🔬 materials science

Vibrational spectroscopy identifies the bond asymmetry of hexagonal diamond

By combining first-principles lattice dynamics with Raman spectroscopy and diffraction data, this study resolves the structural controversy of bulk hexagonal diamond by demonstrating that its interlayer bonds are longer than its intralayer bonds, thereby identifying a specific positive bond asymmetry that contradicts previous conflicting refinements.

Original authors: Li Zhu

Published 2026-08-04

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Li Zhu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where the hardest substance known to humanity, diamond, comes in two flavors: the familiar round, bouncy kind we find in jewelry, and a rarer, hexagonal version that scientists have been chasing for decades. Think of these two flavors as different ways of stacking layers of carbon atoms, like building a tower with either a square grid or a honeycomb grid. For a long time, the hexagonal version was a ghost—mostly found as tiny, messy crumbs in meteorites or created in labs under extreme pressure, making it hard to study. Scientists wanted to know exactly how the atoms were arranged because the strength and behavior of this material depend entirely on the tiny distances between its atoms. Specifically, they were arguing over a microscopic tug-of-war: are the bonds holding the layers together slightly shorter or slightly longer than the bonds holding the atoms together within a single layer? Getting this right matters because if we can master this material, it could lead to super-strong tools or new electronics, but first, we have to agree on what the material actually looks like.

Recently, two different research teams built large, pure samples of this hexagonal diamond and took pictures of the atoms using X-rays. But here is the twist: when they measured the distances between the atoms, they got completely opposite answers. One team said the vertical bonds were shorter, while the other said they were longer. It was like two people looking at the same building and one saying the roof is lower than the walls, while the other insists the roof is higher. This disagreement was so big that it left the scientific community confused about the true nature of this material.

Enter a new study that acts like a detective using a different kind of clue. Instead of just looking at the static picture of the atoms, this team listened to the material "sing." They used a technique called vibrational spectroscopy, which is like tapping a glass to hear its ring. Every molecule has a unique set of notes it can sing based on how its atoms are connected. By simulating how these notes should sound for different atomic arrangements, the researchers could test which of the two conflicting pictures was actually real.

The investigation revealed that the two previous "pictures" were likely blurry or misleading. The new study, using powerful computer simulations, found that the hexagonal diamond naturally wants to stretch its vertical bonds slightly, making them longer than the horizontal ones by about 24 thousandths of an angstrom (a unit so small it's hard to imagine). This stretching happens because of a specific way the layers stack, similar to how a person standing with their feet directly under their shoulders (an "eclipsed" position) feels a bit more tension than someone standing with feet staggered. The simulations showed that this specific stretching creates a distinct "song" or vibrational frequency.

When the researchers compared their simulated songs to the actual songs recorded from the real samples, the results were clear. The two conflicting structures proposed by the other teams simply didn't match the music. One team's structure predicted a song that was way too high-pitched, while the other predicted a song that was too low and had the wrong rhythm. The only structure that matched the real-world "song" was the one where the vertical bonds were slightly longer, confirming the older, less precise measurements from 2003 and the new computer models.

The study also tackled a mysterious high-pitched note that one team had heard at 1,529 cm⁻¹. The researchers concluded this wasn't a note from the perfect hexagonal diamond at all. Instead, they suspect it was a "ghost note" coming from a tiny bit of leftover graphite or disordered carbon that got stretched or squeezed during the experiment, much like a guitar string that is slightly out of tune.

In short, this paper solves a decades-old mystery by listening to the material's vibrations. It suggests that the hexagonal diamond has a specific, slightly stretched structure that makes it unique. It rules out the idea that the bonds are shorter or that the material is perfectly symmetrical. While the author is very confident in their simulations and how well they match the experimental data, they acknowledge that the exact reason for the "ghost note" needs more investigation. They propose that future experiments should look at the material with different types of light to confirm these findings, but for now, the evidence points to a hexagonal diamond that is slightly stretched, not squished.

Technical Summary: Vibrational Spectroscopy Identifies the Bond Asymmetry of Hexagonal Diamond

Problem Statement
Hexagonal diamond (HD, or lonsdaleite) has long been sought as a hexagonal counterpart to cubic diamond with potentially superior mechanical properties. However, its fundamental structural parameter—the asymmetry between its two inequivalent bond lengths—remains contested. Each carbon atom in HD forms three intralayer bonds ( $O_A$ ) within buckled honeycomb layers and one interlayer bond ( $O_B$ ) along the hexagonal axis. Recent bulk syntheses have yielded conflicting structural refinements regarding the sign and magnitude of the bond asymmetry ( $O_B - O_A$ ). One study (Yang et al.) reported a short interlayer bond ( $O_B - O_A \approx -55$ mÅ), while another (Lai et al.) reported a long interlayer bond ( $O_B - O_A \approx +183$ mÅ). These values differ by 238 mÅ, a discrepancy far exceeding the precision of cubic diamond bond measurements. Furthermore, both recent refinements deviate from an earlier 2003 determination ( $O_B - O_A \approx +60 \pm 45$ mÅ) and from each other. The core problem is determining the true equilibrium structure of bulk HD and resolving which of these competing models is physically consistent with experimental observations.

Methodology
The author employs first-principles lattice dynamics based on Density Functional Theory (DFT) to adjudicate between the competing structural models.

