Vibrational spectroscopy identifies the bond asymmetry of hexagonal diamond

Li Zhu · 2026

Read Vibrational spectroscopy identifies the bond asymmetry of hexagonal diamond by Li Zhu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2608-00138.txt.

Introduction

Zhu tests rival structures of hexagonal diamond with vibrational spectroscopy. Bulk hexagonal diamond has been made by independent routes. Its structure stays contested, though. Two recent refinements disagree on the sign of a bond-length gap. They also depart from a 2003 refinement. Zhu uses first-principles lattice dynamics to judge them.

Relaxed hexagonal diamond has an interlayer bond longer than intralayer bonds by 24 milli-angstroms. A bright Raman mode gauges that interlayer bond directly. Neither refined structure reproduces the full measured mode pattern, he finds. The phase-pure sample's spectrum inverts to a small positive asymmetry of 24 plus or minus 3. Two other samples give larger positive values. One spectral feature at 1,529 stays unassigned, an open puzzle.

Catalog record

CREATORLi Zhu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE78 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2608.00138

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Vibrational spectroscopy identifies the bond asymmetry of hexagonal diamond Li Zhu Abstract Bulk hexagonal diamond has been synthesized by independent routes, but its structure remains contested: the two recent refinements disagree even on the sign of the difference between its two inequivalent bond lengths, 238 mÅ apart, and both depart from an earlier 2003 refinement. Here we test the competing structures with first-principles lattice dynamics. Relaxed hexagonal diamond has an interlayer bond longer than the intralayer bonds by 24 mÅ in both functionals, an effect of its eclipsed conformation that scales with polytype hexagonality. The bright zone-center A 1 ​ g A_{1g} mode gauges the interlayer bond at ≈ − 2,100 \approx\!-2{,}100 cm -1 Å -1 , and neither refined coordinate reproduces the full pattern of measured modes. The only structure matching the twinned sample’s three bands requires tens-of-gigapascals confining stress and lattice constants excluded by its own diffraction. Raman spectroscopy and diffraction jointly select a small positive bond asymmetry: inverting the spectrum of the phase-pure sample gives O B − O A = 24 ± 3 O_{B}-O_{A}=24\pm 3 mÅ (95% interval), and two determinations on separate samples give + 33 ± 8 +33\pm 8 and + 60 ± 45 +60\pm 45 mÅ. The 1,529 cm -1 feature cannot be assigned to homogeneous ideal 2H diamond, and the local HRTEM observation remains an open puzzle. Introduction Hexagonal diamond (HD, lonsdaleite) has been pursued for six decades as the hexagonal counterpart of cubic diamond (CD), with predicted mechanical properties rivalling or exceeding it 3 , 12 , 27 . For most of that period it was accessible only as fine, heavily faulted powders from shock or static compression, or as a minority component of meteoritic carbon. Even its status as a distinct crystalline phase was contested: imaging and diffraction analyses argued that much of what had been reported as lonsdaleite is better described as faulted and twinned cubic diamond 25 , 30 , while shock-compression experiments followed its formation from graphite in real time 16 , 35 . Its defining structural feature is the inequivalence of its bonds: each carbon forms three intralayer bonds ( O_{A} ) within buckled honeycomb layers and one interlayer bond ( O_{B} ) along the hexagonal axis. Whether O_{B} is shorter or longer than O_{A} , and by how much, is the most elementary structural question about this material, and the literature of the last two years has returned mutually exclusive answers to it. The question acquired urgency with the recent wave of bulk-HD syntheses: from quasi-hydrostatic compression of single-crystal graphite 36 , from heating of post-graphite phases 4 , and from direct transformation of highly oriented pyrolytic graphite under predominantly uniaxial c -axis stress 39 , 20 . Remarkably, the two structure determinations published since bulk synthesis was achieved contradict each other. Yang et al. 36 refine a short interlayer bond ( O_{B}=1.510 versus O_{A}=1.565 Å, from z\approx 0.0693 ; their Extended Data Fig. 7) and match their ultraviolet Raman spectrum with simulations at O_{B}=1.44 Å. Lai et al. 20 , for a millimetre-sized sample that is phase-pure by diffraction and photoemission, refine