Sciencemadness Discussion Board - A high-temperature mini-furnace for temperatures above 2000 C.

I'm thinking about the possibility of creating a mini-furnace capable of reaching a temperature of just over 2,000 C. The heat source would come from Nernst's lamp. Has anyone looked into this and gotten any results? The furnace would be usable, for example, for melting alumina to create rubies and sapphires, and for similar processes requiring very high temperatures...
https://www.youtube.com/watch?v=aYvWC-mjfdI

At these temperatures the problem isn't providing the heat, rather it's preventing the furnace from melting down.
How do you plan to do that?
Most refractory bricks are made from silca+alumina and some other things, if you want to melt alumina it's going to be tough.

Jenks - 30-11-2021 at 06:35

Now is the question: how could one get (or make) these bars ?

The bottom of the wikipedia article you link mentions that bars of silicon carbide "globars" have replaced Nernst bars for IR spectroscopy. So maybe one of those would do.

metalresearcher - 30-11-2021 at 07:52

Now is the question: how could one get (or make) these bars ?

The bottom of the wikipedia article you link mentions that bars of silicon carbide "globars" have replaced Nernst bars for IR spectroscopy. So maybe one of those would do.

Indeed Globar are SiC bars which can be heated till 1600 C (not that hot as Nernst bar), but does not heat preheating as it conducts at room temperature.

Admagistr - 30-11-2021 at 07:55

The "Nernst stick" has the great advantage of being able to heat up to over 2,000 C in air and not be oxidized!! And is extremely chemically resistant. Tungsten wire doesn't stand a chance in this. The "Nernst stick" can withstand higher temperatures than SiC! It's just a problem of how to make it or where to get it, it's ZrO2 stabilized by Y2O3.
I'd make a furnace lining out of ZrO2 with a little "water glass" mixed in. "Water glass" is approximately 35% Na2SiO3 solution in water. ZrO2 as a furnace lining could be a great choice, it's one of the heat-resistant materials with the highest heat resistance ever. It can hopefully be found in ceramics supplies, in larger quantities.

phlogiston - 30-11-2021 at 13:24

What material is used to make the electrical connections to the actual Y2O3/ZrO2 ceramic?

yobbo II - 30-11-2021 at 15:03

Some ceramic blades (on expensive kitchen knives) are made from this type of ceramic I belived

Admagistr - 30-11-2021 at 15:54

Quote: Originally posted by phlogiston

What material is used to make the electrical connections to the actual Y2O3/ZrO2 ceramic?

True, but which particular one to choose? Occasionally the dealer labels by zirconia ceramics, another kind of similar but totally useless ceramics, with a mixture of Al2O3+ZrO2...Personally, I don't know what's used as a conductor to the Nernst stick, I'd probably choose tantalum? Speaking of SiC, it oxidizes in the air, only superficially, but still. This oxidation protects it from another attack of oxygen by turning part of the silicon into oxide, which prevents carbon from being destroyed. In the case of ZrO2-Y2O3 there is nothing to oxidize

macckone - 1-12-2021 at 03:21

There are only three normal types of heating that will get you to the temperature necessary to melt aluminum oxide.
1) carbon arc furnace (3000C)
2) hydrogen oxygen furnace (3080C)
3) microwave furnace (2200C for commercially available set ups)

All three use graphite crucibles to contain the melt.

Resistance elements top out at about 1850C with MoSi2

From a practical perspective the second two are better as no carbon particles get in the melt.
The microwave is just barely enough to reach the right temperature.
A hydrogen/oxygen type set up is used to make gem quality and/or clear alumina windows.

metalresearcher - 1-12-2021 at 07:09

MoSi2 is the material of which Kanthal Super is made.
It is really horror stuff for amateurs. It is VERY fragile, particularly under 1000 C and requires a strict heating up scheme, so it is only suitable for commercial purposes.

Jenks - 1-12-2021 at 08:38

There are only three normal types of heating that will get you to the temperature necessary to melt aluminum oxide.
1) carbon arc furnace (3000C)
2) hydrogen oxygen furnace (3080C)
3) microwave furnace (2200C for commercially available set ups)

All three use graphite crucibles to contain the melt.

