A hexagonal diamond, in bulk
In the 4 March issue of Nature, a Chinese team reports bulk synthesis of hexagonal diamond, the rarer and theoretically harder form of diamond that until now has appeared only as microscopic traces in meteorite impact sites. Machine-learning-driven simulations played a central role in explaining how it forms.
Diamond has a famous rival, a form of itself that most people have never heard of: hexagonal diamond, sometimes called lonsdaleite. In the ordinary cubic diamond of jewellery and industry, the carbon atoms are arranged in a lattice of interlocking cubes; in hexagonal diamond the same atoms sit in a hexagonal arrangement, and theory has long predicted that this rarer structure should be harder than the cubic form, harder by a margin that materials scientists describe with a certain reverence. The problem has always been making it. Hexagonal diamond appears in nature only as microscopic traces, found at meteorite impact sites where immense shock pressures briefly forced carbon into the hexagonal arrangement, and attempts to synthesise it in bulk have struggled. A paper in the 4 March issue of Nature reports that the feat has now been achieved.
The work is a collaboration between theory and experiment: a team led by researchers at Nanjing University performed machine-learning-driven molecular dynamics simulations, while an experimental group at Zhengzhou University carried out the synthesis, and the two halves of the project inform each other. The simulations explain the microscopic mechanism by which the hexagonal structure forms, revealing the phase transition path that the carbon follows, and the experiments then used that understanding to produce the material in bulk rather than in trace quantities. The paper’s title, “Bulk hexagonal diamond,” is the claim in two words.
The significance lies in what the material promises. If the theory holds, hexagonal diamond’s extra hardness could matter for cutting, abrasion and high-pressure applications, and its properties, which have never been measurable on bulk samples, can now be studied properly. A material that existed only as a curiosity, a footnote in the impact-crater literature, becomes a candidate with a manufacturing path. The simulations are equally notable as method: machine-learning-driven dynamics have matured to the point where they can guide a difficult synthesis, predicting the mechanism well enough for experimenters to target it, a working example of the AI-for-science pipeline that has become one of the most discussed trends in materials research.
There are caveats, and the careful reader will note them. Bulk synthesis is reported, but the scale, stability and practical hardness of the product will be tested by other groups, and the gap between a reported phase and a deployable material is often large. Hexagonal diamond’s theoretical edge over cubic diamond has been argued for decades, and the argument is not settled by one synthesis; measurement on bulk samples will now put the theory to the real test. The claim should be received as an important step, not as the end of the story.
The broader point is the one the collaboration embodies. Diamond’s hexagonal form has been known for nearly sixty years, since it was first identified in meteorite debris, and for most of that time it has resisted making. What changed is not chemistry but method: simulations good enough to show the path, and the willingness to let them set the experimental agenda. A material that nature makes only in violent accidents can now be made deliberately, and the route that got there, theory, simulation and experiment working on the same problem, is as interesting as the diamond itself.