betaFeedback

WorksPhysics and Astronomy

Enhanced Emission from Boron-Vacancy Center in Rhombohedral Boron Nitride

Abstract

Boron nitride is a layered crystal whose properties depend on how its atomic sheets are stacked. Its negatively charged boron vacancy is a well-established magnetic defect that can be prepared and read out optically, but in the common hexagonal form it emits very little light, because the symmetry of the surrounding lattice forbids the relevant optical transition. Here we show, using first-principles calculations, that stacking the sheets in the rhombohedral sequence instead removes this restriction and increases the emitted intensity by one to two orders of magnitude, while the magnetic properties remain comparable or improve. We predict that the resulting emission is bright enough for a single defect to be addressed at room temperature, and that a sharp emission line, absent in the hexagonal form, should appear on cooling. Stacking order therefore acts as a design parameter for tailoring the quantum properties of defects embedded in layered materials.

Introduction

Among two-dimensional (2D) semiconductors, boron nitride (BN) stands out for its wide band gap, excellent physicochemical stability, and mechanical robustness. In layered sp2sp^2 BN, two stable polytypes are commonly encountered [1, 2]: hexagonal boron nitride (hBN) with the AA′AA' stacking sequence and rhombohedral boron nitride (rBN) with ABC stacking, which has only recently been demonstrated experimentally [3, 4, 5]. The stacking sequence strongly influences the optical response and fine structure of BN crystals [2, 5]. Light–matter interaction in monolayer BN is intrinsically weak because of the limited interaction length of a single atomic layer, but it strengthens with increasing layer number [4]. In hBN, inversion symmetry suppresses nonlinear optical interactions between adjacent layers, whereas rBN lacks inversion symmetry and thus naturally supports coherent enhancement of nonlinear optical signals. Accordingly, rBN exhibits distinct advantages, including strong optical nonlinearity, interfacial ferroelectricity, and polarized second-harmonic generation, all arising from its broken centrosymmetry [4, 6]. This symmetry breaking may also enable additional functionalities beyond nonlinear optics.

In the context of quantum technologies, the wide band gaps of 2D BN polytypes make them attractive hosts for color centers. Isolated quantum emitters were initially reported in hBN in the visible [7, 8, 9] and ultraviolet (UV) [10] spectral regions. More recently, it was shown that ABC stacking in rBN can substantially modify the photoluminescence (PL) spectra of UV emitters, and that their spectroscopic fingerprints can be used to identify the rBN host polytype [5]. However, coherent spin control of these UV emitters has not yet been reported, limiting their prospects as qubits for quantum sensing and quantum communication. Here we take a major step toward this goal and show that the lack of inversion symmetry in rBN can be exploited to markedly improve quantum control of embedded, optically addressable defect spins.

Optical initialization and readout of single spins have been reported for several visible quantum emitters in hBN [11, 12, 13, 14, 15, 16]. However, the microscopic origin of these qubits remains under active investigation. Recent studies suggest that donor–acceptor pairs may underlie some of these spin-active emitters, with random separations between constituent species [17, 18, 19, 20], which makes deterministic defect engineering challenging. In contrast, the negatively charged boron vacancy (VB−V_{\mathrm{B}}^-) is a well-established optically addressable defect-spin system in hBN [21, 22], and its concentration can be engineered in a controlled manner [23]. At the same time, coherent control has so far been limited largely to ensembles, because VB−V_{\mathrm{B}}^- is intrinsically dim and has not been substantially brightened in optical cavities [24]. Its extremely low quantum efficiency (<0.1%< 0.1\%) originates from the dipole-forbidden nature of the transition between the lowest triplet excited state and the ground state [25, 26, 27]. Indeed, a recent multireference study of VB−V_{\mathrm{B}}^- in hBN places the radiative rate of the lowest triplet excited state at only 10.1 kHz (τrad=98.8 μs\tau_{\mathrm{rad}} = 98.8\ \mu\mathrm{s}), whereas the competing spin-selective intersystem crossing (ISC) to the singlet shelving manifold proceeds on a 3.2 ns timescale [28], i.e., the optical cycle is overwhelmingly nonradiative.

In this study, we analyze the photophysics of VB−V_{\mathrm{B}}^- in 2D BN and show that the lack of inversion symmetry in rBN can enhance its emission intensity relative to hBN. Our ab initio results predict a one- to two-orders-of-magnitude increase in the radiative decay rate, with coherent zero-phonon-line emission expected to be observed. servable at cryogenic temperatures in the near-infrared. The simulated room-temperature fluorescence spectrum and the ground-state zero-field-splitting tensor reinforce the tentative assignment of a near-infrared color center to VB−V_{\mathrm{B}}^{-} in rBN [29]. Combined with our previous finding of a long spin–lattice relaxation time T1T_1 [30], these results establish VB−V_{\mathrm{B}}^{-} in rBN as a strong candidate for single-spin quantum sensing. More broadly, our work leverages symmetry engineering to turn a technologically mature defect into a viable qubit platform in rBN that is compatible with cavities, resonators, and nanophotonic components, thereby advancing quantum technologies based on two-dimensional materials.

