Sciencemadness Discussion Board - Calcium Sulfate and Titanium Thermite

Calcium Sulfate and Titanium Thermite

MrHomeScientist - 28-8-2012 at 12:54

For the last few weeks, I’ve been trying to follow blogfasst25’s procedure (outlined on his blog: http://developing-your-web-presence.blogspot.com/2008/10/on-... ) for making titanium thermite. In this composition, a heat booster is required to keep the reaction going and to ensure the products are all molten (necessary for good metal/slag separation). He uses calcium sulfate and extra aluminum powder. I have had some troubles getting this to work properly, and I thought I’d post the details of my efforts in case others have had the same issues.

My source for calcium sulfate was DAP brand plaster of paris, which is the hemihydrate (CaSO4 * 1/2H2O). The ingredients listed on the back are calcium sulfate, calcium carbonate, and silicon dioxide. Assuming the latter two were only there in small amounts, I set off using this plaster straight out of the box. After heating some of it in my oven for 2 hours at 425 F to dehydrate it, I mixed up a batch of thermite according to blogfast’s specs:

TiO2 – 30g
CaSO4 – 25.5g
Al – 27g
CaF2 – 17.5g

Here’s how it went: http://www.youtube.com/watch?v=KfqKOaZyllQ
That video is actually my second try, which used the extra booster formula (see the video’s description for specifics). As you can see, the burn was very “sputtery.” While it did produce titanium metal, the pieces were small spheres that didn’t coalesce together like they should if the whole thing were molten. The slag was also very sparkly throughout, likely from tiny pieces of Ti that were trapped before they could flow together. You can see in the picture at the end spheres of white titanium embedded throughout the slag. Apparently the impurities, likely added as fire retardants or to improve its setting properties, were more detrimental than I thought. Here’s the Ti I recovered from this batch:

I looked up the MSDS for this particular brand of plaster: http://www.dap.com/docs/msds/00071008_english.pdf
Turns out, the impurities are pretty significant – CaCO3 at 15% – 25%, and SiO2 at 0% - 1%.

So, to remedy this I tried removing some of the additives. The silica isn’t worth the effort needed to remove it, but the CaCO3 is easy enough to destroy with a treatment in acid. As a bonus, by using sulfuric acid I make more of my desired reagent!

CaCO3 + H2SO4 == CaSO4 + H2O + CO2

I took some of the raw plaster, submerged it in a good amount of water (so it wouldn’t harden up on me), and added roughly 9M sulfuric acid slowly until the bubbling stopped. I filtered off the CaSO4, let it air dry, and then dehydrated the powder in a fused silica crucible over my small butane burner for about 45 minutes. Using this in the same thermite composition listed above, the reaction ran much smoother: http://www.youtube.com/watch?v=DSY4biEB88Y
When recovering the metal afterward, the pieces were much larger and I found 4 or 5 individual pieces about ½” in diameter. This is indicative of a much better reaction temperature, but there were still some problems with this batch. There were a number of large voids in the slag, the Ti metal was incredibly hard to separate from the adhering slag, and the metal pieces were much more tarnished.

My hypothesis is that all of these are due to water of crystallization still in the CaSO4, i.e. I didn’t heat it up enough. The water boiling off would cause the voids, and I suspect it would react with the hot Ti metal to tarnish its surface. The tarnish can be easily polished off, of course, but I feel that I can do better.

=====================

My next attempt was to make my own CaSO4, so I know it’s pure (or at least, I know for certain what’s in it!) I tried two methods to do this.

Method 1: CaCl2 + H2SO4 == CaSO4 + 2HCl
This uses calcium chloride and sulfuric acid, both of which are fairly OTC for me – CaCl2 is used in Damp Rid brand moisture absorbers, and my sulfuric acid is from Liquid Fire brand drain opener. It also produces hydrochloric acid as a byproduct, so this might be a handy way to get much higher purity acid than hardware store brand muriatic acid (which is highly contaminated with, mostly, iron).

Method 2: MgSO4 + CaCl2 == CaSO4 + MgCl2
I like this method a lot because it uses nothing hazardous at all, and both reactants are very easy to source: magnesium sulfate is Epsom salt and calcium chloride is used in Damp Rid.

