DIAMOND CRYSTAL WITH Y-DEFECTS - SPECTROSCOPY AND TRANSMISSION ELECTRON MICROSCOPY
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Russian Geology and Geophysics |
| Published | 2026-04-17 |
| Authors | A.A. Shiryaev, E. A. Vasil’ev, А Ю Васильев, В. В. Артемов, N. V. Gubanov |
| DOI | 10.2113/rgg20254977 |
Transmission electron microscopy links the Y-defect-rich zone of a natural type Ib-IaA diamond to thin vacancy clusters and dislocation debris. More than 40 narrow photoluminescence lines appeared under 787 nm excitation, but their positions and intensities changed over distances of about 1 µm. The authors propose that recombination at nanovoid walls causes this variation. They do not treat that mechanism as proven.
Y-defect absorption falls where nitrogen aggregates
Section titled “Y-defect absorption falls where nitrogen aggregates”In the high-absorption zone, absorption at the diamond Raman frequency reached 2.7 cm⁻¹. C-center absorption at 1344 cm⁻¹ stayed below 0.3 cm⁻¹. At the crystal periphery, A- and C-center concentrations reached 70 and 166 ppm, respectively, while the 1332 cm⁻¹ band fell to 1.4 cm⁻¹.
This measured anticorrelation occurs between high Y-defect absorption and low concentrations of aggregated nitrogen. It conflicts with the proposed view of the Y-center as an intermediate stage in the C to A transformation.
The authors extracted concentrations from N_A = 16.5 × α(A1282) ppm and N_C = 37 × α(C1344) ppm. The A-center estimate is reliable only at the periphery because overlapping one-phonon bands complicate the central envelope.
The 1332 cm⁻¹ band is partly associated with Y-defects, but N⁺ or other unidentified defects may contribute. The paper therefore treats this band as a spatial marker, not a unique Y-center signature.
Sampling and spectroscopy
Section titled “Sampling and spectroscopy”After infrared screening of type Ib-IaA diamonds from the Yubileinaya kimberlite pipe in Yakutia, the authors selected cubic crystal U-14. It had the strongest Y-defect absorption among the screened stones.
FTIR spectra covered 400 to 7800 cm⁻¹ at 1 cm⁻¹ resolution. A VERTEX-70 spectrometer and Hyperion1000 microscope collected 32 scans per point through a 100 × 100 µm aperture. Spectra were normalized to the two-phonon lattice absorption.
IR spectra also contained carbonate bands at 877 and 1450 cm⁻¹ and H₂O bending bands at 1600 to 1700 cm⁻¹. Broad bands from 1000 to 1300 cm⁻¹ were attributed tentatively to silicates. A band near 3680 cm⁻¹ was linked to mica OH groups. These signals show that the Y-defect zone also contained mineral and fluid inclusions.
Photoluminescence was measured at 77 K with a Renishaw InVia spectrometer and a ×50 objective. The authors used 488 nm and 787 nm excitation, with collection ranges of 490 to 750 nm and 790 to 1050 nm, respectively.
Cathodoluminescence used cold-cathode plasma excitation at 5 kV and 100 to 200 µA. Three TEM foils were cut with a Ga⁺ ion beam near (110) and (111) orientations. The OSIRIS TEM/STEM operated at 200 kV with HAADF imaging and energy-dispersive detection.
Photoluminescence varies over micron distances
Section titled “Photoluminescence varies over micron distances”With 787 nm excitation, the spectrum contained an intense 904 nm line and satellites at 930 and 940 nm. The H2 system at 986 nm was stronger at the periphery. More than 40 narrow lines occupied the 800 to 900 nm region.
Lines recorded about 1 µm apart showed irregular changes in wavelength and relative intensity. At room temperature, the individual lines merged and broadened. Lowering the excitation power from 100% to 1% did not change the spectra.
This behavior was absent below the 521.9 nm Raman line under 488 nm excitation. The 488 nm spectra instead contained the H3 system, a 525 nm zero-phonon line, a broad band near 580 nm, and narrow lines at 689 and 698 nm.
Several additional lines at 630, 632, 638, and 685 nm sat on the broad band. The Y-rich zone also showed lines at 622 and 624 nm.
CL bands were too broad for unambiguous assignment. Large H3 and broad A-band contributions were considered plausible.
TEM reveals nanosized voids and dislocations
Section titled “TEM reveals nanosized voids and dislocations”TEM showed three defect classes: polyphase inclusions, dislocations in several configurations, and strongly flattened rhombic plates. Inclusion sizes ranged from tens to 300 nm. Some inclusion vertices generated dislocations or possible microcracks.
The rhombic objects reached 5 to 20 nm across and were only 1 to 3 nm thick. They lay in intersecting (111) planes. No detectable impurity enrichment and changes between imaging modes supported an interpretation as vacancy clusters, or nanovoids.
The plates occurred in both the Y-defect-rich intermediate zone and the periphery, although they appeared less common at the periphery. The periphery also contained weakly contrasted distorted hexagonal formations and rare quasihexagonal plates below 10 nm. The authors suggested that some may be graphitic flakes.
TEM also showed perfect and partial dislocations, loops, dipoles, and clusters of point defects. The clusters were linked to non-conservative dislocation motion and lacked obvious impurity concentrations, so they were presumed to contain vacancies and interstitials.