Results

Symmetry of the boron vacancy in the two polytypes

We demonstrate the concept of tailoring the properties of fluorescent defect spins by embedding them in non-centrosymmetric rBN, using VB−V_{\mathrm{B}}^{-} as a prototypical example. VB−V_{\mathrm{B}}^{-} has a well-defined structure within a single BN honeycomb layer, as shown in Figure 1a. In bulk hBN, the AA′ stacking enforces D3hD_{3h} symmetry for VB−V_{\mathrm{B}}^{-}, where the defective layer acts as a mirror plane that effectively plays the role of inversion symmetry [21, 26]. Previous theoretical studies [25, 26] established that the ground state is the 3A2′{}^{3}A_{2}^{\prime} triplet, while the lowest triplet excited states have 3A1′′{}^{3}A_{1}^{\prime\prime} and 3E′′{}^{3}E^{\prime\prime} character. Here, the Γ′\Gamma^{\prime} and Γ′′\Gamma^{\prime\prime} labels denote states that are even and odd with respect to the mirror plane, respectively, in D3hD_{3h} symmetry.

Electronic structure of the negatively charged boron vacancy in rhombohedral boron nitride

Figure 1. Electronic structure of the negatively charged boron vacancy in rhombohedral boron nitride. a Geometry of the negatively charged boron vacancy with the calculated spin density (yellow isosurface). Nitrogen, boron and the vacant boron site are indicated in the panel. b Single-particle Kohn–Sham electronic structure obtained with the Heyd–Scuseria–Ernzerhof (HSE) hybrid functional within density functional theory (DFT). Arrows denote the two spin channels; a1a_1 and exe_x, eye_y label the fully symmetric and the doubly degenerate defect orbitals, and a star marks the antibonding partners. c Many-body electronic structure from many-body perturbation theory within the GW approximation combined with the Bethe–Salpeter equation (GW+BSE). The composition of the exciton wavefunction in terms of single-particle hole–electron pairs is indicated beside the symmetry label of each multiplet excited state. The radiative lifetimes τrad\tau_{\mathrm{rad}}, in microseconds, are derived from the Bethe–Salpeter spectrum. d Vertical optical excitation spectrum, with the lowest excitonic peak at 1.72 eV. Red and blue denote photon polarization perpendicular and parallel, respectively, to the threefold rotation axis of the defect. The optical transition dipole moments are given in Debye. e Isosurface visualization of the Kohn–Sham orbitals of panel b, in top and side view. The side views show that the mirror plane of the boron nitride sheet is broken, which is what makes the near-infrared transition allowed.

Because electric-dipole transitions between Γ′\Gamma^{\prime} and Γ′′\Gamma^{\prime\prime} are forbidden in D3hD_{3h} symmetry, the corresponding emission is intrinsically weak and becomes allowed only through vibronic coupling [27]. Consistent with coupled phonon–exciton many-body perturbation theory, emission is activated by phonons and coherent zero-phonon-line (ZPL) emission is therefore not expected for VB−V_{\mathrm{B}}^{-} in hBN [27]. The phonon modes responsible for activating emission involve out-of-plane displacements of atoms surrounding VB−V_{\mathrm{B}}^{-}, i.e., they break the mirror symmetry [27]. This can be rationalized by group theory: out-of-plane vibrations transform as Γph′′\Gamma_{\mathrm{ph}}^{\prime\prime} in D3hD_{3h}, so the vibronic excited state transforms as Γel′′⊗Γph′′=Γvibronic′\Gamma_{\mathrm{el}}^{\prime\prime} \otimes\Gamma_{\mathrm{ph}}^{\prime\prime} = \Gamma_{\mathrm{vibronic}}^{\prime}, which renders the transition to the Γ′\Gamma^{\prime} ground state allowed (a Herzberg–Teller mechanism).

A higher-lying 3E′{}^{3}E^{\prime} triplet excited state resides in the visible spectral range and is dipole allowed. This explains why VB−V_{\mathrm{B}}^{-} in hBN can be photoexcited with green light and, following rapid relaxation into the lowest triplet excited states (3A1′′{}^{3}A_{1}^{\prime\prime} and 3E′′{}^{3}E^{\prime\prime}), exhibits weak near-infrared emission via phonon-assisted decay [22, 25, 26]. This analysis suggests that breaking the mirror symmetry should enhance optical emission.