I tried both methods, and they both work exceedingly well. As long as you use dilute solutions of everything (mine were <5M), the precipitation doesn’t happen immediately. That means it is much less likely to form occlusions of impurities and you end up with a beautifully snow white product. I then heated both samples to 480F for two hours to dehydrate – here they are fresh out of the oven:

The Method 2 sulfate is on top, and Method 1 on bottom. The latter was quite a bit more powdery and a little purer white, which is interesting considering it was made from acid that has a dark red color to it.
I was able to use all the sulfate from Method 2 for a single 100g Ti thermite charge. The result can be seen here: http://www.youtube.com/watch?v=6TFty76NyHw
This one was a little slow to start, but burned fairly well once it got going. The slag was especially nice, conforming to the bottom of the flower pot very nicely with only one small void – meaning it was completely molten and (mostly) free of water. The pieces I recovered were about halfway between the first two reactions – larger spheres than the first run with their somewhat shiny cast, but smaller than the lumps of the second.

So that’s where I’m at now. Ideally, I want to produce large, shiny lumps of Ti for my element collection, but I think these will do nicely for now. If anyone else is trying this reaction, I hope my posting my efforts here was helpful!

blogfast25 - 28-8-2012 at 14:01

Very nice work and write-up, MrHS, thank you. I'm convinced the formulation needs a little more CaSO4/Al booster, just giving it that extra bit of end temperature to operate in your 'local conditions'. Let me come up with a suggestion tomorrow.

[Edited on 28-8-2012 by blogfast25]

elementcollector1 - 28-8-2012 at 14:17

Are you kidding?! Those Ti spheres look beautiful! Why wouldn't you use those as your element sample?

bbartlog - 28-8-2012 at 19:32

I think it's very interesting that the bits of Ti from the first run are so much shinier than those from the later, more productive trials. Is it possible that the lower oxygen balance of that mix was the cause? If so you could try to tweak the formulation so that it's somewhere between the first trial and the later ones (i.e. use slightly less than stoichiometric CaSO4) and see whether there is a happy medium to be found...

blogfast25 - 29-8-2012 at 04:16

MrHS:

Currently your [my] formulation is, in mol:

TiO2 ……………….. 1 mol
Al ………………..2.666… (rec.) mol
CaSO4 ……………… 0.5 mol
CaF2 ………………… 0.6 mol

Which is perfectly stoichiometric for both reactions occurring.

My strong impression, weighing all the evidence before me, is that the formulation still runs a little too cool in your local conditions. I suggest to increase the sulphate level by 0.1 mol (add 0.2666… mol Al also) or by 0.2 mol (add 0.5333… mol Al also). Higher end temperature allows more time for the metal to coalesce before the melt freezes up. And a faster reaction gets all the chickens back in the henhouse faster, where they can congregate and coalesce.

All in all the reaction in the video runs remarkably similar to my own, but mine have the edge in speed and smoothness, I think.

With CaSO4 I’ve always obtained globules though, whereas KClO3 does give slabs. I believe this is due to a surface tension effect.

With KClO3 I get metal that is as black as the ace of spades but it’s very, very superficial: buff the metal with a rotating steel brush (on a drill) or such like; highly shiny, silvery metal will be yours.

Remind me again what ignition system you're using?

MrHS, ultimately to make the experience complete you’ll have to try KClO3 or KNO3/NaNO3 as booster systems!

[Edited on 29-8-2012 by blogfast25]

[Edited on 29-8-2012 by blogfast25]

[Edited on 29-8-2012 by blogfast25]

MrHomeScientist - 29-8-2012 at 08:23

elementcollector1: They do look beautiful, I agree! I would just love to be able to produce larger pieces of similar quality

I'll certainly be using them in the collection until I come across something better.

blogfast: Thanks for the advice. I need to make more sulfate, but once I do that I'll try it out with extra booster. Another thing that's bothered me about this particular thermite is how clumpy the CaSO4 and, especially, the TiO2 can be. This makes it a little hard to mix fully, so my final mix always has bits of white in it. That might lead to some of the sputter.

My ignition system is, essentially, Mg ribbon. What I do is make a 'volcano' of thermite, add a small pile of potassium permanganate to the top, and push a piece of Mg ribbon into that so it just pokes out the top but also reaches down into the thermite. Adding glycerin causes it and the permangante to ignite, which lights the Mg ribbon, which starts the thermite. I really like this method because it's worked every time, I don't have to break out the blowtorch to light the ribbon, and it gives me ample time to walk away before the reaction starts.

