At the Renaissance fair a few years back I was watching a smith forge metal into shapes. During this time a very odd question came to me. I was wondering what the furnace was made of. My logic stated that whatever the furnace was made of must have a higher melting point than the materials he was melting. This quickly turned into an elemental arms race resulting in an odd question of how do we melt stuff like refractory metals (more specifically the one with the highest melting point) so we can melt other things inside of it.

Now I know that (for some odd reason I don't understand) rapid cooling can manipulate the strength of an item. Is there a similar property to manipulate the melting point?

Note: My current best guess (like can be done to make weapons harder) is that we take two elements, melt them, and the resulting compound has a higher melting point.

  • 9
    $\begingroup$ The level of positive reaction to this not only makes me not feel stupid for my question but gives me a fuzzy feeling inside. Thank you. $\endgroup$
    – Griffin
    Sep 24, 2014 at 13:04
  • 3
    $\begingroup$ The vessel holding the metal doesn't have to be as hot as the metal itself if you have another way of getting energy into the metal. Some refractory metals can be vaporised from cooled containers by directing electron beams or other energy sources into the metal (not for large scale melting but often used in synthesis or vapour deposition processes). $\endgroup$
    – matt_black
    Apr 28, 2016 at 1:01

5 Answers 5


Tungsten's melting point of 3422 °C is the highest of all metals and second only to carbon's, for which melting occurs only at high pressure (there's no standard melting point). This is why tungsten is used in rocket nozzles and reactor linings. There are refractory ceramics and alloys that have higher melting points, notably $\ce{Ta4HfC5}$ with a melting point of 4215 °C, hafnium carbide at 3900 °C and tantalum carbide at 3800 °C.

Carbon cannot be used to hold molten tungsten because they will react to form tungsten carbide. Sometimes ladles and crucibles used to prepare or transport high melting point materials like tungsten are lined with the various higher melting ceramics or alloys. More typically tungsten and other refractory materials are fabricated in a non-molten state. A process known as powder metallurgy is used. This process uses 4 basic steps:

  • powder manufacture - a variety of techniques are available to generate small particles of the material being worked
  • powder blending - routine procedures are used to blend the constituent particles into a uniform mixture
  • compacting - the blended powder is placed in a mold and subjected to high pressure
  • sintering - the compacted material is subjected to high temperature and some level of bonding occurs between particles.
  • 1
    $\begingroup$ Materials like tungsten may also be shaped through electrochemical machining (ECM) or electrical discharge machining (EDM). $\endgroup$ Sep 24, 2014 at 17:29
  • 1
    $\begingroup$ Ta4HfC5 isn't technically an alloy. It is a ceramic (a mixture of tantalum and hafnium carbides). Tungsten has the highest melting point of any metal or alloy. Alloys typically have lower melting points than the metals used to form them. $\endgroup$ Nov 30, 2015 at 23:58
  • $\begingroup$ "transport high melting point materials like tungsten". I would LOVE to see a container pouring out pure liquid tungsten the way we do with iron or bronze casting. $\endgroup$ Jul 31, 2021 at 23:13

Sorry, can't comment here, but I wanted to more directly answer your question.

Blacksmiths avoid melting their forges because the "heat" that can melt or oxidize iron and steel is actually contained in a ball in the center of the coal. In fact, maintaining coal "structure" is an important skill in blacksmithing.

To clarify better, imagine a hollow in the center of a pile of coal. This is where the temperatures rise past 2000 °F, since the heat reflects back in on itself due to the coal molding into a refractory ball of sorts.

And yes, sometimes your ball does fall apart, or you've structured it poorly - and then you notice that the cast iron drain cover that protects your airflow intake has melted through.

  • 1
    $\begingroup$ There's an interesting analogue to this in modern fusion reactors. For example, in JET, a thin ring of plasma hotter than the core of our sun (and indeed anything known in our galaxy) is held in place by magnetic levitation. While the ring itself is hot enough to destroy any material it touches, the thermal radiation falloff (iirc due to Planck's law and the Stefan–Boltzmann law) allows them to use supercooled graphite panels to protect the reactor, staying well within the 3550°C melting point of carbon. $\endgroup$
    – Polynomial
    Nov 12, 2016 at 11:52

We use an levitating furnace to heat samples of refractory ceramics up to approx $3000~^\circ\mathrm{C}$. Its for research purposes, so the samples are small (2 mm) beads. These are balanced on a jet of argon and heated with $\ce{CO2}$ lasers.

Here is a paper which talks about the technique:
D. Langstaff, M. Gunn, G. N. Greaves, A. Marsing, and F. Kargl, Rev. Sci. Instrum.; 2013, 84, 124901. (Mirror)


One could melt them floating on a pool of high-boiling point denser metal, or in space where they can readily be contained. Or one could create a thick actively cooled shell and melt them inside it, melting part of the shell as well. Finally, it's probably not very practical, but one could use an air jet to keep then suspended away from other matter and then melt them with lasers or superheated air.


There are two alternatives to the other answers here, though whether they can be used on a large scale is open to question.

The first is to use an actively cooled vessel to hold the metal and a method of getting energy into the metal not based on the heat of the crucible. Many metal-vapour reactions (used for small scale chemistry research) do this and provide sufficient energy to vaporise even refractory metals using electron guns. See Malcolm Green's site (and this entry "The synthesis of the first zerovalent compounds of the early, refractory transition metal via the development of the electron-gun metal vapour synthesis experiment").

The other method is to use inductive heating of the metal. This can sometimes work even without a vessel at all as a suitable inductive coil will levitate the lump of metal and the induced eddy currents will dump enough energy into it to melt it. There are plenty of youtube videos of this with non-refractory metals such as aluminium but the principle should still work for high melting metals.


Not the answer you're looking for? Browse other questions tagged or ask your own question.