What Is the Melting Point of Diamond?

Diamond melts at roughly 4,500 K (about 4,200 °C or 7,600 °F), but only if you squeeze it under enormous pressure first. At everyday atmospheric pressure, diamond never reaches a liquid state; it transforms into graphite or simply burns away before melting could begin. That makes diamond’s melting point an unusually slippery number, one that shifts dramatically with pressure and has pushed physicists to develop some of the most extreme laboratory conditions ever created just to pin it down.

Why Diamond Does Not Melt at Ordinary Pressure

If you heated a diamond in open air, you would not see it melt into a puddle of liquid carbon. Instead, two things happen depending on how much oxygen is around. In an oxygen-rich environment, diamond starts to oxidize and effectively burns. In a vacuum or inert atmosphere, diamond converts to graphite, the more stable form of carbon at low pressure. Experiments on diamond powders in vacuum have documented this graphitization process kicking in across the temperature range of about 1,700 to 2,100 K, with two distinct regions governed by different activation energies.¹ This is well below any supposed melting point, which means the diamond crystal structure falls apart and rearranges into graphite layers long before the carbon atoms could shake loose into a true liquid.

Even tiny nanodiamonds follow this pattern. Experiments on optically levitated nanodiamonds in vacuum show that as the surrounding pressure drops and the particle heats up from trapped laser light, the diamond shrinks. First it loses surface water and organic material, and then its outer shell of amorphous carbon burns away or graphitizes.² At no point does the nanodiamond melt. The carbon simply transforms into another solid form or combines with available oxygen and vanishes as gas.

So in any setting you could create in a kitchen, a workshop, or even a typical research furnace, diamond will never melt. It will graphitize, crack, or burn. To actually melt it, you need pressure high enough to prevent the crystal from converting to graphite first.

The Pressure Requirement

Carbon’s phase diagram has a critical landmark called the graphite-diamond-liquid triple point, the unique combination of pressure and temperature where graphite, diamond, and liquid carbon can all coexist. This point sits at roughly 12 gigapascals (GPa) and around 4,500 K. Below that pressure, heating diamond leads to graphitization, not melting. Above it, you can push diamond into a genuine liquid state.

Computational simulations using molecular dynamics have mapped out this triple point and the diamond melting curve extending from it.³ Separate calculations using quantum-mechanical methods have confirmed that these computed melting curves match well with the available experimental data from shock wave studies.⁴ The agreement between theory and experiment gives physicists reasonable confidence in the overall shape of the melting curve, even though directly measuring it remains fiendishly difficult.

To put 12 GPa in perspective, that is about 120,000 times atmospheric pressure, roughly the pressure found deep inside Earth’s mantle. This is why no one accidentally melts a diamond. The conditions required are planetary in scale.

How Scientists Actually Measure Diamond Melting

You cannot put diamond in a crucible, crank up a furnace, and watch it melt. Instead, researchers rely on shock compression, using powerful lasers to slam a wave of pressure through a diamond sample so quickly that it heats and compresses simultaneously. The diamond has no time to graphitize because the pressure ramps up faster than the crystal structure can rearrange.

In one landmark set of experiments at France’s LULI laser facility, researchers fired 527-nanometer laser pulses carrying up to a kilojoule of energy onto layered targets that included a plate of type IIA diamond. The laser drives a shock wave through the target, and the team measured the shock velocity using an interferometric diagnostic that tracks tiny Doppler shifts in reflected light. By measuring how fast the shock moved through the diamond and comparing it to the known behavior of a quartz reference plate, the researchers could reconstruct the pressure and temperature conditions inside the sample.⁵ These experiments lasted just a few nanoseconds, capturing a snapshot of diamond under conditions that exist naturally only inside giant planets.

A separate series of shock experiments pushed diamond to even more extreme conditions, between 0.6 and 1.05 terapascals (600 to 1,050 GPa). The researchers observed a temperature plateau at about 9,000 K, which they interpreted as a signature of melting: the latent heat absorbed by the phase transition holds the temperature steady even as more energy pours in.⁶ That 9,000 K figure applies at pressures roughly 100 times higher than the graphite-diamond-liquid triple point, illustrating how steeply diamond’s melting temperature climbs with pressure.