I'd love to try those other boosters at some point, now that my understanding of thermite is starting to develop. I don't have any chlorate unfortunately - what I need to do is look into making it via electrolysis (another subject that I don't know much about). I do have KNO3 - I'm assuming these other boosters still react with excess Al, correct?

elementcollector1 - 29-8-2012 at 10:08

All I have for my titanium piece is a spork. Granted, sporks are cool, but...
Also, I need someone's advice on my Al powder, as every thermite I've tried with it to date has failed. It's made from foil, and ball milled for a while, and has that dark gray color (almost black) of German Dark. Is there something in the ingredients that causes the Al to be unreactive?

For chlorate, you almost certainly need a platinum, MMO, or lead dioxide anode. Add in a bunch of KCl, let run for a few days, and you should be collecting crystals of chlorate in no time. Or, boil some bleach for sodium chlorate, and react with KCl to precipitate potassium chlorate (less soluble). Recrystallize as necessary.

blogfast25 - 29-8-2012 at 11:49

MrHS:

Yeah, TiO2 can be a bit lumpy. I always sieve off the lumps with a tea strainer. Set lumps aside for later use.

For a tested starting point formulation based on KNO3 boosting, replace the CaSO4 by KNO3, mol per mol.

Adjust the Al content slightly as follows:

1 mol CaSO4 uses 8/3 mol Al, so 0.5 mol uses 4/3 mol Al.
1 mol KNO3 uses 2 mol Al, so 0.5 mol uses 1 mol Al.
So reduce the overall Al content of the formulation by 4/3 – 1 = 0.333… mol.

EC1:

Ball milled or otherwise size reduced Al foil should be fine, not sure why yours doesn’t work. German Black has charcoal in it. Avoid for thermits.

[Edited on 29-8-2012 by blogfast25]

AndersHoveland - 29-8-2012 at 12:08

My experience is that magnesium/calcium sulfates make only very reluctant oxidizers in thermite mixtures. The oxygen seems to be held tightly within the sulfate groups. Perhaps the reason is that the thermal decomposition point of sulfates are much higher than that for Fe2O3 or CuO. The sulfate may mostly only be acting as an oxidizer in its molten phase, which is much less efficient than decomposing into oxygen gas that can pass outward and have a higher surface area with the reducing metal powder.

Just because sulfates contain four atoms of oxygen does not mean they make good pyrotechnic oxidizers.

blogfast25 - 29-8-2012 at 13:10

Quote: Originally posted by AndersHoveland

My experience is that magnesium/calcium sulfates make only very reluctant oxidizers in thermite mixtures. The oxygen seems to be held tightly within the sulfate groups. Perhaps the reason is that the thermal decomposition point of sulfates are much higher than that for Fe2O3 or CuO. The sulfate may mostly only be acting as an oxidizer in its molten phase, which is much less efficient than decomposing into oxygen gas that can pass outward and have a higher surface area with the reducing metal powder.

Just because sulfates contain four atoms of oxygen does not mean they make good pyrotechnic oxidizers.

As so often you talk baseless bullsh*t. Your label ‘Giver of bad advice’ remains apt. I sometimes wonder if your part here is to simply try and disrupt things.

Compare any CaSO4 based booster system to more classic KClO3 or Na,KNO3 based system and there’s hardly a cigarette paper between them. I KNOW because I’ve done it.

And simple stoichiometric mixtures of CaSO4 and Al powder burn like hell. I know because unlike you I’ve actually done it.

I’m convinced you have ZERO practical experience with such booster systems in thermits or elsewhere.

You’re an insufferable TWIT.

And here’s a very public warning: keep up with your charades and I will make a very convincing case to the mods for having you banned permanently from this club. Capisce?

[Edited on 29-8-2012 by blogfast25]

elementcollector1 - 29-8-2012 at 14:21

Can MgSO4 even be used in CaSO4's place?
I do believe the reason the Al is so very faulty is because, according to numerous sources, the foil is actually coated in plastic (making it useless for melting). Perhaps I should melt some aluminum cans, then take a hacksaw to those.

hyfalcon - 29-8-2012 at 15:49

This is the aluminum I use in all my thermite/flash experiments.

http://alphachemicals.com/aluminum_powder

blogfast25 - 30-8-2012 at 03:40

Quote: Originally posted by elementcollector1

Can MgSO4 even be used in CaSO4's place?
I do believe the reason the Al is so very faulty is because, according to numerous sources, the foil is actually coated in plastic (making it useless for melting). Perhaps I should melt some aluminum cans, then take a hacksaw to those.

