Glatzel basically heated molten MnCl<sub>2</sub> with powdered Mg in the presence of (rather a lot) of molten KCl flux/heat sink.

Without a doubt, the initially formed manganese is powdered metal and the second heating phase is designed to take the mixture of slag, possibly unreacted MnCl<sub>2</sub> and KCl to above the MP of manganese (1,246 C) and allow the metal to separate out.

Also, the high molar ratio of MnCl<sub>2</sub> to Mg (0.8 / 0.57) would help explain the near-absence of Mg in the obtained Mn. Theoretical yield: 31 g of Mn.

[Edited on 20-9-2008 by blogfast25]

chloric1 - 20-9-2008 at 10:35

Have yet to try the calcium oxide flux for my chrome thermite but recently read that calcium aluminates can melt as low as 1300 Celcius! After I originally posted this idea, it dawned on me that adding CaO(an alkali) to the dichromate(acidic) boosted thermite might cause side reactions before ignition. Of coarse siad calcium chromates should still yield appropiate calcium aluminate flux.

I can figure out the specific heat of calcium oxide and how much dichromate booster I will need to counter the heat sinking effect. Question is, do I really need stochiometric amounts of CaO to form the 1:1 aluminate or can I use substantially less to just "fluidize" the slag?

blogfast25 - 21-9-2008 at 08:24


Use substantially lower quantities of CaO: for a1:1 ratio of Al2O3/CaO the amount needed of the latter would be impractically high. If the Ca aluminate (or parts of it) has indeed such a low MP, it will act as a slag-fluidiser, even at modest doses.

Personally I've never tried lime as an additive in thermites but many literature references to that use exist. It's recommended for instance for MnO2/Al...

chloric1 - 21-9-2008 at 10:32

Well, it gets better. Supposedly, from a boron by thermite patent of 1962, calcium aluminate is easily soluble in dilute HCl. This was important considering boron does not separate in another phase. This would be really nice for getting clean vanadium,tungsten, molybdenum or silicon for that matter. If the aluminate can be dissolved with little fuss, it might be a source of CaCO3 and Al(OH)3 for latter.

blogfast25 - 22-9-2008 at 08:48

Very interesting [Ca aluminate being soluble in HCl].

The only problems I can see is that to obtain Ca aluminate only you need quite a bit of this inert (and heat sink) substance (CaO + Al2O3 --> Ca(AlO2)2, assuming this is the simple structure of Ca aluminate) and that this will have to be factored into the thermo calcs (end temperature!) The heat to carry the CaO to the MP of the metal (or slag, whichever is highest) will be very considerable and cannot simple be ignored...

Also, high levels of such (basically half a mol of lime per mol of Al) an inert substance may well mess with the kinetics, may make ignition harder and such like.

But for Si this is something I need to try, absolutely, I may in fact calc it tonight.

For V, the reaction with just Al (no CaO) works extremely well and leads to excellent metal coalescence, no need to remove slag chemically (see one spectacular V2O5 thermite over at, results very, very close to my own).

For Mo it's also an interesting proposition.

For W, it could be a way to produce the powdered metal, as the MP is far too high to obtain lump metal.

One could make some Ca aluminate (to study its properties) by igniting a mix of Al, KClO3 and CaO: overall reaction KClO<sub>3</sub> + 2 Al + CaO ---> Ca(AlO<sub>2</sub>;)<sub>2</sub> + KCl

12AX7 - 22-9-2008 at 09:20

Originally posted by chloric1
calcium aluminate is easily soluble in dilute HCl

Easily soluble, or easily reacts with?


blogfast25 - 23-9-2008 at 04:32


There isn't, at first glance, much to be learned from the Tinkerwebs on the properties of calcium mono aluminate, other than that its hydrolysis behaviour appears to be complex. It would sound plausible though that in certain circumstances this aluminate would react with HCl, to produce CaCl2 and AlCl3, analogous to aluminates of Group I. Wait and see, I guess...

