Towards the real diamond formation conditions - insights and perspective
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Nature Communications |
| Published | 2026-10-30 |
| Authors | Mara Murri, Mattia Gilio, Fabrizio Nestola, Thibault Duretz, Federica Marone |
| DOI | 10.1038/s41467-026-79094-4 |
A diamond can sustain about 2.5 GPa of olivine inclusion pressure before minor fracturing begins. The resulting pressure loss is partial, so elastic geobarometry can understate the entrapment pressure by as much as 2.5 GPa at 1150 °C rather than erase the formation record.
Residual pressure can misrepresent formation pressure
Section titled “Residual pressure can misrepresent formation pressure”Mineral inclusions can retain the pressure and temperature at which a diamond formed. Elastic geobarometry combines the residual pressure inside an inclusion with the thermo-elastic properties of the inclusion and diamond. This produces an isomeke, a line of possible entrapment conditions in pressure and temperature space.
About 99% of all diamonds are lithospheric and form at depths of 160 to 190 km. Super-deep diamonds account for less than 1% and form at depths greater than 300 km. Classical geothermobarometry estimates pressure and temperature from inclusion chemistry, but it requires chemical equilibrium that is often difficult to demonstrate. Elastic geobarometry avoids this chemical requirement and can, in principle, apply to any diamond inclusion system.
Its main assumption is less secure. Among more than 500 diamonds studied for elastic geobarometry, about 90% of inclusions are surrounded by cracks and often a thin fluid film. Cracking marks non-elastic behavior. Most calculated residual pressures for olivine inclusions fall below 0.5 GPa. At 1150 °C, their calculated entrapment pressures lie within the graphite stability field even though diamond hosts the inclusions.
Sample and Raman experiment
Section titled “Sample and Raman experiment”The authors examined an octahedral lithospheric diamond from the Udachnaya kimberlite in Siberia. Micro-Raman spectroscopy identified fifteen inclusions: fourteen olivines and one magnesiochromite. The mineral assemblage indicates a peridotitic origin.
Inclusion 1 was rounded and lacked visible fractures. Inclusion 10 was prismatic, elongated, and carried a halo interpreted as a planar fracture. Both were selected to limit stress-field interactions with nearby inclusions. They lay about 50 and 170 μm below the diamond surface, respectively.
Raman spectra were collected from 300 to 1100 K in 100 K increments at atmospheric pressure under argon flux. A free olivine from Kittelfjäll, Sweden, served as the reference. Its Fo92.4 composition closely matched the common Fo92.7 composition of Udachnaya inclusions. The authors state that this small difference did not affect Raman peak positions.
Four olivine modes were tracked at approximately 822, 855, 918, and 961 cm⁻¹. The experiment used a 532 nm laser with about 8 mW at the sample. Spectral resolution was about 2 cm⁻¹, and peak-position reproducibility was about 0.5 cm⁻¹. Each sample remained at the target temperature for 10 minutes before data collection.
Cracking appears during heating
Section titled “Cracking appears during heating”All four olivine Raman modes shifted toward lower wavenumbers as temperature increased. The free crystal and both inclusions showed the same direction of shift. Inclusion peaks remained at higher absolute wavenumbers and had a shallower temperature-dependent slope. The authors attribute this difference to retained residual pressure caused by the elastic contrast between olivine and diamond.
Cooling was generally reversible for inclusion 1 and the free reference. Inclusion 10 behaved differently. Its 918 cm⁻¹ mode dropped at 900 K during heating, while its 961 cm⁻¹ mode changed slightly at 300 K during cooling. The authors interpret these changes as evidence of pressure release.
Post-heating microscopy also showed new fracture propagation around inclusion 2. Synchrotron X-ray tomography detected fracture propagation around inclusion 12 and minor propagation around inclusion 2. The tomography used 3000 radiographs at 16 keV and produced images with 0.24 μm pixels. The imaging therefore documented cracking in inclusions that were not monitored directly by the Raman heating sequence.
The interface contains a thin film
Section titled “The interface contains a thin film”A Raman transect across inclusion 1 detected a strong boundary signal between the olivine and diamond. It contained two broad bands within the 500 to 1000 cm⁻¹ region, with features near 640 and 780 cm⁻¹. The signal disappeared when the laser moved from the rim into either phase. Based on the strongest signals, the authors estimated a film thickness of about 3 to 4 μm.
The authors treat this feature as a possible fluid film. Earlier work proposed that such films may contain Si2O(OH)6 dimers and Si(OH)4 monomers. The present experiment did not determine the film chemistry.
The thermo-elastic model omitted the film, yet reproduced the measured pressure rise within experimental uncertainty. Angel et al. reported in 2023 that a continuous 1 μm water film produced inclusion-pressure differences within the experimental uncertainty of 0.3 GPa. These results support a moderate influence of interfacial fluid on the calculated residual pressure.
A radial model reproduces thermal pressurisation
Section titled “A radial model reproduces thermal pressurisation”The authors used a one-dimensional finite-difference thermo-elastic model with radial symmetry. The model represented the olivine inclusion and diamond host with a cohesive interface and zero pressure at the outer boundary. Incremental heating was treated as instantaneous, so heat diffusion was neglected.
