# Is it possible to melt diamonds into a liquid?

Is it possible to melt diamonds into a liquid? I mean if you heat diamond at open air it will start to burn around 700 degrees Celsius, reacting with oxygen to produce carbon dioxide gas. In the absence of oxygen it will transform into graphite, a more stable form of carbon, long before turning into graphite. My question is, is it possible to melt diamonds into liquid? If so, then, how? If not, then why?

• "...I mean if you heat diamond at open air it will start to burn around 700 degree celsius, reacting with oxygen to produce carbon dioxide gas....." If it burns reacting with oxygen, producing carbon dioxide, then it mean there is no diamond or diamond liquid, it has all got converted into carbon dioxide. I don't know whether what I am saying makes sense or not, but that is what I thought when I read this question. There might exist liquid diamond during heating, but it gets converted to carbon dioxide, in a specific time interval. – Immortal Player Nov 6 '13 at 19:03
• @CURIE To avoid oxidizing the diamond, it would have to be heated in a vacuum or inert atmosphere, yes. Exposed to oxygen, it would burn completely far before it ever got close to melting. But this is not particularly special, for many substances would react with our atmosphere before melting/boiling. – Nicolau Saker Neto Nov 6 '13 at 23:31
• See this related question about liquid carbon… the answers include a phase diagram of carbon, which answers your question – F'x Nov 7 '13 at 7:41

Liquid carbon does indeed exist, but perhaps surprisingly, relatively little is known about it. It only exists above around $4000\ \mathrm{K}$ and $100\ \mathrm{atm}$, which are not trivial conditions to sustain and probe. There certainly are many theoretical studies into the properties of liquid carbon, though. You can find a phase diagram for carbon here, though it is likely just a rough guide. The specific bonding properties of carbon atoms in the liquid phase still seems uncertain to some extent because carbon is very versatile and can create very dissimilar structures (easily seen comparing its two most common phases, graphite and diamond). This source suggests that bonding in the liquid varies continuously between linear acetylenic chains, graphite-like and diamond-like structures, depending on the exact conditions.
You'll need a lot of pressure though! Rigorously speaking, according to the phase diagram above, to obtain liquid carbon directly from melting a diamond would require a pressure of at least $10\ \mathrm{GPa}$, which is about $10^5\ \mathrm{atm}$. To put that into perspective, it's about $1/30$ of the pressure at the centre of the Earth. Amazingly, we actually can reach such absurd pressures in the lab by compressing two diamonds against each other. However, if the conditions required conditions for melting diamond (and sustaining it in molten form) are reached, it would probably compromise the integrity of the instrument and cause catastrophic (possibly explosive) failure, so I don't think anyone will be purposefully melting diamonds any time soon! Other techniques could be used (laser inertial confinement, light-gas guns, etc), but I believe most would only transiently produce liquid diamond.