For the hypothetical reaction of silica with Al in the presence of lime (overall reaction:

SiO<sub>2</sub> + 4/3 Al + 2/3 CaO ---> Si + 2/3 CaAl<sub>2</sub>O<sub>4</sub>

boosted with:

KClO<sub>3</sub> + 2 Al + CaO ---> CaAl<sub>2</sub>O<sub>4</sub> + KCl)

The overall stoichiometric formulation composition would be:

SiO2 .......................... 1 mol
KClO3 ....................... x mol
Al ............................... (4/3 + 2 x) mol
CaO ........................... (2/3 + x) mol

Applying a few reasonable assumptions a thermocalc shows that in strictly adiabatic conditions at x = 0.2 this mixture would reach approximately 2,500 K (well above the MP of alumina).

I will test this 'shortly' [cough!]...

blogfast25 - 24-9-2008 at 07:26

Interestingly, the lime route may even open up a new possibility of solving the MnO<sub>2</sub> thermite problem of Mn volatility, by cooling the reaction with lime. Provided the calcium mono aluminate can form somewhat <i>below</i> the MP of actual alumina, this would be a way of keeping the reaction temperature lower than the BP of Mn, yet have a nicely fluid slag all the same. Industrially, Mn is produced by thermite reduction of Mn<sub>3</sub>O<sub>4</sub> in the presence of lime. Unfortunately I don't have a pret-a-porter formulation at hand...

It would also be interesting to test the lime in a TiO2 thermite boosted with CaSO<sub>4</sub>, largely by replacing the CaF<sub>2</sub> slag fluidiser with CaO.

chloric1 - 24-9-2008 at 09:00

I was thinking that the calcium aluminate would lose the calcium to the HCL and leave part of the alumina as a hydrous jelly. Aqueous aluminum chemistry requires alot of free acid. So double theoretical amounts of HCl should do the trick.

The lime trick might work for manganese. What about using calcium chloride as a flux in small amounts?

blogfast25 - 24-9-2008 at 11:46

Originally posted by chloric1

The lime trick might work for manganese. What about using calcium chloride as a flux in small amounts?

That's generally speaking far too volatile (BP below 2,000 C) and would be blown off. Could be useful in chloride reductions though...

blogfast25 - 27-9-2008 at 06:18

Well, the two first attempts at using CaO (quicklime)/CaAl2O4 as a slag fluidiser weren't very successful at all.

Both were calcium sulphate boosted formulations and in the first I set the mol of CaO to half the mol of Al, so that the CaO/Al2O3 ratio was exactly 1. The thermite burned well but considerably slower than its CaF2 analog and the slag was very viscous and didn't even make it to the bottom of the crucible. There was no lump metal whatsoever.

Reasoning the high CaO level was perhaps proving too much of a heat sink, in the next test I cut it right down to CaO/Al2O3 = 3/8, less than half the original dose. This one burned faster and hotter, with more (but not all) slag collecting at the bottom of the crucible but no lump metal was found at all...

In both cases the slag was properly fused and completely non-porous. In the first test the slag composition was approximately (in mol) CaS/Al2O3/CaO = 0.5 / 4/3 / 4/3. A lump of it, dunked in 32 w% HCl did react slightly at first and H<sub>2</sub>S could be smelled. On the plus side, 24 h later and the (approx.) 5 g lump of slag had completely disintegrated and mostly actually dissolved. The solution also tests strongly positive for the target metal.

It's not clear whether this dissolution is partly due to hydrolysis of the CaS or due to hydrolysis of the alleged Ca aluminate (or both), as I don't remember how this type of slag (but minus the CaO) normally behaves in HCl.

The next test will be in a chlorate (no CaS is then formed) boosted SiO2 thermite, possibly in conjunction with a small amount of CaF2. If lump metal forms and the slag does indeed dissolve in HCl then the Si-metal(loid) recovery problems I've been having may be solved by use of CaO. But right now I'm waiting for a new consignment of Al powder...


Here's an interesting thermite based patent (EP19920106428) abstract, that, for once, actually gives some detail on the formulations used.

<i>"Ferroniobium was produced, using niobium ore (niobium oxide) and iron oxide as starting metallic oxides, aluminum as a reducing agent, sodium chlorate as an exothermic agent and a mixture of fluorite and quick lime as a slag forming material."</i>

In addition they also used small quantities of size-reduced ferroniobium (FeNb), as a heat sink (not entirely sure why).