Resistance elements top out at about 1850C with MoSi2

From a practical perspective the second two are better as no carbon particles get in the melt.
The microwave is just barely enough to reach the right temperature.
A hydrogen/oxygen type set up is used to make gem quality and/or clear alumina windows.

Back to light bulbs, the original Edison bulb used carbon (mp 3,550C) as the filament, the novelty being evacuating the bulb. A graphite crucible would also have to be protected from oxygen at high temperature. What I don't understand about the above furnaces is that it would seem simpler to heat the graphite crucible by simply passing an electrical current through it. Why not that?

Twospoons - 1-12-2021 at 11:33

What I don't understand about the above furnaces is that it would seem simpler to heat the graphite crucible by simply passing an electrical current through it. Why not that?

Because you have to get the current into the crucible with lower resistance connections than the crucible itself, otherwise the heating goes into the contacts. And of course the contact material also has to withstand the furnace temperatures.

So while in theory you could use fat graphite lead-ins for the current, they also would be very good thermal conductors, which presents a heat loss problem.

Personally I would go the microwave route - you can completely insulate the crucible for better heating efficiency, and cheap microwave ovens are readily available.

Admagistr - 1-12-2021 at 17:47

Quote: Originally posted by Twospoons

What I don't understand about the above furnaces is that it would seem simpler to heat the graphite crucible by simply passing an electrical current through it. Why not that?

Thanks you for the advice! You mean a regular microwave for kitchen use? When I tried to melt a powdered Al2O3 in it with a little amount of Cr2O3, by placing a closed ring of 1 mm thick Al wire on this mixture surface, I had to terminate the experiment prematurely after about 20 seconds, because the microwave started to stink and something was probably burning. How do you make not burn the magnetron or any other part of the microwave? Would help the perfect thermal insulation of the crucible? Otherwise, the experiment went well, I created tiny rubies and one formed on the end of a overburnt aluminum wire. Thank you!

Twospoons - 1-12-2021 at 19:54

Ideally you want to modify the oven by inserting a circulator or isolator between the magnetron and waveguide. This can be used to dump reflected power into a dummy load, protecting the magnetron. But this is into the realm of RF engineering - not trivial.

It would probably help to have a bigger susceptor in the oven for heating, probably SiC or graphite rather than Al metal. Al is too conductive to use as a heater - and it melts too low.

The other thing worth doing is to use a variac to control the magnetron power, but to do this you will need a separate transformer for the filament supply - which you can pull from a second MW oven ( same model preferably).

So a bit of engineering there, but probably easier than trying to make a zirconia element.

Jenks - 2-12-2021 at 06:33

Quote: Originally posted by Twospoons

What I don't understand about the above furnaces is that it would seem simpler to heat the graphite crucible by simply passing an electrical current through it. Why not that?

The heat loss isn't ideal, but otherwise your description sounds very simple - take a carbon rod and narrow it in the middle, drill a hole in the top of the narrowed part to serve as a crucible, surround the thing with a steel box or can using a ceramic to insulate the carbon from the can, purge with nitrogen and pass enough current through to heat the narrowed portion to the desired temperature. Maybe spray the can with water if it gets too hot. I like that this doesn't require costly or complicated technology. Is there some reason this wouldn't work?

Twospoons - 2-12-2021 at 11:18

Of course that would work - but that wasn't your original question was it? Making a graphite heating element with a narrow mid section is not the same as passing current through a thick graphite crucible.

Looking around the net it would seem induction heating of graphite crucibles is a common method. Safer than microwave engineering, but you'd need an induction heater supply.

Fulmen - 2-12-2021 at 12:43

@Admagistr: What is your current level of proficiency? You can't just jump in at 2000°C, that's wizard level alchemy...

yobbo II - 2-12-2021 at 12:55

Thread here of use.

http://www.sciencemadness.org/talk/viewthread.php?tid=70334

Super Kanthal can be run from a large welder transformer and a variac. I have done it. The Super K is hard to purchase cheap and you need a large variac (and welder).

Yob

wg48temp9 - 2-12-2021 at 13:02

Induction heating of a graphite crucible with water cooled copper tube coil would be the way to go. However at 2,000C and higher the graphite will be burning in air and the heat loss will be very high.