Because the ABC stacking of rBN removes the mirror symmetry (Figure 1e), we test this principle for VB−V_{\mathrm{B}}^{-} in rBN.

The defect symmetry is reduced to C3vC_{3v}, such that the prime/double-prime parity labels no longer apply and the associated selection rule is relaxed. We therefore proceed to compute the optical and spin properties of VB−V_{\mathrm{B}}^{-} in rBN.

Optical properties

For VB−V_{\mathrm{B}}^{-} in rBN, we find several occupied defect levels of fully symmetric a1a_1 and doubly degenerate ee symmetry. In our calculations with the Heyd–Scuseria–Ernzerhof (HSE) hybrid functional within density functional theory (DFT), the highest occupied ee level is filled by two electrons, yielding the mS=+1m_S = +1 component of an S=1S = 1 state. Because two a1a_1 and two ee orbitals appear in the band gap, we label them by their energetic ordering in the spin-majority channel; spin polarization can reorder these levels in the spin-minority channel. All of these orbitals are localized on the three nitrogen dangling bonds adjacent to the vacancy.

Previous work on VB−V_{\mathrm{B}}^{-} in hBN [26] showed that several vacancy-localized orbitals with similar energies contribute to strongly correlated many-body excited states. Motivated by this, we investigate the excited-state manifold using many-body perturbation theory within the GW plus Bethe–Salpeter equation (GW+BSE) framework (Methods), which captures correlated excitations as superpositions of single-particle transitions, an essential aspect for VB−V_{\mathrm{B}}^{-} in two-dimensional BN [26]. The GW+BSE calculations resolve the low-energy intra-defect excitations and provide the corresponding excitation energies. They also enable analysis of the character of each excited state via its excitonic composition (Figure 1c) and yield the optical transition strengths from the ground state.

Throughout these calculations, the atomic geometry is fixed at the DFT-optimized ground-state structure; the resulting spectrum therefore corresponds to vertical excitations energies and neglects ionic relaxation upon photoexcitation (Figure 1d). Consequently, the computed spectrum is not intended for direct quantitative comparison with experiment; instead, it provides a reliable map of the optically accessible transition pathways between the ground state and the relevant excited states.

We find that the lowest excited state can be described as a combination of a1→e∗a_1 \rightarrow e^{*} and e→e∗e \rightarrow e^{*} hole–electron pairs which transforms as 3E{}^{3}E, see Figure 1c for visualization. In stark contrast to VB−V_{\mathrm{B}}^{-} in hBN with D3hD_{3h} symmetry, this lowest excited state is optically allowed from the ground state with perpendicular polarization of photons in the optical transition with C3vC_{3v} symmetry. The excitation wavelength of the lowest excited state falls in the near-infrared (NIR) region. As can be seen in the side views of Figure 1e, the planar symmetry is broken, which is responsible for the observable optical transition in the NIR region. We identify another optically allowed state with parallel polarization of photons also in the NIR region, and finally another state in the visible wavelength region (green color) with perpendicular polarization of photons which has the strongest optical transition dipole moment (see Figure 1d). The last optical transition also occurs in hBN for VB−V_{\mathrm{B}}^{-} and can be used to efficiently excite the defect in hBN. Our calculations imply that VB−V_{\mathrm{B}}^{-} in rBN can be excited in the NIR region too.

Next we focus on the lowest energy triplet excited state which provides luminescence upon appropriate illumination. According to GW+BSE results, the state has 3E^{3}E symmetry which is subject to Jahn–Teller distortion. We apply the constrained-occupation (Δ\DeltaSCF) method within HSE DFT to compute the adiabatic potential energy surface of this excited state. To this end, we promoted an electron from the a1a_{1} level to the e∗e^{*} level in the spin-minority channel. After the self-consistent field procedure in the C3vC_{3v} symmetry, the resulting hole orbital has broken symmetry and shows a mixture of a1a_{1} and exe_{x} characters, which implies a multiplet exciton excited state—similar to the GW+BSE result.

We then allowed the ions to relax to a lower symmetry, which indeed occurs due to the e⊗Ee \otimes E Jahn–Teller (JT) effect [31]. Our simulations reveal substantial JT distortion that lowers the total energy by 0.21 eV in the 3E^{3}E state. The barrier between different JT distorted configurations is 159 meV.

