Quantifying Spin Defect Density in hBN via Raman and Photoluminescence Analysis

Negatively charged boron vacancies ($\mathrm{V_B^-}$) in hexagonal boron nitride (hBN) are emerging as promising solid-state spin qubits due to their optical accessibility, structural simplicity, and compatibility with photonic platforms. However, quantifying the density of such defects in thin hBN flakes has remained elusive, limiting progress in device integration and reproducibility. Here, we present an all-optical method to quantify $\mathrm{V_B^-}$ defect density in hBN by correlating Raman and photoluminescence (PL) signatures with irradiation fluence. We identify two defect-induced Raman modes, D1 and D2, and assign them to vibrational modes of $\mathrm{V_B^-}$ using polarization-resolved Raman measurements and density functional theory (DFT) calculations. By adapting a numerical model originally developed for graphene, we establish an empirical relationship linking Raman (D1, $E_\mathrm{2g}$) and PL intensities to absolute defect densities. This method is universally applicable across various irradiation types and uniquely suited for thin flakes, where conventional techniques fail. Our approach enables accurate, direct, and non-destructive quantification of spin defect densities down to $10^{15}$ defects/ cm${}^3$, offering a powerful tool for optimizing and benchmarking hBN for quantum optical applications.

💡 Research Summary

The manuscript presents a fully optical methodology for quantifying the density of negatively charged boron‑vacancy (V_B⁻) spin defects in thin hexagonal boron nitride (hBN) flakes. The authors generate V_B⁻ centers by focused nitrogen ion‑beam irradiation across a patterned array of 4 µm × 4 µm tiles with systematically varied fluences. Monte‑Carlo SRIM simulations provide a first‑order estimate of the vacancy distribution, revealing that the defects are confined within ~100 nm of the surface and are roughly equally split between boron and nitrogen vacancies. The authors acknowledge that SRIM does not capture defect annealing, charge state dynamics, or interstitial formation, so experimental calibration is required for absolute density determination.

Raman spectroscopy, performed with excitation energies of 1.96 eV (633 nm), 2.33 eV (532 nm) and 2.62 eV (473 nm), reveals three characteristic peaks: the intrinsic E₂g mode at ~1365 cm⁻¹, and two defect‑related modes labeled D1 (~1290 cm⁻¹) and D2 (~450 cm⁻¹). The D1 and D2 intensities increase monotonically with ion fluence, while the E₂g intensity remains essentially constant. Polarization‑resolved Raman measurements show that D1 is isotropic with respect to the incident linear polarization, whereas D2 exhibits a pronounced cos 2θ dependence. High‑resolution spectra resolve D2 into two sub‑components, D2a (≈459 cm⁻¹) and D2b (≈352 cm⁻¹), which appear only under parallel (∥) polarization; D2b disappears under perpendicular (⊥) polarization, confirming the symmetry‑derived selection rules. The authors interpret D2 as arising from out‑of‑plane vibrational motions localized around the V_B⁻ defect, while D1 likely corresponds to in‑plane defect‑induced vibrations.

Density functional theory (DFT) calculations of the vibrational spectrum of a V_B⁻ defect in hBN predict localized modes at frequencies matching the experimental D1 and D2 peaks, providing a solid theoretical assignment. The photoluminescence (PL) spectra display a broad near‑infrared band centered at 1.53 eV (≈810 nm), which is independent of the boron isotope composition (¹⁰B vs. ¹¹B) and scales with defect density, consistent with previous reports of V_B⁻ optical emission.

To translate Raman and PL intensities into absolute defect densities, the authors adapt the Tuinstra‑Koenig model originally developed for graphene. In graphene, the ratio I(D)/I(G) is inversely proportional to the crystallite size L_a, which is directly related to defect spacing. By analogy, they propose an empirical relationship I(D1)/I(E₂g) = C · L_d⁻¹, where L_d is the average distance between V_B⁻ defects and C is a calibration constant determined from the set of irradiated tiles. Combining this with the SRIM‑derived vacancy fluence‑to‑density conversion yields an absolute density scale that is accurate down to ~10¹⁵ cm⁻³. The method is validated by showing a linear correlation between the Raman intensity ratio, the integrated PL intensity, and the simulated vacancy concentration across the tile series.

Key advantages of the approach are its non‑destructive nature, applicability to ultrathin flakes where electron‑microscopy‑based techniques fail, and its universality across different irradiation sources (ions, electrons, neutrons). The technique requires only a standard confocal Raman/PL microscope, making it readily adoptable in most labs. Limitations include reduced signal‑to‑noise at very low fluences, the need for experimental calibration because SRIM does not account for defect recombination or charge‑state evolution, and saturation of the Raman defect peaks at very high defect densities (>10¹⁸ cm⁻³), where the linear model breaks down.

The authors suggest several future directions: (i) low‑temperature or high‑pressure annealing to improve defect ordering and Raman peak sharpness; (ii) time‑resolved Raman and PL to monitor defect creation and annealing dynamics in situ; (iii) extension of the Raman‑PL quantification framework to other two‑dimensional materials hosting spin defects, such as transition‑metal dichalcogenides. Overall, the presented all‑optical quantification protocol provides a practical, scalable tool for benchmarking and optimizing hBN‑based quantum photonic devices, facilitating reproducible fabrication of spin‑defect ensembles for quantum sensing and information processing applications.