One of the formulations gives a good idea of the ratios and amounts of CaF2/CaO used (quantities in kg):

Nb ore .......................... 1,000
Fe ore ........................... 162
Al ................................... 350
NaClO3 ........................ 50
CaO .............................. 60
CaF2 ............................. 80
FeNb ............................. 50

So the quantities of slag former and slag fluidiser used are really quite small (but Nb2O5 is a real scorcher, due to its high oxygen content).

[Edited on 27-9-2008 by blogfast25]

[Edited on 27-9-2008 by blogfast25]

blogfast25 - 2-10-2008 at 09:04

A few more tests with CaO at a lower level and in combination with CaF2 didn't really yield anything useful, at least not in sulfate boosted reactions.

I decided to give it one last shot in a silicon thermite, potassium chlorate boosted. The formulation was:

SiO2 ............................... 1
KClO3 ........................... 0.36
Al ................................... 2.05
CaO ............................... 0.72
CaF2 .............................. 0.41

The CaO/Al2O3 target ratio was 0.70 (I was trying to get 1 but was a little short of lime).

A 20 g batch burned quite hot but no slag collected at the crucible bottom, rather a greyish, porous mass was formed. And then something strange happened.

Dunking the lightly crushed slag in 30 w% HCl, I noticed strong formation of gas, presumably hydrogen. This would be very unusual as silicon doesn't react with HCl and if the reaction had proceeded stoichiometrically there should be no free Al worth speaking of... I left the beaker and contents to stand for a few minutes. After I came back, a hot column of foam had started to flow over the beaker's rim and crackling noises could be heard. In fact, miniature 'explosions', including very small but clearly visible flames could be observed. I suspect that these were due to silane (SiH<sub>4</sub> and possibly higher silanes) spontaneously bursting into flame upon contact with air. There was also considerable generation of heat.

Possibly the lime reacts partly with the silica, forming a Ca silicate of sorts, which may be harder to reduce than pure silica. Some unreacted Al would then be able to alloy itself with any silicon formed, forming an aluminium silicide of sorts, which with HCl would evolve silane(s) and hydrogen...

12AX7 - 2-10-2008 at 19:38

Aluminum and silicon form a simple eutectic system, no intermetallics ("silicide").

Possibly a calcium silicide?


DerAlte - 2-10-2008 at 21:31

@Blogfast & chloric, the foremost exotic thermiters:-

I was recently surprised to read - in a "usually reliable" text - that both Ba and Sr can be produced by a thermite process (but no references given.)

Sr has MP=777C, BP=1382C (a bit low).
Ba has MP=727C, BP=1897C (better),
both according to CRC.

Haven't looked at the energy angle, but both metals are rather energetic.

Ever thought of these two?

Regards, Der Alte

12AX7 - 3-10-2008 at 06:48

Brauer gives prep with aluminum-containing briquettes under heat and vacuum. Not really exothermic so not really fair to call it thermite.


blogfast25 - 3-10-2008 at 07:00

@Tim and DerAlte:

Ignoring for a minute the entropies, the reaction enthalpies (ΔH) for MeO + 2/3 Al ---> Me + 1/3 Al<sub>2</sub>O<sub>3</sub> (with Me = Ca, Sr or Ba, resp. HoFs - 635 kJ/mol, - 592 kJ/mol and - 548 kJ/mol) are all positive, so no spontaneous change of state is to be expected (but things might be different at much higher temperature if the entropic terms (- TΔS) kicks in - without gaseous reaction products the entropic term is usually really small though).

Certainly I wouldn't expect CaO, SrO or BaO to react with Al in backyard thermite conditions (what might be achieved by strongly heating such mixtures and by siphoning off the volatile earth alkali metal reaction product, assuming some forms in equilibrium, is a different matter). DerAlte, it may be possible that the thermite references to SrO and BaO may have been in reference to that: strongly heated mixtures of Al and MeO, with removal of any formed Me.

Magnesium, at high enough temperature and subject to removal of both the formed CO and Mg can be produced carbothermically from MgO and cokes, for instance... For an aluminothermic high temperature process to produce Sr or Ba, low BP could actually be an advantage...

Tim, I can't really see how else to explain the formation of silane(s), as formation of Ca in significant quantities seems highly unlikely in these conditions. Wouldn't an alloy of Al and Si, perhaps in particular ratios, generate some silane? Just a guess...

[Edited on 3-10-2008 by blogfast25]

 Pages:  1  2    4    6