The phonon sideband in the luminescence spectrum is determined within the Huang–Rhys (HR) theory [32, 33, 34, 35, 36, 37, 38] (see the original theory in Ref. 32 and our implementation in Ref. 39 that is based on Ref. 35). We calculated the PL spectra of VB−_{\mathrm{B}}^{-} in rBN at 4 K and 300 K (Figure 2a). The defect shows strong room-temperature PL centered at λ≈734\lambda\approx734 nm (1.69 eV). The computed HR factor is Stot=3.64S^{\mathrm{tot}} = 3.64, which results in a relatively broad PL spectrum, but the coherent ZPL peak becomes visible at low temperatures. We note that the sideband originates predominantly from the JT-active modes transforming according to the EE representation of the C3vC_{3v} group, as indicated by the decomposition of the total HR factor Stot=∑iSiS^{\mathrm{tot}} = \sum_i S_i into EE (JT-active) and A1A_{1} (totally symmetric) contributions, with SA1tot=0.52S_{A_{1}}^{\mathrm{tot}} = 0.52 and SEtot=3.12S_{E}^{\mathrm{tot}} = 3.12, respectively (see Supplementary Figure 1 in Supplementary Note 1 for details). The resulting Debye–Waller factor, fDW=e−Stot=2.6%f_{\mathrm{DW}} = e^{-S^{\mathrm{tot}}} = 2.6\%, is about six times larger than the 0.4% reported for VB−_{\mathrm{B}}^{-} in hBN [28], for which the ZPL has never been unambiguously resolved. The computed PL spectrum agrees well with an observed one at room temperature [29] that was associated with VB−_{\mathrm{B}}^{-} in rBN.

Figure 2. Photoluminescence and electron–phonon coupling of the negatively charged boron vacancy in rhombohedral boron nitride. a Photoluminescence (PL) spectrum simulated with the Heyd–Scuseria–Ernzerhof hybrid functional within density functional theory, at 4 K (purple) and 300 K (green). The zero-phonon line (ZPL) and a structured phonon sideband emerge at low temperature. The computed ZPL energy could have an inaccuracy of about 0.1 eV, so the computed spectrum may be shifted when directly compared to experimental data. b Electron–phonon coupling underlying the sideband in panel a. Vertical bars are the partial Huang–Rhys (HR) factors SiS_i of the individual vibrational modes (left axis), resolved into totally symmetric A1A_1 modes (blue) and Jahn–Teller-active EE modes (red); the solid curves are the same data broadened by Gaussians for visibility. The HR factors are dimensionless by definition, Si=ωiqi2/2ℏS_i = \omega_i q_i^2/2\hbar, where ωi\omega_i is the frequency and qiq_i the mass-weighted displacement of mode ii, hence the left axis carries no unit. The dotted black curve is the multiphonon spectral function S(ℏω)S(\hbar\omega) of the sideband (right axis, in eV−1^{-1}), plotted against the phonon energy ℏω\hbar\omega (bottom axis, in eV). The two dominant EE modes, at 24 and 39 meV, are indicated by arrows. The symmetry-resolved sums given in the legend are SA1tot=∑i∈A1Si=0.52S_{A_1}^{\mathrm{tot}} = \sum_{i\in A_1} S_i = 0.52 and SEtot=∑i∈ESi=3.12S_E^{\mathrm{tot}} = \sum_{i\in E} S_i = 3.12, giving the total Stot=SA1tot+SEtot=3.64S^{\mathrm{tot}} = S_{A_1}^{\mathrm{tot}} + S_E^{\mathrm{tot}} = 3.64.

The corresponding radiative lifetime (τrad\tau_{\mathrm{rad}}) is 1.4 μ\mus, i.e., on the order of a microsecond, estimated as

Γrad=1τrad=nDEZPL3μ23πε0c3,(1)\Gamma_{\mathrm{rad}} = \frac{1}{\tau_{\mathrm{rad}}} = \frac{n_{\mathrm{D}} E_{\mathrm{ZPL}}^{3} \mu^{2}}{3\pi\varepsilon_{0} c^{3}}, \tag*{(1)}

where ε0\varepsilon_{0} is the vacuum permittivity, ℏ\hbar is the reduced Planck constant, cc is the speed of light, nD=2.1n_{\mathrm{D}} = 2.1 is the refractive index of rBN at the ZPL energy EZPLE_{\mathrm{ZPL}}, and μ=0.64\mu= 0.64 Debye is the optical transition dipole moment of the given transition at the GW+BSE level of theory (Figure 1d). In hBN, the corresponding emission is symmetry-forbidden at the level of the static D3hD_{3h} structure and is rendered only partially allowed by planar-symmetry-breaking phonons and by the JT distortion, whereas the zero-phonon coherent emission remains symmetry-forbidden [27]. The theoretical estimates of the hBN radiative lifetime accordingly span 11–

99 μ\mus, from ∼11 μ\sim11~\mus used to model time-resolved optically detected magnetic resonance (ODMR) data [40], through 20 μ\mus including Herzberg–Teller intensity borrowing [25], to the multireference value τrad=98.8 μ\tau_{\mathrm{rad}} = 98.8~\mus (an upper limit, since electron–phonon coupling shortens it) [28]. In contrast, the lower C3vC_{3v} symmetry of rBN allows a direct emission with a radiative lifetime that is shorter by one to two orders of magnitude. This is the most important finding of our study, which turns VB−_{\mathrm{B}}^{-} addressable at the single defect level in rBN.