Yes, MgSO4 would work too. But Epsom Salt is highly hydrated.

Al fois isn't coated plastic, AFAIK. There's no need: Al passifies; there's no need for further protection.

Are you sure your thermits are formulated properly?

MrHomeScientist - 30-8-2012 at 05:25

Quote: Originally posted by elementcollector1

I don't believe Al foil is plastic coated, but I know soda cans are. There's the label on the outside, of course, but there is also a plastic lining inside that prevents corrosion from the acidic soft drink. Melting the cans would probably burn off some of this and leave the rest as a slag, so you should be alright going that route.

blogfast25 - 30-8-2012 at 05:44

Quote: Originally posted by MrHomeScientist

Apparently the way it’s done industrially is to dunk the Al junk into a eutectic bath of KCl-NaCl. The molten Al sinks to the bottom (where it’s regularly tapped off) and the paint, paper, plastic etc floats on top as dross which is periodically removed.

Melting down small amounts of Al is harder than you might think, as my experience showed. You need to get a bit of a bath going and because the molten metal forms an oxide layer that can take a while to happen. But once you’ve got a few cm3 molten then you can feed small chunks into that and they will melt like ice in water, if you keep heating the melt. Start with the cleanest metal you’ve got. Hammer down some Al foil balls for instance.

watson.fawkes - 30-8-2012 at 09:22

Quote: Originally posted by blogfast25

Melting down small amounts of Al is harder than you might think, as my experience showed. [...] Start with the cleanest metal you’ve got. Hammer down some Al foil balls for instance.

In addition, heat the raw material to drive off residual and adsorbed water. Water dissociates in molten aluminum. This reaction is driven by the relatively high solubility of H₂ in molten aluminum. The oxygen reacts forming alumina, a large component of slag. If you're casting aluminum, dissolved H₂ is the dominant source of porosity.

You can also use fluxes when recovering Al in the backyard foundry. This page (found with a search) provides a reasonable introduction.

blogfast25 - 30-8-2012 at 11:45

Well, well, there's a whole world of fluxes out there!

CrossxD - 6-2-2016 at 04:41

where I can find calcium fluride? or can I use other flux?

[Edited on 6-2-2016 by CrossxD]

j_sum1 - 6-2-2016 at 05:37

I bought some from onyxmet for the purpose but have not used it yet. I didn't get much and it was a bit pricier than I would like.
The reaction definitely needs a flux. My last attempt did give me Ti but it was intermingled with everything else. You also need to dry the hemihydrate calcium sulfate. Trying to drive off the water during the thermite reaction is enough of an energy burden to make a big difference to your product.

blogfast25 is definitely the expert on this particular reaction. Although he was vague in his previous post and referring to Al recycling, I am certain that he could recommend a few alternative fluxes for the thermite.

blogfast25 - 6-2-2016 at 06:51

Quote: Originally posted by j_sum1

blogfast25 is definitely the expert on this particular reaction. Although he was vague in his previous post and referring to Al recycling, I am certain that he could recommend a few alternative fluxes for the thermite.

The one to try is definitely quick lime - CaO. CaO and alumina form a number of low melting aluminates, aiding slag fluidity.

I tested this recently in the hardest of them all - MnO₂ with 'good' results. I say 'good' in the sense that MnO₂ thermites rarely give more than 30 % metal yield and the one with CaO flux did about the same.

It would be really interesting to test CaO in a TiO₂/CaSO₄ boosted formulation! :cool:

Going by the Wikipedia entry, the lowest melting calcium aluminate is for a CaO/alumina molar ratio of about 2/1. But that would make for a very high CaO loading. I think the optimal loading must be something like Al powder/CaO molar ratio between 4 and 2 in the thermite formulation.
<hr>
I don't see many other candidates as slag fluidizers in aluminothermy besides the trusted CaF₂ and CaO. Slag fluidisers need to be inert (irreducible by Al), very high BP and with an MP significantly below 2000 C, to be effective as fluxes. There's not that many compounds that jump to mind, meeting those criteria.