A third-order Birch Murnaghan equation of state represented the mechanical contribution. An Einstein model represented thermal pressure, and temperature-dependent shear moduli were included for both phases. The Julia code used MineralEOS.jl, which was validated against EosFit7c.
Residual pressures were calculated from the four Raman modes using the calibration of Liu et al. from 2024. The experimental and modelled results were as follows.
| Inclusion | Condition | Pinc at 300 K | Pinc at 1100 K | Modelled Pinc at 1100 K |
|---|---|---|---|---|
| 1 | Rounded, no visible fracture | 0.8 ± 0.3 GPa | 2.9 ± 0.5 GPa | 2.59 GPa |
| 10 | Elongated, pre-existing planar fracture | 0.5 ± 0.1 GPa | 2.5 ± 0.5 GPa | 2.36 GPa |
An independent EosFit7c calculation gave 2.78 GPa for inclusion 1 at 1100 K, consistent with the Raman estimate. Deviatoric stress around the inclusion reached about 2 GPa. At 1100 K, the three normal stress components within the inclusion were equal. Host stresses decayed proportionally to 1/r³ and approached zero within about three inclusion radii.
Fracturing causes partial pressure loss
Section titled “Fracturing causes partial pressure loss”Fast adiabatic decompression during kimberlite eruption disrupts the thermo-elastic equilibrium established during diamond growth. Olivine is softer than its diamond host, so decompression can generate a large pressure difference inside the inclusion. For a pressure drop from 5.9 to 0.5 GPa, that difference can reach 3.3 GPa.
The heating experiments show that diamond can sustain about 2.5 GPa before small fractures form. Fracturing occurs while the inclusion remains in a compressional regime, so the pressure loss is partial. The measured pressure still contains information about the original entrapment conditions.
This differs from zircon inclusions in garnet. Their transition from a compressional to a tensile regime causes a complete elastic reset and removes the original entrapment information. Diamond fractures around olivine retain partial information instead.
Inclusion 1 had a residual pressure of 0.8 GPa at 300 K, allowing a recalculated entrapment pressure close to the diamond stability field. Inclusion 10 had only 0.5 GPa. Its expected residual-pressure range for lithospheric diamond formation is 0.9 to 1.5 GPa, so at least 0.4 GPa of pressure was probably released. The common residual pressure near 0.5 GPa therefore does not necessarily indicate a complete reset.
Limits for natural diamond pressure studies
Section titled “Limits for natural diamond pressure studies”Hydrostatic pressure estimates varied among the four Raman modes. For seven inclusions, three modes generally clustered at positive pressures. The 961 cm⁻¹ mode instead produced negative values. The authors attribute this contrast to non-hydrostatic stress, consistent with earlier single-crystal X-ray diffraction results.
The radial model simplifies several real features, including inclusion geometry, fluid rims, anisotropy, and fracture damage. The authors state that the fracture threshold for a natural diamond is difficult to determine because it may vary with impurities, mineral assemblages, inclusion shapes, existing fractures, and fluid chemistry.
Future modelling must combine anisotropic thermo-elasticity, inclusion geometry, kinetic effects, damage mechanics, and plastic deformation. Such models could estimate the pressure lost through fracturing and correct the calculated formation pressure.
For researchers selecting natural diamonds as geological pressure records, a low residual pressure should not automatically disqualify an inclusion. It can still retain formation information. An uncorrected isomeke may nevertheless lie below the diamond stability field and understate the original depth. Fracture history, Raman-mode consistency, inclusion geometry, and interfacial chemistry must be considered before assigning a formation pressure.
View Original Abstract
Pressure and temperature conditions of diamond formation provide valuable insights for reconstructing paleo-geotherms and the deep carbon cycle. These conditions can be determined by investigating mineral inclusions in diamonds using elastic geobarometry, which can be applied to any mineral inclusion in a diamond. However, it assumes purely elastic host-inclusion behaviour throughout exhumation. In reality, almost all mineral inclusions in diamonds are surrounded by cracks. Here, we combine in situ high-temperature micro-Raman spectroscopy with thermo-elastic modelling to evaluate the pressure-release process in one of the most common diamond inclusions, olivine [(Mg,Fe)2SiO4]. Our results indicate that a diamond can sustain pressures of ~2.5 GPa before undergoing minor fracturing, which induces only partial loss of inclusion pressure rather than a complete reset. The estimated pressure loss can lead to an underestimation of up to 2.5 GPa in the entrapment pressure at 1150 °C, a typical temperature for diamond formation. This study shows that olivine in diamond can undergo partial release of pressure during kimberlite eruptions due to fast adiabatic decompression, leading to cracked olivine inclusions and entrapment pressures underestimated by up to 2.5 GPa at 1150 °C
Single Crystal Cvd Diamond Plate
View Specs →
Advanced Thermal Simulation Service Diamond Heat Spreader Performance Analysis
View Specs →