Nonradiative decay and quantum efficiency

The brightness of a color center is set by the branching ratio between radiative and nonradiative decay, η=Γrad/(Γrad+Γnr)\eta= \Gamma_{\mathrm{rad}}/(\Gamma_{\mathrm{rad}} + \Gamma_{\mathrm{nr}}), so the enhancement discussed above is only meaningful if Γnr\Gamma_{\mathrm{nr}} is not enhanced along with it. Two nonradiative channels compete with photon emission from the lowest triplet excited state: (i) direct multiphonon internal conversion to the 3A2^{3}A_{2} ground state, and (ii) spin-selective ISC into the singlet shelving manifold. Channel (i) is negligible in both polytypes: bridging the ≈1.7\approx1.7–1.9 eV electronic gap requires the simultaneous emission of ≳40\gtrsim40 phonons of the dominant 24–39 meV modes, and the corresponding rate is exponentially suppressed [28]. Channel (ii) therefore dominates, with ΓISC≈(3.2 ns)−1\Gamma_{\mathrm{ISC}} \approx(3.2~\mathrm{ns})^{-1} in hBN [28], consistent with the measured excited-state decay rates of 1–2 ns−1^{-1} [40].

Crucially, the ISC channel is not symmetry-forbidden in D3hD_{3h} hBN—it is already fully allowed and fast—so removing the mirror plane cannot switch it on. Its rate is governed by the spin–orbit matrix elements between the triplet excited state and the singlets, by the singlet–triplet energy gap, and by the vibrational overlap, all of which are properties of the three nitrogen dangling bonds within a single BN sheet. The ABC versus AA′ registry perturbs these intra-layer quantities only weakly, as evidenced by the near-identical ground-state zero-field splitting (D=3.44D = 3.44 GHz here versus 3.45 GHz measured in both polytypes [22, 29]), by the nearly unchanged PL band position, and by the fact that room-temperature ODMR—a signal that exists only by virtue of the spin-selective ISC—is readily observed for VB−_{\mathrm{B}}^{-} in rBN under essentially the same conditions as in hBN [29]. Adopting Γnr≃ΓISC\Gamma_{\mathrm{nr}} \simeq\Gamma_{\mathrm{ISC}} unchanged between the two polytypes, we obtain ηhBN≈3×10−5\eta_{\mathrm{hBN}} \approx3 \times10^{-5} and ηrBN≈2×10−3\eta_{\mathrm{rBN}} \approx2 \times10^{-3}; that is, the quantum efficiency inherits essentially the full radiative enhancement. We emphasize that this estimate is deliberately conservative: it assumes the least favorable case in which none of the nonradiative rate is reduced in rBN. A quantitative ab initio treatment of the rBN ISC rates would require a multireference description of the quasi-degenerate singlet–triplet manifold beyond the scope of the present work [28], but any such refinement would have to make the ISC in rBN more than two orders of magnitude faster than in hBN to offset the radiative gain, which we consider implausible given the weakness of the interlayer perturbation.

Spin properties

The 3A2^3A_2 ground state’s spin density is localized on the three nitrogen dangling bonds (see Supplementary Figure 2). The respective hyperfine constants are listed in Supplementary Tables 1 and 2. The C3vC_{3v} crystal field splits the mS=0m_S = 0 and mS=±1m_S = \pm1 manifolds of the triplet by dipolar spin–spin interactions. The calculated zero-field splitting (ZFS) parameter is D=3.44D = 3.44 GHz, which agrees well with that of a recently observed ODMR center in rBN (D≈3.45D \approx3.45 GHz at room temperature) that was associated with VB−V_\mathrm{B}^{-} [29]. We simulated the respective continuous-wave (cw) ODMR spectrum using the calculated hyperfine tensors, which we show at zero magnetic field and at 9 mT in Figure 3 (see Supplementary Note 4 for details about simulation parameters). Both transitions exhibit resolved hyperfine splitting into seven lines, originating from the nuclear spins (I=1I = 1) of three equivalent 14N^{14}\mathrm{N} atoms adjacent to the vacancy in the rBN plane. The nearest-neighbor 14N^{14}\mathrm{N} nuclei yield a splitting of Azz≈48.3A_{zz} \approx48.3 MHz in the ground state. The hyperfine broadening of the simulated cw-ODMR spectrum and the broadening in the observed cw-ODMR spectrum agree remarkably well, which reaffirms the identification of the ODMR center in rBN.