[Edited on 6-2-2016 by blogfast25]

RogueRose - 6-2-2016 at 09:01

I'm doing something similar and have a question about the purity of my CaSO4. I suspect there is some NH4NO3 in it at at most 2.5% max concentration of the CalSul. I wouldn't think this would cause any mayor problems as it seems that it should add some extra fuel but at a low enough concentration that it shouldn't pose any problem with causing an explosion of any kind.

could this cause any problems with the mixture such as causing ignition or burn problems?

blogfast25 - 6-2-2016 at 09:19

Quote: Originally posted by RogueRose

I'm doing something similar and have a question about the purity of my CaSO4. I suspect there is some NH4NO3 in it at at most 2.5% max concentration of the CalSul. I wouldn't think this would cause any mayor problems as it seems that it should add some extra fuel but at a low enough concentration that it shouldn't pose any problem with causing an explosion of any kind.

could this cause any problems with the mixture such as causing ignition or burn problems?

Not in my opinion.

careysub - 6-2-2016 at 11:18

Quote: Originally posted by CrossxD

where I can find calcium fluride? or can I use other flux?

I see it on eBay for not unreasonable prices.

People use sodium fluoride as a flux, which is available cheap from pottery supply places.

If calcium fluoride it must be, you could make it by dissolving the sodium fluoride in water and precipitating with a soluble calcium salt (calcium chloride, perhaps).

BTW: I will point out (form early comments on this thread) that most plaster-of-paris compositions (and DAP in particulare) are 25% calcium carbonate, not an insignificant amount.

[Edited on 6-2-2016 by careysub]

blogfast25 - 6-2-2016 at 12:25

Quote: Originally posted by careysub

People use sodium fluoride as a flux, which is available cheap from pottery supply places.

If calcium fluoride it must be, you could make it by dissolving the sodium fluoride in water and precipitating with a soluble calcium salt (calcium chloride, perhaps).

BTW: I will point out (form early comments on this thread) that most plaster-of-paris compositions (and DAP in particulare) are 25% calcium carbonate, not an insignificant amount.

NaF, with a BP of about 1700 C simply boils off in Thermites.

NaF is poorly soluble, about 1 M limit, so not ideal for that purpose.

My no-frills wall-filler contained no CaCO₃ whatsoever. It's easy to test for and easy to remove, if needed.

careysub - 6-2-2016 at 13:41

Quote: Originally posted by blogfast25

NaF is poorly soluble, about 1 M limit, so not ideal for that purpose.

I'm not sure that is such a problem since the desired product is the precipitate.

I have had problems trying to do metathesis with more concentrated solutions (ammonium sulfate and calcium nitrate) where a massive unworkable water-holding precipitate formed. The recommendation I encountered there was to use a much more dilute solution (and which made it much less attractive since the product remained in solution).

A gallon of water would dissolve 160 grams of sodium fluoride, yielding 210 grams of CaF2. Mix it up in an old 1 gallon milk jug and put it aside to settle seems pretty easy to me. Do it as often as needed to make more.

blogfast25 - 6-2-2016 at 13:56

Quote: Originally posted by careysub

I'm not sure that is such a problem since the desired product is the precipitate.

Ever precipitated an insoluble fluoride? Going by NdF₃ and CaF₂, the precipitates are quite slimy, gelatinous barst**ls. No sandy, easily filtered off substances.

What you suggest is possible, just not very easy.

j_sum1 - 6-2-2016 at 16:03

I used CaO as a flux. (Admittedly on this run, I did not dehydrate my calcium sulfate.)
The reaction was quite interesting. I got a clean-burning reaction with no sputtering. It seemed to be very hot but not at all violent.

The product was a single cake of solid material about 20mm thick and 80mm diameter. Maybe a bit like a hockey puck. It was light coloured and dusty on the surface. Breaking it open showed that the light coloration was only a mm thick. The interior was a charcoal grey colour with fine needle-like Ti crystals lined up in parallel through the puck. They were finer than the thickness of a hair and sparkled somewhat in the light. But they were obviously embedded into the surrounding slag. I didn't get any nice nuggets of metal.

Interestingly also, there was no sulfide smell -- not even much when treated with acid.

blogfast25 - 6-2-2016 at 17:10

Quote: Originally posted by j_sum1

I used CaO as a flux. (Admittedly on this run, I did not dehydrate my calcium sulfate.)
The reaction was quite interesting. I got a clean-burning reaction with no sputtering. It seemed to be very hot but not at all violent.

