Diamond’s first nuclei reveal how graphite changes as crystals grow
High-pressure experiments and atom-level simulations connect the birth of diamond inside graphite with its later growth. By restricting growth in carbon nanotubes, researchers preserved early structures for electron microscopy. The peer-reviewed study finds that the apparent transformation pathway depends on the scale being observed, helping reconcile differences between earlier calculations and experiments.
Science··Evening
The beginning of a diamond
Diamond formation inside graphite begins differently from the way a larger crystal grows, according to a peer-reviewed investigation combining pressure experiments with atomic simulations. Both materials are carbon, but their atoms have different bonding arrangements. The researchers connected fleeting bonds between graphite layers, lasting diamond nuclei and subsequent crystal growth. The sequence offers an explanation for why a calculation following the first atoms can appear to disagree with an experiment examining a much larger transformed specimen.[1]
Confinement preserves early structures
To hold back growth long enough to examine early structures, the team used the narrow space within carbon nanotubes. It compared these confined specimens with larger samples of strongly aligned graphite, where later growth could proceed. High-resolution electron microscopy examined material recovered after exposure to high pressure and temperature. The images preserve structural snapshots from the experiments, allowing the initial nuclei to be compared with features observed during more extensive growth.[1]
Growth takes different directions
Once permanent nuclei form, growth advances unevenly in different directions. Neighbouring nuclei can merge, and the boundaries between crystalline grains rearrange as the material develops. Small nuclei also carry strain and lattice dimensions unlike those of larger diamond crystals. Pressure, temperature and the starting carbon structure influence the pathway. The accelerated simulations help interpret the recovered shapes, while their computational steps cannot be read as the real elapsed time of a transformation.[1]
