Hypersonic Impact Turns Diamond to Graphite in a Diamond–cBN–Cobalt Composite

Diamond, among the hardest natural materials, can turn into graphite under hypersonic impact, researchers have confirmed. When space debris or other objects strike a structure at thousands of meters per second, the material experiences enormous stress in an extremely short time. In such extreme environments, a material may not only deform or fracture; the bonding state of its atoms can change as well.

An international team including researchers from Texas A&M University and the University of Brasília developed a ternary composite made of diamond, cubic boron nitride (cBN), and cobalt, and conducted high-velocity impact experiments using a two-stage light-gas gun. According to a paper published in the journal Materials Today (vol. 99, article 103477, published online August 6, 2026; preprint: arXiv:2602.13188), analysis of post-impact fragments by electron microscopy and other methods confirmed that diamond particles inside the composite had undergone a phase transition to graphite.

The team suggests that this phase transition may have contributed to dissipating impact energy. However, the results come from laboratory materials evaluation and do not mean that practical protective armor has been completed. How much of the impact energy the phase transition accounted for has not been quantified, and further verification is needed to judge its contribution to protective performance.

Spark plasma sintering produced a composite that retained diamond

A major challenge in using diamond as a structural material is how difficult it is to shape and process. Near atmospheric pressure, graphite is the thermodynamically stable crystalline phase of carbon, and diamond is metastable. As a result, when diamond powder is sintered at high temperature, graphitization can progress before the particles bond sufficiently, and the material may fail to retain its original hardness and crystal structure.

To address this problem, the team adopted spark plasma sintering (SPS). The raw materials were synthetic diamond powder with a particle size of 50–100 μm, cubic boron nitride powder of roughly 0.5–1 μm, and cobalt powder of roughly 1–2 μm, mixed in approximately equal weight proportions. Cubic boron nitride has hardness second only to diamond, along with excellent heat resistance and chemical stability, and serves as the matrix of the composite. Cobalt acts as a binder phase that joins the hard particles and aids formation of the sintered body.

For sintering, the chamber was evacuated and then filled with high-purity argon, and the temperature was raised at 50°C per minute. The sample was then held at 1,400°C under a uniaxial pressure of 90 MPa for 60 minutes, producing a cylindrical composite disk. SPS combines electrical heating with applied pressure, and the team succeeded in consolidating the hard particles while suppressing excessive graphitization of the diamond.

The density of the sintered disk, calculated from its dimensions and mass, was about 3.32 g/cm³. X-ray diffraction (XRD) showed clear peaks for diamond and cubic boron nitride, along with a faint graphite peak, suggesting that some of the diamond graphitized during sintering.

Raman spectroscopy also showed the characteristic diamond peak at about 1330.9 cm⁻¹, as well as peaks corresponding to the transverse optical (TO) and longitudinal optical (LO) modes of cubic boron nitride at about 1057.08 cm⁻¹ and 1307.64 cm⁻¹, respectively. The team also tried sintering formulations without cobalt or cBN, and formulations with different particle sizes, but these either failed to yield disks of sufficient strength or showed substantial graphitization. These were comparisons to find suitable sintering conditions, however, not control experiments comparing impact protection performance.

{
  "type": "bar",
  "title": "Raw material mix ratio of the ternary composite",
  "unit": "%",
  "categories": [
    "Diamond (50–100μm)",
    "Cubic boron nitride (0.5–1μm)",
    "Cobalt (1–2μm)"
  ],
  "series": [
    {
      "name": "Weight ratio (approx.)",
      "values": [33.3, 33.3, 33.3]
    }
  ],
  "caption": "Three powders mixed in roughly equal weight proportions. The disk sintered at 1,400°C and 90MPa had a density of about 3.32g/cm³",
  "source": "Materials Today (2026), DOI: 10.1016/j.mattod.2026.103477"
}

This yielded dense composite disks that largely retained the crystal structure of the diamond particles. The team used these samples to examine the impact response in the hypersonic regime.

Two impact experiments above 2,500 m/s

The impact tests used a 45-foot (about 13.7 m) two-stage light-gas gun at Texas A&M University's Hypervelocity Impact Laboratory. A two-stage light-gas gun is an experimental device that uses high-pressure gas to accelerate projectiles to extremely high speeds. The team treated relative velocities of 1,700 m/s or more as hypersonic impacts and carried out two experiments under different conditions on composite disks made by the same method. One disk was tested under each condition.

In the first experiment, multiple spherical particles of aluminum alloy (Al 2017-T4), each 1 mm in diameter, were fired simultaneously. The impact velocity reached 2584.7 m/s (equivalent to Mach 7.5), and high-speed shadowgraph imaging recorded many particles striking the composite surface. After impact, the disk remained intact overall, with no large visible cracks or separated fragments. Slight erosion marks were left on the surface, but XRD and Raman spectroscopy detected signals indicating partial graphitization at the impact site.