Figure 3. Simulated magnetic resonance spectrum of the negatively charged boron vacancy in rhombohedral boron nitride. Continuous-wave optically detected magnetic resonance (ODMR) spectrum simulated at a magnetic field of 9 mT (blue), computed from the ab initio zero-field-splitting and hyperfine tensors reported in this work; the simulation parameters are given in Supplementary Note 4. In-set: the corresponding spectrum at zero magnetic field (magenta), plotted on the same microwave-frequency axis. Both traces show the seven-line hyperfine pattern arising from the three equivalent nearest-neighbor nitrogen-14 nuclei, each of nuclear spin I=1I = 1, that surround the vacancy. The ODMR contrast is given in arbitrary units.

We note that the reduced symmetry may open additional relaxation channels when compared to those for VB−V_\mathrm{B}^{-} in hBN. The calculated spin–lattice relaxation time for VB−V_\mathrm{B}^{-} in rBN is T1≈25 μsT_1 \approx25\ \mu\mathrm{s} at room temperature, which is analyzed in detail in our previous study [30]. This indicates that VB−V_\mathrm{B}^{-} in rBN can be applied as a room-temperature qubit.

We further note that by solving the e⊗Ee \otimes E JT Hamiltonian [39] for the 3E^3E excited state with the use of our developed code [41], we obtained a Ham reduction factor of 0.006, implying strong quenching of spin–orbit coupling and off-diagonal ZFS tensor components [42] due to electron–phonon interactions. As a consequence, the usual D=3/2DzzD = 3/2D_{zz} component yields the energy splitting between mS=0m_S = 0 and mS=±1m_S = \pm1 levels in the 3E^3E excited state. On the other hand, the computation of the ZFS tensor is not straightforward with DFT due to the multi-determinant nature of the excited state (see Supplementary Note 3 for details). Nevertheless, we roughly estimate the DD constant of the lowest energy 3E^3E state to be approximately Dtheoryex.≈1.5D_{\mathrm{theory}}^{\mathrm{ex.}} \approx1.5 GHz. We note that this ab initioab\ initio value is relatively close to the experimentally observed 2.1 GHz associated with the ZFS of the excited state of VB−V_\mathrm{B}^{-} in hBN [40, 43, 44, 45, 46]. Given the similarity of the ground-state ZFS of VB−V_\mathrm{B}^{-} in hBN and rBN, our results suggest that the excited-state ZFS of VB−V_\mathrm{B}^{-} in rBN should also be around 2.1 GHz.

Discussion

In this work, we have theoretically demonstrated that the rhombohedral polytype of boron nitride provides an environment in which the optical and spin properties of the negatively charged boron-vacancy center are markedly enhanced. By combining density functional theory and many-body perturbation theory, we revealed that symmetry reduction in rBN enables optically allowed transitions with radiative lifetimes on the order of microseconds and photoluminescence that makes the coherent emission visible at low temperatures. The predicted zero-field splitting and hyperfine structure are in excellent agreement with experimental observations, supporting the reliability of our approach.

Our results establish rBN as a promising host material for solid-state spin defects, overcoming the low brightness that has limited the usability of VB−V_{\mathrm{B}}^{-} centers in hBN. This finding introduces a materials design pathway where layer stacking is not merely a structural detail but a powerful degree of freedom to engineer the quantum properties of embedded defects. Such control opens opportunities for realizing room-temperature single-spin quantum sensors, scalable photonic devices, and hybrid van der Waals quantum technologies.

This enhancement should be kept in perspective, and it has not yet been tested experimentally: the only existing measurement on VB−V_{\mathrm{B}}^{-} in rBN was not designed as a photometric comparison and reports no defect-density-normalized count rates [29]. A microsecond radiative lifetime is still three orders of magnitude longer than the nanosecond lifetimes of the bright visible and UV single-photon emitters of BN [7, 10], so VB−V_{\mathrm{B}}^{-} in rBN remains a comparatively dim emitter. What changes is that it crosses the threshold of single-defect detectability. With the optical cycle limited by the metastable singlet (τcycle≈0.04\tau_{\mathrm{cycle}} \approx0.04–0.34 μs0.34\ \mu\mathrm{s} [28]) and ηrBN≈2×10−3\eta_{\mathrm{rBN}} \approx2 \times10^{-3}, the saturated photon emission rate is ∼104 s−1\sim10^{4}\ \mathrm{s}^{-1}; assuming a few per cent collection efficiency for a high-numerical-aperture objective, ∼50%\sim50\% optical throughput and ∼60%\sim60\% detector quantum efficiency in the near infrared, this corresponds to ∼102\sim10^{2}–10310^{3} counts s−1\mathrm{s}^{-1}, which is sufficient for photon-correlation measurements on a single center. The same estimate for hBN gives only ∼1\sim1–1010 counts s−1\mathrm{s}^{-1}, at or below typical background, consistent with the fact that single VB−V_{\mathrm{B}}^{-} centers have never been isolated optically in that polytype [29].