Computational Framework: Calculations utilize the dispersion-corrected meta-GGA functional (r2SCAN+rVV10) and the PBE functional for control. Structural relaxations are performed for the $P6_3/mmc$ space group (2H polytype).
Lattice Dynamics: Phonon frequencies and Raman activities are calculated using finite-displacement methods (Phonopy) on 72-atom supercells. The study maps the zone-center phonon frequencies and Raman intensities as a function of the internal coordinate ( $z$ ) and bond lengths ( $O_A, O_B$ ).
Spectroscopic Gauge: The author establishes a sensitivity map linking the frequency of the bright zone-center $A_{1g}$ mode to the interlayer bond length ( $O_B$ ). They invert the measured Raman spectra from phase-pure and impact-diamond samples to infer bond lengths, using the calculated sensitivity ( $\approx -2,100$ cm $^{-1}$ Å $^{-1}$ ) as a gauge.
Comparative Analysis: The calculated spectra for the relaxed structure and the two published Rietveld refinements (Yang and Lai) are compared against the measured Raman spectra of the same samples. The study also tests "fixed- $O_A$ " constructions and least-squares compressed structures to see if any mechanically viable state can reproduce the full set of observed bands, including the high-frequency 1,529 cm $^{-1}$ feature reported by Yang et al.

Key Contributions and Results

Equilibrium Bond Asymmetry: Full relaxation of HD reveals a robust, small positive bond asymmetry where the interlayer bond is longer than the intralayer bonds ( $O_B - O_A \approx +24$ mÅ). This result is consistent across functionals and is attributed to the "eclipsed" conformation of the interlayer bond, which induces filled-orbital repulsion. This finding aligns with the sign and magnitude of the 2003 refinement and contradicts the negative asymmetry reported by Yang et al.
The $A_{1g}$ Mode as a Structural Gauge: The study identifies the $A_{1g}$ zone-center mode as a direct gauge of the interlayer bond length. Its frequency shifts linearly with $O_B$ at a rate of $\approx -2,100$ cm $^{-1}$ Å $^{-1}$ . Crucially, this mode carries the dominant Raman activity near equilibrium.
Inconsistency of Recent Refinements:
- Lai et al.: The structure with a long interlayer bond ( $O_B - O_A \approx +183$ mÅ) predicts an $A_{1g}$ frequency (959 cm $^{-1}$ ) far below the measured envelope and fails to reproduce the observed pattern.
- Yang et al.: The structure with a short interlayer bond ( $O_B - O_A \approx -55$ mÅ) predicts an $A_{1g}$ frequency (1,399 cm $^{-1}$ ) that does not match the measured dominant band. While a specific "fixed- $O_A$ " construction with $O_B \approx 1.44$ Å reproduces the 1,529 cm $^{-1}$ band, it fails to reproduce the lower-frequency bands (predicting them at $\approx 1,000-1,160$ cm $^{-1}$ ) and requires unphysical confining stresses ( $\approx 77$ GPa) and lattice constants excluded by the author's own X-ray diffraction data.
Inversion of Measured Spectra: Inverting the Raman spectrum of the phase-pure sample (Lai et al.) using the DFT-calibrated sensitivity map yields a bond asymmetry of $O_B - O_A = 24 \pm 3$ mÅ. This value is consistent with the relaxed DFT prediction and the 2003 diffraction refinement.
Origin of the 1,529 cm $^{-1}$ Feature: The high-frequency band observed in some samples (e.g., Yang et al.) cannot be assigned to a homogeneous ideal 2H diamond structure. The author proposes it arises from extrinsic sources, likely strained $sp^2$ carbon (graphitic) at grain boundaries or twins, which can shift the G-band to 1,529 cm $^{-1}$ under $\approx 1\%$ biaxial tension. This assignment is supported by the variability of this band across different samples and its absence in the phase-pure sample.

Significance and Claims
The paper claims to resolve the structural controversy of hexagonal diamond by demonstrating that vibrational spectroscopy provides a sharper discrimination of structure than the available diffraction data.

Resolution of Controversy: The study concludes that the intrinsic structure of hexagonal diamond possesses a long interlayer bond ( $O_B > O_A$ ), with an asymmetry of approximately $+24$ mÅ. This supports the oldest determination and the theoretical prediction based on eclipsed bonding, while rejecting the two most recent, mutually exclusive refinements.
Methodological Validation: The work establishes that the $A_{1g}$ Raman mode is a reliable, direct gauge for the interlayer bond length in HD, allowing for the inversion of spectra to determine structural parameters.
Limitations of Current Data: The author asserts that the two recent Rietveld refinements are inconsistent with the measured Raman spectra of their own samples and with each other. They caution that the refined coordinates should be used with care pending re-refinement.
Extrinsic Origins: The paper argues that anomalous spectral features (such as the 1,529 cm $^{-1}$ band and the negative bond asymmetry suggested by HRTEM in specific nanometric regions) are likely artifacts of local strain, stacking faults, or minority $sp^2$ phases, rather than intrinsic properties of the ideal bulk crystal.

The study does not propose new experimental protocols but calls for polarization-resolved ultraviolet Raman and statistical HRTEM surveys to further validate these conclusions. It emphasizes that the "eclipsed" bond mechanism is a local, transferable property of hexagonal stacking that dictates the structural asymmetry.

More like this