The product was a single cake of solid material about 20mm thick and 80mm diameter. Maybe a bit like a hockey puck. It was light coloured and dusty on the surface. Breaking it open showed that the light coloration was only a mm thick. The interior was a charcoal grey colour with fine needle-like Ti crystals lined up in parallel through the puck. They were finer than the thickness of a hair and sparkled somewhat in the light. But they were obviously embedded into the surrounding slag. I didn't get any nice nuggets of metal.

Interestingly also, there was no sulfide smell -- not even much when treated with acid.

That's very interesting and hard to explain. The fine crystals point to lack of time to coalesce the metal, of course. But why?

It would be worth trying this again with a KClO₃ or KNO₃ boosted TiO₂ thermite.

I've been meaning to test this in a Classic ferric oxide thermite, CaF₂ and CaO, back-to-back comparison... One day...

[Edited on 7-2-2016 by blogfast25]

j_sum1 - 6-2-2016 at 17:24

My thoughts were that it didn't get up to a high enough temperature. Or more to the point my setup had a strong thermal gradient which lead to directional crystal growth and insufficient time for the metal to coalesce. Working with the hydrated CaSO4 can't have helped.

Unfortunately I haven't had nearly as much time in the lab as I would have liked to experiment further. The KNO3 version is definitely on the list.

blogfast25 - 6-2-2016 at 18:01

Quote: Originally posted by j_sum1

My thoughts were that it didn't get up to a high enough temperature. .

Your earlier observation:

Quote:

The product was a single cake of solid material about 20mm thick and 80mm diameter. Maybe a bit like a hockey puck.

Doesn't really support that (all slag was molten and found at the bottom) but it's possible end-temperature was a bit lower and that reduces cooling time and thus also time for the metal to coalesce into worthwhile puddles.

I noticed systematically that with TiO₂/KClO₃/CaF₂ I always obtained one single regulus, with CaSO₄/CaF₂ always several.

Lots of 'fiddling' with formulations possible, so little time...

j_sum1 - 7-2-2016 at 02:51

Hmmm.
I had another look at the result this afternoon and took a pic. Unfortunately, I had crushed up most of it including the best looking pieces and the photo is just from my antiquated phone. It does not really show well the needle-like crystals that went from one side to the other-- at least in some of the pieces.

I am not really sure what to make of it. I'm pretty sure that there is metallic Ti there but it is definitely not the result I was hoping for. I think the next attempt will be using KNO3. I will probably stick with the CaO flux for a while since I have plenty and i do not have much CaF2. I will attempt the same with silicon as well. (But I think I need some better sand. Last attempt had a really low yield using Al/S. And the byproduct was decidedly green -- almost Cr(III) colour. I have no idea what was in it.)

blogfast25 - 7-2-2016 at 07:10

There seems to be metal in there but very little. Very strange...

macckone - 7-2-2016 at 10:23

Source of fairly pure calcium fluoride.

http://www.nmclay.com/ProductDesc.aspx?code=FLUOR&ty...

This is an artificial product rather than mined from what I can tell.

MrHomeScientist - 8-2-2016 at 07:00

Minimum order 10 lbs! That's a lot of thermite!

careysub - 8-2-2016 at 07:48

Maybe someone would like to order and split it? I would take some.

semiconductive - 19-6-2018 at 16:55

Quote: Originally posted by j_sum1

Sorry to resurrect this after so long, but I was very curious:
Did you ever get a chance to try again?

I bought 20LBS of TIO2, for $10. Aluminum is about $2/LB. I don't have a lot of CaF2.
So, I was wondering about carbonate fluxes, or perhaps boric acid; when I came across this thread.

In the original post, CaCO3 was a contaminant; but that becomes CaO under heat.
Some people also say that CaCO3 will burn with aluminum powder; which means it would create the flux at the same time as the thermite reaction proceeded. I have both CaCO3 and MgCO3 to experiment with; and I know how to microvawe calcine them into impure CaO and MgO at home. I also have slaked lime (CaOH).

But I was curious, did the CaO flux ever work out for you ?