The burning can be easily solved with an inert atmosphere or a vacuum. A vacuum would have the advantage of eliminating convective cooling but that still leaves radiative cooling which at 2,000C and above will be very high even in a vacuum. That heat loss can be reduced with a one or more reflective shields and if sufficiently far from the crucible can be made from stainless steel or even copper. The shields may have to be water cooled unless they are made of a refractory metal.

With a small 20 ml crucible in a vacuum with reflectors as above, perhaps only one or two kW of heating power would be required.

A 2.5kW (probably more accurately half of that) high frequency induction heating supply is relativly cheap from China £90.5 on ebay see https://www.ebay.co.uk/itm/402913450608?hash=item5dcf835a70:...

Twospoons - 2-12-2021 at 14:10

I'd put the graphite crucible inside a zirconia crucible for a first layer of insulation. Vacuum would be more of a PITA than inert gas, probably not worth the effort since radiative heat transfer will far outweigh convective heat transfer (from memory radiative goes up as the fourth power of delta T).
I hadn't realised those chinese induction heaters were so cheap - might have to put one on my wish list to play with.

Admagistr - 2-12-2021 at 17:05

Quote: Originally posted by Fulmen

@Admagistr: What is your current level of proficiency? You can't just jump in at 2000°C, that's wizard level alchemy...

@Fulmen I'm sorry, I don't know if I understand your question correctly? English is not my mother's language. What exactly do you find unrealistic? I'm not an electrician or a radio amateur,..I have some theoretical knowledge, but theory is not the same as practice...

wg48temp9 - 3-12-2021 at 06:43

Quote: Originally posted by Twospoons

I'd put the graphite crucible inside a zirconia crucible for a first layer of insulation. Vacuum would be more of a PITA than inert gas, probably not worth the effort since radiative heat transfer will far outweigh convective heat transfer (from memory radiative goes up as the fourth power of delta T).
I hadn't realised those chinese induction heaters were so cheap - might have to put one on my wish list to play with.

I wounder how insulating zirconia crucibles with a lid would be. Perhaps with an outer high alumina insulation it could be heated in a domestic microwave oven and reach 2,000C. With a lid the oxidation of the graphite would be significantly reduced.

You need a multi-kilowatt 48V power supply to power the induction heater. I noticed in the "12V PSU for electrolysis" thread referred to a cheap server 70A@12V power supply. I checked on ebay and yes they are cheap for used and even tested ones they cost about £10 to £15. Three or four of them would easily drive the induction heater. From memory they are also suitable for powering carbon gouging rods (copper removed) used as heating elements.

PS: There are youtube videos showing how these server power supplies can be modified to change the output voltage or even make it variable.

They also show what the output connections are and how to enable them.

The following link has the connections and mod info for cheap and commonly available server power supplies: https://www.xsimulator.net/community/threads/the-ultimate-po...

[Edited on 12/4/2021 by wg48temp9]

metalresearcher - 3-12-2021 at 09:18

These cheap induction furnaces from ebay are very picky on overpowering. They are not protected to overpowering which is dependant on the contents of the crucible. The MOSFETS decease easily when overpowered.

wg48temp9 - 3-12-2021 at 21:32

These cheap induction furnaces from ebay are very picky on overpowering. They are not protected to overpowering which is dependant on the contents of the crucible. The MOSFETS decease easily when overpowered.

Can you give more detail as to what you mean by "overpowered MOSFETS"?

My understanding is as follows:
1) The minimum component count circuits don't have protection against both MOSFETS being on at the same time which can happen during switch "on" the resulting high dissipation in the MOSFETS destroys them.
2) There is usually little to no limits on what the generators can drive other than the maximum power supply voltage and current drawn. The more expensive generators do include a coil which I would assume the generator is capable of driving at least without anything loading the coil.

I am also aware that maximum voltage, current and in particular power handling specs of cheap modules from China tend to be optimistic.