In the second experiment, a larger single projectile was used. When a 4 mm diameter aluminum alloy (Al 2017-T4) sphere was fired at 2900.8 m/s (equivalent to Mach 8.45), the composite disk could not withstand the impact and shattered into several large fragments. Analysis of the recovered fragments confirmed substantial graphitization of the diamond particles inside.

However, these results cannot directly determine the composite's failure limit. In the two experiments, the projectile diameter, number, impact configuration, and velocity all changed at the same time. It is impossible to separate how much the increase in projectile mass, the concentration of kinetic energy, and the higher velocity each contributed to the disk's shattering under Condition 2. In addition, each condition was tested only once, so there is insufficient data to assess statistical reproducibility or failure probability.

Shadowgraph imaging recorded the projectiles in flight, the moment of impact, and the scattering of fragments and ejecta. This imaging method cannot directly observe changes in atomic-level bonding, however. The evidence for graphitization came from microstructural analysis of samples recovered after impact, not from real-time measurement of the phase transition as it progressed.

Electron microscopy reveals a diamond–graphite interface

A detailed examination of the sample shattered under Condition 2 revealed major changes in crystal structure inside the composite. In XRD measurements, the diamond diffraction peak, strongly observed in the as-sintered sample, weakened substantially, and a strong peak corresponding to the (002) plane of graphite appeared. Raman spectroscopy likewise showed the diamond characteristic peak diminishing, while the G band (about 1579.70 cm⁻¹), characteristic of graphite's crystal structure, and the D band (about 1345.06 cm⁻¹), which reflects lattice defects and structural disorder, were clearly observed.

The team describes this change as a "nearly complete phase transition" of the diamond particles. However, what graphitized was the carbon component of the composite, and the cubic boron nitride framework was largely preserved. In other words, the entire composite did not turn into graphite; rather, the crystal structure of the embedded diamond particles changed substantially.

To investigate the details of the phase transition, the team cut thin sections from the recovered fragments using a focused ion beam (FIB) and analyzed them with transmission electron microscopy (TEM), high-resolution TEM (HRTEM), four-dimensional scanning transmission electron microscopy (4D-STEM), and STEM-EELS, which combines scanning transmission electron microscopy with electron energy loss spectroscopy. This made it possible to examine not only the orientation relationships of crystal planes but also the local bonding state of carbon atoms.

STEM-EELS confirmed that regions of $s{p}^3$ bonding, which form diamond's characteristic tetrahedral structure, and regions of planar $s{p}^2$ bonding, characteristic of graphite, lay adjacent to each other across a clear boundary. HRTEM and fast Fourier transform (FFT) analysis further showed that the (-111) plane of diamond and the (002) plane of graphite were in contact with a specific orientation relationship. The measured interplanar spacings were about 3.4±0.1 Å for the graphite (002) plane and about 2.09±0.05 Å for the diamond (-111) plane. One angstrom (Å) equals 0.1 nm.

{
  "type": "bar",
  "title": "Interplanar spacings measured at the diamond–graphite interface",
  "unit": "Å",
  "categories": [
    "Graphite (002) plane",
    "Diamond (-111) plane"
  ],
  "series": [
    {
      "name": "Measured interplanar spacing",
      "values": [3.4, 2.09]
    }
  ],
  "caption": "Interplanar spacings measured by HRTEM. Measurement error is ±0.1Å for graphite and ±0.05Å for diamond",
  "source": "Materials Today (2026), DOI: 10.1016/j.mattod.2026.103477"
}

Such an interface structure suggests that the diamond crystal structure changed into graphite while maintaining a certain orientation relationship. However, the interface was not a perfectly coherent crystal lattice; lattice mismatch, local curvature, dislocations, and stacking faults were present. The team confirmed that the interface was locally semi-coherent, and disorder in the stacking direction was also seen in the graphite layers. These are thought to be traces of rapid structural change, but they are not a direct observation of the order in which atoms moved.

Electrical resistivity also changed markedly. The volume resistivity of the as-sintered disk was about 217 mΩ·cm, whereas that of the sample shattered by impact fell to about 36 mΩ·cm. This roughly sixfold change is consistent with the analysis that conductive $s{p}^2$-bonded graphite had formed. However, because the composite also contains metallic cobalt, and fracturing may alter the internal conductive pathways, resistivity alone cannot determine the degree of graphitization. Moreover, the drop in resistivity reflects a change in microstructure and does not directly imply improved protective performance.