The main issue with CaCO3 contamination was that it caused spattering because it out-gassed. But, if the thermite reaction was done in a deep hole and buried in sand; it wouldn't explode (sand is used for casting steels, and outgassing is common) but I was wondering if that might contain the spattering.

j_sum1 - 19-6-2018 at 19:04

Funny you should mention. Answer, not yet. Shortly after that experiment my lab went into boxes for two years. I have a significant backlog of interesting experiments to do.

Just today I showed MrHomeScientist's thermite videos to a science class and we will do a couple of basic ones later this week.

I recommend watching the thermite series by the gayest person on youtube. Really good stuff. One of the things he plays with is various mixtures of fluxes/slags. It is pretty insightful. I am like you: having some oxides and Al powder but not a whole lot of CaF2. A cheaper and more accessible flux would be welcome. My CaF2 took a while to locate and was quite pricey.

semiconductive - 20-6-2018 at 15:26

Quote: Originally posted by j_sum1

...
I recommend watching the thermite series by the gayest person on youtube.

I found the videos for the Exotic Thermite, videos 1,2,3. Video 3 Has Ti, in it @ 5:45 minutes in.

https://www.youtube.com/watch?v=LsdesMWC37g

There is supposed to be a fourth video, but apparently was never uploaded? ... the experimenter says in Episode 1, that he's limiting flux to only CaF2 and cryolite. He only mentions other fluxes in the first video ... but doesn't try any of them.

Also, his thermite mixes for extracting Ti uses potassium chlorate or perchlorate, which outgasses a lot of potassium chloride gas at molten titanium temperatures. So he looses a lot of the titanium metal as blow-off from the oxidizer.

The process in this thread with plaster, I think, is more efficient.

Perhaps you were thinking of a different person who experimented with fluxes ??

This is what I know of various fluxes:

Cryolite (Na3 Al F6) reduces alumina's melting point to 1000C, and itself melts at 1016C. (Wikipedia) also listed as melting at 950C (Wikipedia). (See Cryolite, vs. Sodium hexafluoroalumiante): It decomposes rather than boil, but the breakdown temp isn't listed even on MSDS sheets. I don't have any of this unpredictable stuff.

Fluorspar (CaF2) melts at 1418C BP 2533C

Boric oxide (B2O3) melts at 450C Sublimes ~1500C and BP 1860C (Wikipeda).
So, boric oxide is similar in melting to Fluorspar; except that aluminum dust can reduce it to metal MP 2076C , BP 3207C.

Calcium Oxide is higher melting: MP 2613C, BP 2850C.

I haven't located Eutectic melting points yet for various oxides, so I'm not sure how to calculate amount of flux being optimum except by trial and error.

I've been thinking that I could perhaps mix aluminum with calcium carbonate or slaked lime, pre-burn it, and then grind up the result to make a flux. Either that, or microwave Ca[OH]2 with graphite to reach 800C in an alumina crucible. The other way I was thinking to attack the problem, was to make a separate thermite mix of flux on TOP of the mixture without flux; then the outgassing would be over when the molten flux ignites the Ti mixture underneath.

I've got a rock polishing tumbler that can be used as a ball mill on solid slag, to re-powder it, pretty easily; but I'm not sure if carbon contamination will ruin the experiment to make Ti Metal. My experience is limited.

Earlier this year, I tried making a ceramic by partially de-oxidizing titanium with plaster of paris and aluminum powder. That was before I knew making Ti Metal was possible that way. My idea was that CaTiO3 has a melting point of 1975C and might precipitate out if I only added enough aluminum dust to remove some of the oxygen and sulpher. The waste alumina is high temperature refactory; So I didn't care that it remained in the ceramic. I just wanted a cheap high temperature refractory sponge. What I got was a bluish ceramic with extremely large air-holes in it.

I used DAP plaster, just as the OP did (ACE Hardware). But I didn't realize it had so much CaCO3 in it. The MSDS says 20-50%, but at the same time -- that can't be right; because the hemihydrate is also listed at 20-50% ...

(It doesn't add UP!)

Having 20% carbonate does seem to explain why DAP plaster becomes so weak after setting. I've cast hydrostone branded gypsum then put it in a ceramics kiln and it will shrink ~20% at 900C -- but it will still he rock hard and strong. But DAP plaster from the hardware store cracks and falls apart about the second time its heated in a regular kitchen oven. Now I know why.

I followed this threads idea, and mixed the DAP plaster with just enough ROOTO drain opener (H2SO4) to where it no longer bubbled. So, I verified that DAP has a lot of chalk in it.