Fulmen - 4-12-2021 at 00:53

@Admagistr: My advice; start with a 1000-1200°C furnace. The challenge grows exponentially with temperature, working at 2000°C with no prior experience isn't realistic.

macckone - 4-12-2021 at 09:01

+1 Fulmen

a kitchen oven is 250C.
a propane forge is about 1000C and an electric furnace capable of 1000C is not as easy as it seems.
Getting to even 1300C requires some engineering challenges (upgrading to fuel oil).
Even though the flame temperature of propane is higher than 1300C getting a working forge or furnace that heats that hot is not as intuitive as it may seem.
Most kilns do not go to cone 10 which is 1305C.
Kanthal A1 is usable up to 1400C with the right wire guage.
But you have to have thick walls (6" of insulation) and the right wattage.

JohnnyBuckminster - 5-12-2021 at 05:04

I have been looking for references about option 2, using a graphite crucible in a hydrogen-oxygen furnace, but there doesn't seem to be that common.

Does anybody have any info on this, i.e. how do you use a graphite crucible in a hydrogen-oxygen furnace?

wg48temp9 - 5-12-2021 at 08:03

Red hot carbon reacts with water vapor to produce CO and H2. Graphite probably does the same. So a red hot graphite crucible, particularly one at 3,000C, may not last very long when heated with a hydrogen oxygen flame.

Admagistr - 5-12-2021 at 17:06

Quote: Originally posted by Fulmen

@Admagistr: My advice; start with a 1000-1200°C furnace. The challenge grows exponentially with temperature, working at 2000°C with no prior experience isn't realistic.

I worked with a laboratory crucible furnace with a temperature up to 1000 C. Some time ago, I purchased a very precisely controlled modern laboratory furnace, with PID regulation, from a European manufacturer, up to a temperature of 1250 C. So some little experience I have...

Admagistr - 5-12-2021 at 17:16

Also, I haven't heard anything about oxy-hydrogen furnaces with graphite crucible in which Al2O3 melts...Carbon reacts with water vapor at high temperatures and creates dangerous gaseous products...In our country, it's called "Water Gas."

macckone - 6-12-2021 at 09:57

You have to use a reverb type set up.
The aluminum oxide is heated from the top, not the bottom.
Hydrogen oxygen flame will ignite graphite.
The graphite crucible is often heated from the bottom with traditional electrical heating to 1400C.
Since graphite is a pretty good insulator, the aluminum oxide in the crucible stays fluid with the flame heat input and the flame doesn't impinge on the graphite.
The aluminum oxide is preheated in a chute as it is dropped into the crucible.

This how they do it for large panes of aluminum oxide.

Fulmen - 6-12-2021 at 11:12

@Admagistr: That's good to know, it's hard to give advice without knowing your experience.

Admagistr - 6-12-2021 at 15:51

You have to use a reverb type set up.
The aluminum oxide is heated from the top, not the bottom.
Hydrogen oxygen flame will ignite graphite.
The graphite crucible is often heated from the bottom with traditional electrical heating to 1400C.
Since graphite is a pretty good insulator, the aluminum oxide in the crucible stays fluid with the flame heat input and the flame doesn't impinge on the graphite.
The aluminum oxide is preheated in a chute as it is dropped into the crucible.

This how they do it for large panes of aluminum oxide.

That might work well, thanks for the tip! I've thought about it, too, and I haven't seen a major problem. There's a company in our country that makes sapphire tubes and crucibles by modified Stepanov method. I spoke to their production experts and they told me they use large vessels of molybdenum and tungsten, working in a vacuum, or in an argon atmosphere. Those sapphires can then be surface contaminated with molybdenum. I suggested that they reheat them in the air at a high temperature to oxidize traces of Mo to MoO3, which they did, and when the product came to me it looked completely clean. I had to pay extra for that. Otherwise, their finished sapphire products have a chemical purity of 99.999%!
I'm also thinking and dreaming about the possibility of melting Al2O3 with an infrared CO2 laser. I don't know if any offered by Chinese retailers would be appropriate, I'd be more inclined to make my own CO2 laser, there's a few detailed instructions on how to do it on the internet, but it's not quite simple... The idea is very attractive for me, the laser heating would maintain perfect chemical purity...I even found a video on YouTube where Al2O3 melts in the focus of a giant solar furnace!