Molecular dynamics reproduces rapid bond changes

To investigate the mechanism of a phase transition that proceeds over an extremely short time, the team also ran simulations using reactive force field molecular dynamics (ReaxFF). ReaxFF can calculate the processes by which bonds between atoms break and form, allowing changes in crystal structure during impact to be tracked at the atomic level.

The calculations simulated a high-speed particle striking a diamond crystal block made of 4,000 carbon atoms. The impact produced bond-angle distortion and strong shear stress, and within just 5 picoseconds ($5\times 10^{-}12$ seconds), the fraction of $s{p}^3$ bonds fell to about 42% while $s{p}^2$ bonds reached about 58%. A similar bond change was confirmed in a model expanded to 10,976 atoms, showing that graphitization can progress in an extremely short time at the atomic level.

However, this simulation does not completely reproduce the composite used in the experiments. The model represents a pure diamond crystal and does not include the cubic boron nitride, the cobalt, or the complex microstructure formed by sintering in the actual material. Nor is it a calculation that directly reproduces an impact by a millimeter-scale aluminum alloy sphere.

Therefore, the result that $s{p}^2$ bonds reached about 58% within 5 picoseconds is a value within an idealized computational model. It does not mean that graphitization proceeded at the same speed or to the same degree throughout the actual composite, nor is it a figure that directly indicates fracture strength or energy absorption.

Regarding the mechanism that triggered the phase transition, the team believes that compression by shock waves, shear stress, non-equilibrium changes during unloading, and a local temperature rise were involved. At the same time, the catalytic action of the cobalt in the composite cannot be ignored. Transition metals such as cobalt can participate in the dissolution and re-precipitation of carbon at high temperatures and may promote graphitization of diamond.

It has been suggested that in these impacts, too, cobalt may have acted on the diamond surface under the momentary high-temperature, high-pressure environment and accelerated graphitization. That is, mechanical bond changes caused by the shock may have combined with thermal reactions involving the metal catalyst. However, the respective contributions have not yet been separated quantitatively, and it remains unclear which played the dominant role.

Can diamond graphitization be used in protective armor?

The key possibility this study points to is that the diamond phase transition might be used to dissipate impact energy. In the change from diamond to graphite, carbon bonds rearrange from $s{p}^3$ to $s{p}^2$, and the crystal structure and density change. In general, graphite is less dense than diamond, so this phase transition is accompanied by an increase in volume. The team presents the possibility that such rapid structural change contributed substantially to energy dissipation during impact.

However, it is not accurate to regard the phase transition itself as a simple endothermic reaction. According to thermochemical data from Argonne National Laboratory, the change from diamond to graphite under standard conditions is an exothermic-direction reaction in which enthalpy decreases. Because shock involves bond breaking and reformation, deformation, and fracture, irreversible structural change may play a role in dissipating mechanical energy. But the mere fact that graphitization was observed does not allow one to judge how much energy the phase transition absorbed or dissipated.

The paper also does not quantitatively measure how the kinetic energy delivered by the impact was distributed among the phase transition, the creation of new surfaces through fracture, the scattering of fragments, deformation of the material, heat, and so on. The team discusses the possibility that the phase transition was a major factor in energy dissipation, but no detailed energy budget supporting this is presented. While the conversion of diamond to graphite was confirmed experimentally, how much that change suppressed fracture of the material remains unresolved, which defines the standing of this result.

No comparison tests have been conducted against existing ceramic protective materials such as silicon carbide, or composite armor used in aerospace. That the disk survived impact by multiple 1 mm particles indicates impact resistance under those test conditions, but it does not prove superiority over common protective materials. Likewise, the result of the disk shattering under a single 4 mm sphere provides insufficient information to quantitatively determine the material's limits.

A research summary published by Texas A&M University on November 3, 2026 explains that the result could lead to the design of materials for aircraft, spacecraft, and various protective applications. Materials to protect spacecraft from high-speed debris are one envisioned application. However, this is a future research direction, not a stage at which practical protective performance has been verified or certified.

Going forward, impact tests that vary projectile mass, diameter, velocity, and shape individually will need to be repeated to examine the conditions under which fracture occurs and its reproducibility. It will also be important to separate shock-induced phase transition from metal-catalyzed graphitization, for example by using comparison samples that contain no cobalt. Furthermore, the energy balance before and after impact must be measured precisely to clarify how much energy dissipation the phase transition actually accounts for.

This study demonstrates, at the level of microstructure, that the crystal structure of diamond changes rapidly in the extreme environment of hypersonic impact. This phenomenon may make it possible to design materials that respond to impact through a mechanism different from conventional ones. However, evaluating practicality as a protective material will require further elucidation of the phase transition mechanism and quantitative demonstration of its energy-dissipating effect.