Sixty Years After Its Discovery in a Meteorite, Scientists Make Hexagonal Diamond in Bulk

When the Canyon Diablo meteorite punched through the Arizona desert some 50,000 years ago, it carried with it a strange passenger. Locked inside fragments of the impactor, crystallographers in the 1960s found what appeared to be a new form of carbon, structurally distinct from ordinary diamond, its atoms stacked in a different geometric pattern. They named it lonsdaleite, after crystallographer Kathleen Lonsdale. Then, almost immediately, the arguments started.

For six decades, lonsdaleite occupied an awkward position in materials science. It existed, or perhaps it didn’t. Some researchers argued the anomalous diffraction signals attributed to it were just ordinary cubic diamond riddled with stacking faults, the crystalline equivalent of a misfolded protein. Others maintained it was genuinely real. Without a pure, sizeable sample to study directly, nobody could settle the matter.

Now a team at Zhengzhou University in China has done something that eluded everyone before them: they synthesised hexagonal diamond in bulk, millimetre-scale pieces of the stuff, phase-pure and stable at room conditions. Writing in Nature, Shoulong Lai, Xigui Yang and colleagues report not just the material itself but its properties, measured directly for the first time. The upshot is that the sixty-year controversy appears resolved, and the winner is the material that theorists have long suspected might be even harder than ordinary diamond.

Harder than diamond. Worth pausing on that.

The team’s route to lonsdaleite exploits a subtlety in how graphite converts to diamond under pressure. Ordinary graphite, compressed hard enough and heated enough, tends to produce cubic diamond by default. The graphitic carbon layers, free to slide over one another, shuffle into the arrangement that thermodynamics prefers. Getting the hexagonal phase instead requires locking those layers in place before the transformation happens, which means the starting material matters enormously. Lai and colleagues used highly oriented pyrolytic graphite, a lab-grown form of graphite whose layers are aligned almost perfectly along a single axis. They then compressed it specifically along that axis, using alumina plugs to concentrate stress in the right direction, at pressures around 20 gigapascals and temperatures between roughly 1,300 and 1,900°C.

The window for pure hexagonal diamond turns out to be quite narrow. A few hundred degrees hotter, and cubic diamond takes over. Colder, and the conversion is incomplete. Only within that tight band does the hexagonal form crystallise cleanly, the HOPG layers bonding to their neighbours whilst retaining their relative positions, the ABAB stacking sequence characteristic of the hexagonal structure preserved through the transition.

Confirming what they had made required an arsenal of techniques. Synchrotron X-ray diffraction at the Shanghai Synchrotron Radiation Facility produced patterns consistent with a P63/mmc hexagonal structure and decisively inconsistent with faulted cubic diamond. Certain diffraction peaks, labelled (101), (102) and (103) in the crystallographic notation, appear only in genuine hexagonal diamond and not in cubic material with stacking defects. They were there. Atomic-resolution electron microscopy revealed the ABAB stacking sequence directly, carbon rings viewed along the hexagonal axis appearing empty at their centres where a column of atoms would sit if the material were cubic. Density measurements came in at 3.51 grams per cubic centimetre, nearly identical to cubic diamond at 3.52. Every test pointed the same way.

The hardness measurements are the figures most likely to raise eyebrows. Vickers hardness testing put the material at 114 gigapascals along one direction, compared with natural diamond’s roughly 110 gigapascals on a comparable surface. Young’s modulus, a measure of stiffness, came in at 1,229 gigapascals versus 1,087 for single-crystal cubic diamond. The differences are not dramatic, but they’re consistent with theoretical predictions made since the late 2000s suggesting the hexagonal form should edge out its cubic cousin in hardness, owing to differences in how the bonded carbon planes resist deformation under load. The team also found that lonsdaleite starts to oxidise at a slightly higher temperature than natural diamond, which is perhaps unexpected for a material researchers have barely been able to study before now.

Understanding why the hexagonal phase forms at all, rather than the thermodynamically preferred cubic form, required molecular dynamics simulations on a scale that would have been impractical a few years ago. The team’s simulations (153,072 atoms, trained on quantum mechanical calculations) showed that the key is the presence of pre-existing bonds between graphite layers at defect sites in the HOPG precursor. These interlayer bonds serve two functions at once: they act as nucleation seeds where the diamond phase can begin to crystallise, and they physically suppress the sliding of graphene layers that would otherwise redirect the transformation towards cubic diamond. Simulations starting with perfect, defect-free graphite produced no hexagonal diamond at all. Simulations starting with disordered graphite produced cubic diamond instead, the wrinkled layers sliding into the familiar arrangement. Only the well-ordered HOPG, with its particular defect structure and the uniaxial stress, yielded lonsdaleite. The mechanism is more specific than anyone had appreciated.

There’s also a forensic dimension here. Lonsdaleite has been proposed as a mineralogical fingerprint of high-velocity cosmic impacts, the reasoning being that the shock conditions of a meteorite strike might uniquely produce the hexagonal form. That idea got murkier in 2014, when a detailed analysis suggested the “lonsdaleite” in ureilite meteorites was actually faulted cubic diamond. The new work doesn’t fully resolve the meteorite question (real impact conditions are messier than a lab press), but it does establish that hexagonal diamond can be made pure and stable, and it provides a definitive structural reference against which meteoritic samples can now be compared.

What it opens up industrially is harder to say at this stage. Bulk hexagonal diamond won’t be spilling out of fabrication facilities any time soon; the synthesis conditions are extreme even by diamond-making standards, and the samples are currently only a millimetre or so across. But the properties measured here, the combination of hardness, stiffness and oxidation resistance, suggest the material could matter for cutting tools and other applications where ordinary diamond is already the benchmark. For now, though, the more immediate significance is simpler: a material that spent sixty years as a rumour has finally been made real, held in your fingers, and measured.

DOI: https://doi.org/10.1038/s41586-026-10212-4