a structure whose interlayer bond is instead long ( O_{B}=1.691 Å versus O_{A}=1.508 Å, from z=0.0479(5) ; their Extended Data Table 1). The two refinements thus place the bond asymmetry O_{B}-O_{A} at -55 and +183 mÅ respectively, a 238 mÅ disagreement in a material whose cubic polymorph has bond lengths known to sub-mÅ accuracy. Yang et al. 36 additionally report a local high-resolution transmission-electron-microscopy (HRTEM) measurement of the same sign as their refinement but larger magnitude, nominally O_{B}-O_{A}=-80 mÅ; we keep that observation distinct from the refinements throughout, and return to it in the Discussion. The two reported structures have not been reconciled. Both also depart from an earlier determination that the recent literature has overlooked. Yoshiasa et al. 38 refined hexagonal diamond synthesized in a Kawai-type apparatus against a two-phase model of HD and stacking-faulted cubic diamond, and obtained c/a=1.668(6) , an internal parameter u=0.380(8) (equivalently z=0.060(4) ), and an apical (interlayer) bond longer than the basal (intralayer) one, 1.59(4) versus 1.53(2) Å, giving O_{B}-O_{A}=+60\pm 45 mÅ. Neither Yang et al. 36 nor Lai et al. 20 cites this work. Measured on the common internal coordinate, and in units of the 2003 refinement’s own uncertainty ( \sigma=0.008 on u ), the 2003 value sits 0.7\sigma from the relaxed first-principles structure reported below and 2.3\sigma and 3.0\sigma from the two recent refinements respectively. We do not read that 0.7\sigma as a tight match: an uncertainty of \pm 0.008 is wide enough that agreement is easier to come by than disagreement, and a loosely determined coordinate will sit close to many structures. The force of that refinement lies elsewhere. Its axial ratio c/a=1.668(6) is well determined ( 5.8\sigma above the ideal 1.633) and by itself implies a long interlayer bond: at the ideal internal coordinate, that cell already gives O_{B}-O_{A}=+29 mÅ. Its bond-length budget and stated bond-strength ordering match ours, and the two recent refinements fall outside it in opposite directions. The sign of its refined asymmetry, by contrast, is only a 1.3\sigma statement on its own bond errors. The elementary question is therefore not open in the sense of being untouched, but contested: three determinations exist, the oldest agrees with theory, and the two newest disagree with it and with each other. Here we show that the vibrational spectrum both groups measured adjudicates this contradiction and calls both refined internal coordinates into question. Density-functional-theory (DFT) lattice dynamics establishes three facts. First, relaxed HD has a small, robust asymmetry of the opposite sign to Yang’s determination: O_{B} exceeds O_{A} by 24 mÅ, in both a generalized-gradient-approximation (GGA) and a dispersion-corrected meta-GGA functional, with an identifiable local mechanism, the eclipsed conformation of the interlayer bond. This is the sign, and close to the magnitude, that Yoshiasa et al. 38 measured; theory here corroborates the oldest determination rather than adjudicating alone. Second, the bright A_{1g} zone-center mode is, to an excellent approximation, the interlayer-bond stretch. Its frequency moves at \approx\!-2{,}100 cm -1 Å -1 with O_{B} along the sampled path, and it carries the dominant Raman activity near equilibrium, remaining dominant or co-dominant across the range. The Raman spectrum is therefore a direct gauge of the contested bond. Third, applying this gauge to the published structures shows both Rietveld coordinates to be inconsistent with the spectra measured on the same samples, within the ideal-crystal harmonic model, whereas the relaxed structure is consistent with both at the level of the fitted component positions, and with a third, independently measured lonsdaleite spectrum 15 . No mechanically viable, diffraction-consistent P6_{3}/mmc structure reproduces Yang’s 1,529 cm -1 feature together with their other two bands, leaving extrinsic sp 2 carbon (strained, disordered, or both) as the leading candidate for its origin. Results Relaxed hexagonal diamond has a longer interlayer bond Full relaxation of HD ( P6_{3}/mmc , 4f Wyckoff; the 2H polytype in Ramsdell notation) with the dispersion-corrected meta-GGA functional r 2 SCAN+rVV10 13 , 28 , 26 gives a=2.5024 Å, c=4.1679 Å, z=0.06277 , hence O_{A}=1.537 Å ( \times 3) and O_{B}=1.561 Å ( \times 1). The interlayer bond is thus longer by 24.1 mÅ (Fig. 1 a). The

Read the complete text