The Unusual Shape of Diamond’s Melting Curve

Most materials have a straightforward melting curve: as pressure goes up, the melting temperature goes up too. Diamond’s melting curve bends the other way over part of its range. At pressures above about 500 GPa, the melting temperature actually decreases as pressure increases. This negative slope is a relatively rare phenomenon among solids, and it tells physicists something fundamental about how carbon atoms rearrange when diamond transitions to liquid.

The laser-shock experiments at LULI provided some of the first direct evidence for this negative slope.⁵ In practical terms, a negative-slope melting curve means that at very high pressures, applying more pressure can actually make diamond easier to melt rather than harder. Water ice has a similar quirk at low pressures, which is why ice skating works the way it does (though the analogy is imperfect). For diamond, the phenomenon kicks in at pressures so extreme that it has no everyday consequence, but it reshapes how physicists think about carbon’s behavior deep inside planets like Neptune and Uranus, where pressures of hundreds of gigapascals are routine.

What Happens Beyond a Terapascal

At the most extreme pressures studied so far, diamond’s story takes another unexpected turn. Theory has long predicted that at pressures above roughly 1,000 GPa, diamond should transform into a different crystal structure called BC8, a denser arrangement of carbon atoms. For nearly 40 years this remained a purely theoretical prediction, with no direct experimental confirmation.

Recent shock-compression experiments have finally produced evidence consistent with this transition. Measurements of the Hugoniot curve, the relationship between shock speed, pressure, and temperature along the compression path, closely follow the negative slope of the diamond melting curve between 700 and 830 GPa, then shift in a way that is compatible with a diamond-BC8-liquid triple point at roughly 830 GPa and 7,200 K.⁷ Above that pressure, the data are consistent with a BC8 melting curve extending from 830 to 1,000 GPa. If confirmed by further experiments with direct structural measurements, this would mean that at terapascal pressures, carbon prefers to exist as BC8 rather than diamond, and the “melting point” at those conditions is the melting point of BC8, not of diamond itself.

These pressures are staggering. One terapascal is ten million atmospheres. The only natural settings where such conditions exist are deep inside gas giants and certain exoplanets. Understanding how carbon behaves there matters for modeling planetary interiors and predicting what structures might form in the layers of ice giants.

What Liquid Carbon Is Actually Like

Once diamond does melt, the resulting liquid is exotic. It is not like molten metal, and it is not like any liquid you have encountered. Measurements of liquid carbon’s electrical resistivity at the melting point show values ranging from about 470 to 630 microhm-centimeters depending on pressure, placing it in the category of a “poorly conductive metal,” worse at conducting electricity than typical liquid metals but far better than an insulator.⁸

Even more interesting is how liquid carbon’s character shifts with pressure. At lower pressures (around 1.4 to 4 GPa), its resistivity drops as pressure increases, behaving like a metal in the sense that resistivity rises with temperature. But above roughly 4 to 5.6 GPa, the trend reverses: resistivity starts climbing with pressure, and the temperature dependence flips to a pattern more like a semiconductor. This suggests that the liquid undergoes a structural transformation, with carbon atoms bonding differently at different pressures.⁸ Liquid carbon, in other words, is not a single kind of liquid. Its internal structure changes as you squeeze it harder.

Femtosecond spectroscopy experiments, using ultra-short laser pulses to melt diamond and graphite surfaces and then immediately probe the resulting material with light, have shown that both graphite and diamond melt into the same liquid state. The liquid is more reflective than either crystalline precursor, and the melting transition completes within roughly a picosecond after the laser pulse delivers enough energy to overcome the critical threshold.⁹ The fact that graphite and diamond yield the same liquid reinforces that “liquid carbon” is a single phase, regardless of which solid form you started from.

