Raman Spectroscopy and Bulk Modulus of 2–5 nm Nanodiamond: Evidence of Quantum Confinement Effects

Background

Over the past three decades, nanodiamond research has advanced considerably, yet the influence of quantum confinement on its mechanical properties and Raman response remains underexplored. Quantum confinement becomes significant when the crystal size approaches the exciton Bohr radius; for diamond this radius is 1.57 nm, making the 3 nm regime particularly relevant. Parallel electron energy‑loss spectroscopy (PEELS) and nuclear magnetic resonance (NMR) confirm that nanodiamond contains no sp² carbon; instead, quantum confinement increases the bandgap and discrete energy levels at the band edges, leading to higher elastic moduli. Prior work has reported a bulk modulus of ~500 GPa for nanodiamond based on pressure–volume data, while lattice parameters remain indistinguishable from natural diamond.

Raman studies have identified a downshift of the bulk diamond band from 1333 to 1325 cm⁻¹ due to phonon confinement. Additional features near 1250, 1590, 1640, and 1740 cm⁻¹ arise from surface groups and sp² carbon. However, the assignment of the 1325‑cm⁻¹ band to sp³ carbon and the 1600‑cm⁻¹ band to sp² carbon conflicts with resonant Raman cross‑sections, which are 50–200 times larger for sp² at visible wavelengths but equal at 257 nm. Our work demonstrates that the relative intensities of the 1325 and 1600‑cm⁻¹ bands are excitation‑wavelength independent in the 257–532 nm range, suggesting both arise from phonon features of nanodiamond. A third band shifts from 1500 to 1630 cm⁻¹ when the excitation wavelength changes from 458 to 257 nm. The bulk modulus, estimated at ~560 GPa, supports the presence of quantum confinement effects.

Methods

We employed detonation‑produced 2–5 nm nanodiamond from SINTA (Belarus). To remove surface contaminants, the particles were milled in a planetary ball‑mill with 25 wt % Si or NaCl, using ceramic silicon nitride bowls and 10 mm balls. This procedure yields homogeneous nanocomposites free from ball‑material contamination. For high‑pressure studies, a 25‑nm monocrystalline diamond powder from Microdiamant AG (HPHT synthesis) was used, prepared by drying the aqueous suspension.

Raman spectra were collected with a TRIAX 552 spectrometer (Jobin Yvon), equipped with a 2KBUV CCD detector and razor‑edge filters. TEM imaging was performed on a JEOL JEM‑2010, and X‑ray diffraction used a PANalytical Empyrean. A diamond anvil cell (DAC) provided pressures up to 53 GPa; pressure calibration relied on the Raman shift of the diamond anvil. XRD data were analyzed via the MAUD program using Rietveld refinement, yielding an average particle size of ~5 nm and a lattice parameter of 3.567 ± 0.002 Å—identical to natural diamond.

Results and Discussion

Raman spectra of the 2–5 nm nanodiamond, irrespective of whether the sample was a pure powder or mixed with Si or NaCl, consistently show three bands: 1325 cm⁻¹ (with a shoulder at ~1250 cm⁻¹), 1600 cm⁻¹, and a band that shifts from 1500 to 1630 cm⁻¹ when the excitation wavelength changes from 458 to 257 nm. The 1600‑cm⁻¹ band shows no resonance effect—the intensity remains constant across 458‑ and 257‑nm excitation—indicating it originates from phonon modes of the nanodiamond rather than sp² carbon.

Increasing the laser power from 0.7 mW to 7 mW in the Si‑contaminated sample triggers the formation of SiC and sp² carbon clusters, as evidenced by the appearance of SiC bands near 790 cm⁻¹ and a ~50‑fold increase in the Raman cross‑section of sp² features. This demonstrates that the planetary milling effectively removes surface‑bound functional groups while any residual contamination remains trapped in the Si or NaCl matrix.

Under hydrostatic pressure up to 50 GPa, the 1600‑cm⁻¹ band’s half‑width and intensity remain unchanged, in stark contrast to the pressure‑induced broadening of the G band in graphite or amorphous carbon. The 1325‑cm⁻¹ band disappears under 50 GPa, implying a bulk modulus exceeding 524 GPa. A least‑squares fit of the pressure dependence of the 1600‑cm⁻¹ band yields a bulk modulus of 564 GPa for the 2–5 nm nanodiamond.

To test the size dependence, we examined 25‑nm nanodiamond under the same pressure conditions. Its Raman band at 1329 cm⁻¹ shifts to 1483 cm⁻¹ at 50 GPa, matching the expected behavior of bulk diamond with a 443 GPa bulk modulus. A ~1580‑cm⁻¹ band, characteristic of sp² carbon, vanishes under pressure and exhibits the expected 50–100‑fold intensity drop when the excitation wavelength shifts from 532/458 to 257 nm. These observations confirm that the extraordinary properties observed in the 2–5 nm range arise from quantum confinement and disappear for particles larger than ~10 nm.

Conclusions

Nanodiamond in the 2–5 nm size interval displays three Raman features at 1325, 1500–1630, and 1600 cm⁻¹. The 1600‑cm⁻¹ band is unequivocally linked to phonon modes of the nanodiamond, not to sp² carbon, as evidenced by its excitation‑wavelength independence and pressure resilience. The presence of higher‑frequency bands relative to bulk diamond indicates an elevated elastic modulus, corroborated by a bulk modulus of 564 GPa derived from high‑pressure Raman data. These findings highlight the significant role of quantum confinement in enhancing the mechanical properties of sub‑10 nm diamond nanocrystals.