Rooto has some organic impurities in it (a few %), but the impurities are soluble in water or alcohol and tend to follow the steam to the surface when cooked. The picture is of 50g DAP plaster + 591 grams water (2.5 cup) to prevent DAP from setting up hard.

I microwaved the plaster mud for 15 minutes on low to boil most of the water, then kicked the microwave up to high for 40 minutes. The mud shrank to a dry, friable mass (because it was slightly acidic), with lots of bubble holes inside. I put an alumina plate under the pan, because the microwave overheats the tempered glass when firing it on high with little water left. It's also to preserve the microwave, when I heat other things at 700 - 800C with graphite. Alumina doesn't absorb much microwave energy and only gets warm to the touch. The porcelain bread pan ... OTOH ... can get up to 200C or so and crack...

If washed with alcohol, the brown impurites in the picture (organics) mostly dissolve. Scraping them off the top of the gypsum may be sufficient, but alcohol can be easily re-distilled. Alcohol is easier to work with than water, because vacuum filtration is not needed to separate it from they gypsum again.

I think the microwave selectively heats organics more than water ... so likely that's why the they mostly spattered out of the gypsum onto the dish. Convenient.

If the cooked gypsum is put into water again, it turns to mud and won't harden -- So I think it's probably anhydrite. But if there is remaining water of crystallization, I am not sure.

[Edited on 21-6-2018 by semiconductive]

semiconductive - 21-6-2018 at 22:19

Barium Chloride is apparently a good flux for Titanium !

Let us know when you get around to experimenting with lime again. I'm working on a way to make cheap microwave susceptors out of magnetic sand, so I can make relatively pure lime. But I'm not there yet ... meanwhile, I found out that the U.S. military used BaCl2 as an acceptable flux for titanium welds that excludes oxygen.

It has to be dehydrated before use, though.

BaCl2 -- MP 962C BP 1560C

Barium carbonate is fairly cheap; about $4/LB on ebay by seller rakugoldpottery.
$21 delivers 3Lbs to your door. I can make HCl for $2 a gallon 37% ... so that seems like a worthwhile experiment.

I figure posting some thermodynamic data might be useful for future reference...

dH @ 298K, and S at 298K are given in a military report. (http://www.dtic.mil/dtic/tr/fulltext/u2/402468.pdf) The values are not exactly the same as the wikipedia values, even after units conversion; but they're all in one place for easy reference.

TiO2 dH=-225500 cal/mol S*=12.0 cal/mol.deg
CaF2 dH=-290300 cal/mol S*=16.45 cal/mol.deg
CaO dH=0151900 cal/mol S*=9.5 cal/mol.deg
Na2O dH=100400 cal/mol S*=17.0 cal/mol.deg
BaO dH=-133400 cal/mol S*=16.8 cal/mol.deg

Ti Liq dH=3670 cal/mol S*=8.61 ca/mol.deg
CaF2 Liq dH=-277030 cal/mol S*=27.51 cal/mol.deg
BaCl2 Liq dH=-20000 cal/mol S*=34.6 cal/mol.deg

TiF4 gas dH=-438000 cal/mol S*=71.7 cal/mol.deg
Ca gas dH=42600 cal/mol S*=37 cal/mol.dK
Ba gas dH=490000 cal/mol S*=40.67 cal/mol.deg

So, neither CaF2 nor BaCl2 will react with reduced titanium metal. The barium flux is even less likely to react than the fluoride flux. However, one of the discoveries in the report is that fluorides react with titanium oxide above 1000C to dissolve titanium oxide.
The fluorides will partially reduce the titanium oxide, and cause it to "stick" in the slag.

So, the gayest person on youtube was wrong about CaF2 not reacting at all. Fluorides do attack the titanium oxide and absorb it into the slag. This could potentially reduce yields...

Good news/note:
I was able to electro-plate bright clean titanium onto copper wire at about 125C in a proprietary molten electrolyte. (No Water).
So, If I can get titanium metal in crude form with little or no oxygen, I am pretty confident I can make pure titanium metal.
I'll be curious as to how expensive it is (net cost) per pound, because the electrolyte breaks down with time. ( I doubt it's cost effective. )

[Edited on 22-6-2018 by semiconductive]

[Edited on 22-6-2018 by semiconductive]