How Shear Forces Lower the Melting Point

The melting points discussed so far assume that the diamond is under uniform pressure, squeezed equally from all sides. Real-world conditions are rarely so tidy. When diamond is subjected to shear, where forces push one part of the crystal in a different direction than another, the melting point drops substantially.

Molecular dynamics simulations have revealed that under certain shear strains, diamond’s melting temperature can fall by as much as 1,150 K compared to the equilibrium value at the same pressure.¹⁰ The same study found that diamond’s mechanical strength drops by about half at 3,000 K, even though its elastic stiffness, how much it resists small deformations, stays nearly constant. In effect, the diamond feels just as rigid until it suddenly gives way. This “brittle until it isn’t” behavior has implications for any application where diamond tools or anvils operate at high temperatures: the crystal can fail catastrophically at temperatures well below its nominal melting point if shear stresses are present.

This finding also matters for the diamond anvil cells used in high-pressure physics labs worldwide. These devices use two gem-quality diamond tips to squeeze tiny samples to enormous pressures. The combination of high temperature and shear forces at the anvil tips can cause the diamonds to fail at temperatures lower than would be predicted by the equilibrium phase diagram alone.

When Diamonds Are Very Small

Nanodiamonds, diamond crystals just a few nanometers across, do not simply scale down the behavior of bulk diamonds. Their enormous surface-to-volume ratio means surface effects dominate, and the temperatures at which they graphitize can differ sharply from what bulk diamond does.

Molecular dynamics simulations have predicted that nanodiamonds with a core as small as 0.8 nanometers can remain kinetically stable, meaning the diamond structure persists because there is not enough thermal energy to trigger the rearrangement, up to about 1,500 K.¹¹ That is significantly smaller than the roughly 1.9 nanometer lower limit where nanodiamond is thermodynamically the most stable form of carbon. Below that size threshold, the diamond structure is technically less stable than graphite, but the transformation is slow enough that the nanodiamond survives at moderate temperatures.

Experimental studies on ultra-small nanodiamonds paint a more complicated picture, though. Work on nanodiamonds of around 1.4 nanometers found that graphitization occurs throughout the entire particle rather than starting at the surface, as happens for slightly larger nanodiamonds of about 3.4 nanometers.¹² The theoretical expectation that very small nanodiamonds might actually be more thermally stable was not confirmed. Instead, the smallest particles graphitize in bulk, suggesting that below a certain size, the interior of the nanodiamond is just as vulnerable to transformation as the surface. For anyone working with nanodiamond powders in industrial or biomedical applications, particle size is a critical variable in determining how much heat the material can tolerate.

Why the Answer Keeps Changing

Diamond’s melting point is not a settled, textbook-ready number in the way that the melting points of, say, copper or table salt are. Every decade brings new experiments that push the accessible pressure range higher, and each increment reveals more complexity in carbon’s phase diagram. The discovery of the negative-slope melting curve was a surprise. The emerging evidence for the BC8 phase at terapascal pressures was another. The structural transition within liquid carbon itself was yet another.

Part of the difficulty is that these measurements happen in nanoseconds. A shock-compression experiment might compress a diamond to a million atmospheres, hold it there for a billionth of a second, and then release. Extracting clean temperature, pressure, and phase data from that fleeting moment requires extraordinary precision in diagnostics, and different experimental groups sometimes disagree on interpretation. The temperature plateau at 9,000 K observed in one set of experiments, for instance, relies on interpreting thermal radiation from the shock front, a measurement that involves assumptions about how opaque the shocked material is and how the light travels to the detector.⁶

The practical upshot is that if someone asks “what is the melting point of diamond?” the honest answer depends on how much pressure you are willing to apply. At the minimum pressure where melting is possible, around 12 GPa, the answer is close to 4,500 K. At hundreds of gigapascals, the melting temperature rises to 7,000 K or higher, and the melting curve’s slope becomes negative. At terapascal pressures, diamond may transform into BC8 before it even gets the chance to melt, shifting the question entirely. Diamond does not have one melting point. It has a melting curve, and the curve itself is still being drawn.