The second consequence is spectroscopic. Because the ZPL is no longer symmetry-forbidden and the Debye–Waller factor rises from ≈0.4%\approx0.4\% to ≈2.6%\approx2.6\%, a sharp line should emerge near 1.69 eV upon cooling—a qualitative signature insensitive to defect density and collection efficiency, and thus the most direct test of our prediction. Once the ZPL can be identified, it can be selectively enhanced in an optical cavity [24], where the Purcell effect acts on the coherent line alone and redistributes emission from the phonon sideband into it. This is the route by which stacking engineering turns a technologically mature but dim defect into a usable single-spin–photon interface. More broadly, our study underscores the potential of stacking engineering in two-dimensional materials as a general strategy for tailoring defect-based quantum systems.

Methods

We employed two complementary approaches—density functional theory (DFT) and many-body perturbation theory (MBPT) within the GW approximation—to describe the electronic structure of the VB−V_{\mathrm{B}}^{-} defect at different levels of theory. For the DFT calculations, a plane-wave basis set with a cutoff of 450 eV and projector augmented-wave (PAW) atomic potentials [47] were used as implemented in VASP version 5.4.1 [48, 49]. The convergence criteria for total energy and atomic forces were set to 10−410^{-4} eV and 0.01 eV/Å, respectively. Structural relaxations, excited-state calculations within constrained occupation DFT (or ΔSCF method) [50], hyperfine tensors [51], and spin–spin zero-field splitting (ZFS) tensors [52] as implemented by Martijn Marsman were obtained using the Heyd–Scuseria–Ernzerhof (HSE) hybrid functional [53] with a 0.32 fraction of exact exchange and a screening parameter of 0.2 Å−1^{-1}. To eliminate spin contamination in the spin–spin ZFS tensor, we applied the correction scheme of Ref. 54. The supercell model consisted of a 9×9×19 \times9 \times1 486-atom rBN supercell embedding a single negatively charged boron vacancy. The optimized interlayer distance was 3.31 Å, obtained using the DFT-D3 dispersion correction method of Grimme [55]. Optically detected magnetic resonance (ODMR) spectra were simulated with the EasySpin package version 6.0.0 [56] running under MATLAB R2023b, using the pepper routine in hybrid mode; the full input script is reproduced in Supplementary Note 4. The Jahn–Teller Hamiltonian was solved with our jahn-teller-dynamics Python package [41].

Since excitation spectra based purely on DFT energy levels are not always reliable for comparison with experiment [57], we also employed MBPT within the GW approximation and the Bethe–Salpeter equation (BSE) to predict the optical properties of both defects and the host lattice. MBPT explicitly accounts for electron–hole interactions in the excited state, where the excited electron experiences the Coulomb potential of the remaining hole; these correlated excitations are described by a two-particle Green’s function that fulfills the BSE [58]. Due to the high computational cost of MBPT, we used a smaller 6×6×16 \times6 \times1 216-atom supercell. We verified explicitly that this cell is converged for the quantities of interest: repeating the Δ\DeltaSCF excited-state calculation in N×N×1N \times N \times1 (N=5,6,7,9N = 5, 6, 7, 9) and 6×6×N6 \times6 \times N (N=1,2,3N = 1, 2, 3) supercells changes the excitation energies by less than 20 meV between the 6×6×16 \times6 \times1 and the production 9×9×19 \times9 \times1 cell, and by less than 50 meV over the whole series, while the transition dipole moments vary by less than 20% (Supplementary Note 6 and Supplementary Figure 4). We emphasize that the long-range 1/r1/r finite-size errors that plague charged-defect formation energies largely cancel here, because the ground and excited states carry the same charge; the residual defect–defect coupling is of much shorter range (exponential in the orbital overlap, or 1/r31/r^3 dipole–dipole). For bulk pristine rBN, our GW+BSE calculations yielded a band gap of 5.82 eV at the KK point, in good agreement with experiment [4, 59].

Quasiparticle energies were obtained within the evGW0_0 scheme, starting from HSE DFT results as input. This method provides an efficient and accurate approximation to the GW formalism. Quasiparticle wave functions and energies were updated until convergence was reached after four iterations. The plane-wave basis for the orbitals was kept at the 450 eV cutoff used in the DFT step, which exceeds the largest recommended cutoff of the B and N PAW potentials; the response function was expanded up to ENCUTGW = 270 eV, i.e. the VASP default of 2/32/3 of the orbital cutoff, which is the standard converged choice for sp2sp^2 BN. For the self-energy calculations, 6500 bands were included, corresponding to more than 30 times the number of occupied states of the 216-atom supercell. We note that the quantities we report are energy differences between states described in one and the same basis, which converge substantially faster

with respect to both the cutoff and the number of bands than the absolute quasiparticle energies do. Brillouin-zone sampling was restricted to the Γ\Gamma point, which proved convergent for both DFT and GW+BSE calculations. The defect-level energy differences obtained from GW were consistent with those from HSE (see Supplementary Note 5), validating their reliability for subsequent BSE optical property calculations.