A possible formation route
Section titled “A possible formation route”The authors propose that dislocation dipoles annihilated and left planar vacancy clusters in (111). Subsequent consumption of vacancies produced during non-conservative dislocation motion enlarged the clusters.
The smallest separation within a dipole was about 11 nm. This is comparable to a cited experimental value of 7.4 ± 0.8 nm and a calculated critical separation of 9.9 ± 4.4 nm.
Alternative brittle-deformation mechanisms were considered. The absence of twinning evidence in electron diffraction and the lack of one common plane argue against Rose-channel formation, but the connection to slip planes remains possible.
The stabilization and growth mechanism remains unknown. The dislocation density was also markedly lower than in some brown diamonds with dislocation forests.
What the Y-center model still lacks
Section titled “What the Y-center model still lacks”No accepted structural model explains the Y-center. One proposal treats it as an N-C-N complex without interstitials. Other models include complexes involving neutral substitutional nitrogen and nitrogen or carbon interstitials. Possible oxygen association remains difficult to separate from Ni-related systems.
Only the 1358 cm⁻¹ peak is described as unambiguously correlated with the Y-center. A proposed 1387 cm⁻¹ assignment conflicts with annealing results and earlier infrared mapping. The 1332 cm⁻¹ feature also needs care because N⁺ may contribute.
The authors cite quantum-chemistry predictions for four interstitials with active modes at 1332, 1355, 1389, 1722, and 1741 cm⁻¹. Agreement with the spectrum is qualitative and does not establish the Y-center structure.
Limits of the evidence
Section titled “Limits of the evidence”The study examined one natural crystal, so it does not establish a universal defect structure or formation history. The paper reports difficulty correlating individual PL features with TEM objects.
Central-zone IR decomposition remained uncertain, and the A-center contribution could not be extracted reliably. PL spectra were not corrected for wavelength-dependent detector sensitivity. Broad CL bands prevented firm defect assignments.
The authors also did not determine whether hydrogen decorates the nanovoid walls. They present donor-acceptor recombination on those walls as a plausible explanation for the 800 to 900 nm lines. Void-size scatter is proposed as the source of line-energy differences.
The natural origin of U-14 matters for transfer to engineered diamond. The cited literature had not reported this absorption in synthetic diamonds.
Implications for material selection
Section titled “Implications for material selection”For natural type Ib-IaA diamond, characterize defect distributions by mapping rather than by single-point infrared spectra. The 2.7 versus 1.4 cm⁻¹ change at 1332 cm⁻¹ occurs across one crystal. Raman-frequency absorption alone cannot isolate the Y-center.
For near-infrared emitters, test spatial stability directly. The 800 to 900 nm lines change position and strength over about 1 µm and broaden at room temperature. The data do not support assuming fixed emission wavelengths or stable intensity.
Treat nanovoids as candidate recombination centers, not established ones. Any device program based on these lines still needs structural correlation and stability testing under its operating conditions.
The study also separates two material properties. Nitrogen aggregation governs the A and C centers, while the Y-rich zone has lower aggregated nitrogen and abundant vacancy-related damage. This zoning should be considered when selecting natural diamond for optical studies.
View Original Abstract
The paper presents results of investigation of a natural Ib-IaA diamond containing Y-defects from Yubileinaya kimberlite pipe, Yakutia. Analysis of spatial distribution of nitrogen-related A- and C-centers and intensity of infrared (IR) absorption at Raman frequency (1332 cm-1) reveals anticorrelation between these defects. Transmission electron microscopy of a zone with abundant Y-defects shows presence of dislocations in various configurations and numerous clusters of point defects generated by non-conservative dislocation movement. Extended defects with shape resembling thin (1-3 nm) rhombic plates with the largest dimension up to 5-20 nm are observed. Analysis of contrast of these defects shows that they represent nanosized voids (vacancy clusters). It is suggested that the defects were formed by annihilation of dislocation dipoles with subsequent growth by consumption of vacancies produced by non-conservative motion of dislocations. Upon excitation by 787 nm laser, numerous narrow photoluminescence lines are observed between 800-900 nm, their intensity and position show pronounced spatial heterogeneity on scale of few microns. Qualitatively similar behavior was earlier noted for hydrogenated nanodiamonds. It is suggested that the unusual behavior of the luminescence lines may be explained by recombination processes on internal walls of the discovered nanovoids.
High Concentration Quantum Grade Single Crystal Diamond
View Specs →
Single Crystal Cvd Diamond Plate
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →Original source
Section titled “Original source”References
Section titled “References”- 1994 - The relationship between infrared absorption and the A defect concentration in diamond [Crossref]
- 2025 - From mono- to hexa-interstitials: Computational Insights into carbon defects in diamond
- 1984 - One-phonon infrared absorption in diamond [Crossref]
- 1982 - Optical studies of vibronic bands in yellow luminescing natural diamonds [Crossref]
- 2025 - Infrared spectroscopy of natural Type Ib diamond: insights into the formation of Y-centers and the early aggregation of nitrogen [Crossref]
- 2025 - The first-formed hydrogen and nitrogen defects in natural diamond [Crossref]
- 2022 - Size-dependent thermal stability and optical properties of ultra-small nanodiamonds synthesized under high pressure [Crossref]
- 2010 - Spectroscopic and microscopic characterizations of color lamellae in natural pink diamonds [Crossref]
- 2013 - Extended defects in natural diamonds: an Atomic Force Microscopy investigation [Crossref]
- 2001 - Defect states of substitutional oxygen in diamond