Twospoons - 6-12-2021 at 23:49

A laser is a rather inefficient way to put heat into bulk material. CO2 lasers are about 10% wall-plug to light output.

JohnnyBuckminster - 7-12-2021 at 00:22

There are some reports about holding Al3+ in graphite crucibles, it is problematic because of the formation of alumina carbides, see for example Hoseinpur & Safarian,

"Results showed that Al in Si–Al melt infiltrates into graphite leading to the formation of aluminum carbides, which accompanies with volume expansion and therefore the crucible destruction."

That might be why it is necessary to use molybdenum / tungsten crucibles.

Admagistr - 7-12-2021 at 16:08

On the other hand, Henry Moissan, winner of the Nobel Prize in Chemistry, successfully used his electric arc furnace and graphite crucible to synthesize rubies! I wrote a new topic here recently in a forum. So you can see that even graphite can be used, albeit in a limited way...

JohnnyBuckminster - 7-12-2021 at 23:50

Moissan, didn't he just throw everything into water, so the crucible was "consumed" in the experiment?

Forth and back, I have been thinking about a small scale furnace, ~50 ml volume, for high-temperature experiments, > 2 000 C. I ruled out the electric arc furnace because it seems difficult to control. A hydrogen-oxygen furnace is appealing, highly controllable, can reach temperatures up to 3 000 C, and you don't need to worry about blowing the fuses. But, unwanted chemistry can be induced, because of the atmosphere.

A Nernst lamp is a very interesting option as a heat source, it should be highly controllable, and can operate in an ordinary atmosphere.

The company in your country, that makes sapphire tubes in molybdenum and tungsten crucibles, how do they heat the crucible?

metalresearcher - 8-12-2021 at 03:32

Maybe this Chinese supplier of induction furnace might give you an idea ?

https://www.cdocast.com/category/precious-metal-melting-furn...

Or (cheaper and easier) the Youtube channel of Nighthawkinglight, he has some nice ideas of obtaining high temperatures.

Admagistr - 8-12-2021 at 07:37

Moissan doesn't say if the crucible went down or not, it just says what voltage and current he used and how long the meltdown lasted, it was only 10 to 15 minutes! Our manufacturer uses induction heating, it's standard procedure for the Czochralski method and its like. The manufacturer is CRYTUR and is based in Turnov, Czech Republic. It has a great tradition and experience, now it is focused on the American market. I thought of powering up that Moissan furnace with four 12V batteries, designed for automobiles, and using an induction, a silencer, some coil to guard the batteries for overload and to keep them from exceeding the prescribed current. An electrician told me that he had once built a similar battery power suply and it worked as a welder for him. In Prague, at Charles University, they use an optical furnace to synthesize rubies and sapphires and other gems, when the necessary temperature is generated in a pipe of quartz glass, in the focus of ellipsoidal mirrors! They use for it powerful halogen lamps. I spoke to their expert and he invited me to tour their labs. I haven't been there yet, because of Covid. Thanks so much for the link and the tips, Greetings German colleagues and friends!

[Edited on 8-12-2021 by Admagistr]

JohnnyBuckminster - 8-12-2021 at 20:22

Did you get any information about why they work in vacuum or in an argon atmosphere, is it to protect the crucible or the melt?

macckone - 9-12-2021 at 11:03

They use an inert environment to protect the crucible.
The melt is a very stable oxide.
Graphite crucibles used in industry have a very limited life span but are considerably cheaper than molybdenum or tungsten in the same size.
Once the large chunk is made, they cut off the edges that are 'imperfect' then slice it to make windows.
Diamond saw blades etc, the waste dust is used as a polishing abrasive.

Ubya - 9-12-2021 at 18:41

https://www.youtube.com/watch?v=VTzKIs19eZE

i have no idea the temperature that could be reached with this arc furnace, but it is small, uses "common" materials and it is pretty simple to operate (even though i wouldn't call it 100% safe foor newbies).

the electrodes get burned pretty quickly, and the insulating brick he used in the video got pretty beat up by the heat, so i suppose finding a refractory/insulating brick that survives 2.000°C is a challange in itself

i know this post was about a small furnace that uses nerst elements, i have 0 experience with nerst lamps (i only know of them because i had an IR spectroscopy course)

Admagistr - 9-12-2021 at 19:38

Quote: Originally posted by Ubya

https://www.youtube.com/watch?v=VTzKIs19eZE

i have no idea the temperature that could be reached with this arc furnace, but it is small, uses "common" materials and it is pretty simple to operate (even though i wouldn't call it 100% safe foor newbies).

the electrodes get burned pretty quickly, and the insulating brick he used in the video got pretty beat up by the heat, so i suppose finding a refractory/insulating brick that survives 2.000°C is a challange in itself

i know this post was about a small furnace that uses nerst elements, i have 0 experience with nerst lamps (i only know of them because i had an IR spectroscopy course)

Great video, thanks! I'll be making a brick of zirconia (ZrO2) reinforced with sodium silicate solution! Zirconia has a melting temperature of 2700C!

yobbo II - 10-12-2021 at 16:43

Once you 'reinforce' the ZrO2 you will decrease its melting temp.
PURE ZrO2 has a melting temp. of 2700C.

Pure bricks are very expensive. I think it's because they are pure or perhaps they are a low volume produced product or both?

Admagistr - 10-12-2021 at 17:38

Quote: Originally posted by yobbo II

Once you 'reinforce' the ZrO2 you will decrease its melting temp.
PURE ZrO2 has a melting temp. of 2700C.

Pure bricks are very expensive. I think it's because they are pure or perhaps they are a low volume produced product or both?

That's a good point, and unfortunately true...Thanks for that. In which case, it would be better to use sodium zirconiate, or a zirconium dioxide water gel... Some of the "zirconia" bricks are actually ZrSiO4, not ZrO2. Also, ZrO2 is mixed with Al2O3 for those purposes. Why these bricks are so expensive I don't know, but maybe because the sintering process that they're probably making takes a long time and high temperatures. Plus ZrO2 is much more expensive than, say, Al2O3...

yobbo II - 10-12-2021 at 18:42

There is a link here showing Zirconia blocks. They range from 99% to 99.999%

https://www.americanelements.com/zirconium-blocks-7440-67-7

Funny thing is they give the same melting temperautres?

Bricks here at about 12 bucks each!
https://www.ebay.co.uk/itm/264205633418?_trkparms=amclksrc%3...

Tubes of ZiO2+Yit Oxide here but they are probably a silly price.

https://www.preciseceramic.com/products/zirconia-ysz-tube/

Yob

[Edited on 11-12-2021 by yobbo II]

[Edited on 11-12-2021 by yobbo II]

metalresearcher - 10-12-2021 at 23:36

You can also use magnesia (mp 2800 C). I tried to find a supplier of MgO bricks, but could not find one.

Admagistr - 11-12-2021 at 15:50

You can also use magnesia (mp 2800 C). I tried to find a supplier of MgO bricks, but could not find one.

Thanks a lot! That would be a cheaper solution, I believe

. Moissan used calcium oxide in his furnace as thermal insulation, which surprises me, because of its, to water rather reactive and then corrosive nature...Before it turns into CaCO3.

macckone - 12-12-2021 at 14:02

Magnesia is easily synthesized.
Start with sorel cement, heat until all the HCl is given off.
To make it impermeable you can treat it with milk of magnesia (use the low sodium variety).
Then treat with linseed oil, which will crosslink and carbonize under heat.
At the working temperature you are hoping to achieve, the carbon with graphitize.
This yields a magnesia/graphite composite that is very durable and capable of withstanding all but the highest temperatures (> 2500C) and is more resistant to oxidizing conditions.

metalresearcher - 13-12-2021 at 01:04

Magnesia is easily synthesized.
Start with sorel cement, heat until all the HCl is given off.To make it impermeable you can treat it with milk of magnesia (use the low sodium variety).
Then treat with linseed oil, which will crosslink and carbonize under heat.
At the working temperature you are hoping to achieve, the carbon with graphitize.
This yields a magnesia/graphite composite that is very durable and capable of withstanding all but the highest temperatures (> 2500C) and is more resistant to oxidizing conditions.

I thought cement is basic so it never can contain HCl ?

Admagistr - 13-12-2021 at 14:35