Data Availability

The data supporting the findings of this study are computational. All quantities underlying the reported results—optimized supercell geometries, Kohn–Sham and quasiparticle eigenvalues, BSE excitation energies and transition dipole moments, partial Huang–Rhys factors and vibrational frequencies, and the hyperfine and zero-field-splitting tensors—are reported in full in the main text and in the Supplementary Information, which together constitute the minimum dataset needed to interpret, verify and extend this work. The underlying raw electronic-structure output files (VASP OUTCAR, vasprun.xml, POSCAR and CONTCAR files) are archived on the institutional storage of the HUN-REN Wigner Research Centre for Physics and are available from the corresponding author (A.-G. [email protected]), who will respond to requests within four weeks. No restrictions apply to their use. No experimental datasets were generated or analysed during the current study.

Code Availability

No custom code was used to generate the results of this study beyond the items listed here. The electronic-structure calculations were performed with the commercial plane-wave code VASP version 5.4.1, which is available under licence from the VASP Software GmbH; the zero-field-splitting module used within it is the standard implementation by Martijn Marsman distributed with that release, modified only by removing the 1/[S(2S−1)]1/[S(2S-1)] prefactor for the broken-symmetry singlet configuration, as described in Supplementary Note 3. The optically detected magnetic resonance spectra were simulated with EasySpin version 6.0.0, an open-source package freely available at https://easyspin.org; the complete input script is reproduced verbatim in Supplementary Note 4, so that the simulations can be repeated without any additional code. The photoluminescence lineshape was obtained with our own Huang–Rhys implementation, described in Ref. 39 and based on the theory of Alkauskas and co-workers [35]. That code is not distributed publicly but is available from the corresponding author on request; it does not constitute a key advance of the present paper, since the advance is the physical result rather than the software, and it is fully documented in the cited publication. The e⊗Ee \otimes E Jahn–Teller Hamiltonian was solved with the jahn-teller-dynamics Python package, which is freely available from the public repository linked in the code paper of Ref. 41.

Acknowledgements

A.G. acknowledges access to high-performance computational resources provided by KIFÜ (Governmental Agency for IT Development, Hungary). G.T. acknowledges the support from the János Bolyai Research Scholarship of the Hungarian Academy of Sciences. N.E. acknowledges the scholarship support from the Iranian Ministry of Science, Research, and Technology.

Funding

A.G. discloses support for the research of this work from the European Commission [SPINUS, Grant No. 101135699] and the European Commission [QuSPARC, Grant No. 101186889]. G.T. discloses support for the research of this work from the National Office of Research, Development and Innovation of Hungary (NK-FIH) [STARTING Grant No. 150113]. N.E. and I.A.S. disclose support for the research of this work from the Iranian National Science Foundation (INSF) [Project No. 4021945].

Author Contributions

N.E. and G.T. performed the computations. The analysis of the results was carried out by N.E., supervised by I.A.S., G.T., and A.G. A.G. conceived and supervised the project. All authors contributed to manuscript writing.

Competing Interests

The authors declare no competing interests.

Application of Huang–Rhys Theory

Within the Huang–Rhys (HR) theory, one takes the following approximations:

  • Both the ground and excited states exhibit the same parabolic confinement potential with the same set of vibration modes: (ℏωi)(\hbar\omega_i), where i=1…(3N−3)i=1\ldots(3N-3) spans all vibration modes of a rBN supercell hosting NN atoms.

  • The electron–phonon coupling manifests itself only by that these two adiabatic potential energy surfaces (APES) are shifted from each other by a fixed amount by 2αi2\alpha_i.

Optical transitions are accompanied by phonon sidebands. The corresponding lineshape can be described within the model of two displaced quantum harmonic oscillators. Because the minima of the two potential energy surfaces are displaced relative to each other, not only the zero-phonon-line (ZPL) transition is allowed, but transitions involving different phonon quanta also acquire finite intensity due to the nonzero overlap of their vibrational wavefunctions. The square of this dimensionless displacement is known as the partial Huang–Rhys (HR) factor, Si=2αi2S_i=2\alpha_i^2, which characterizes the strength of the electron–phonon coupling for each vibrational mode. Accordingly, the potential energy surfaces of the ground and excited electronic states, V−(xi)V_-(x_i) and V+(xi)V_+(x_i), can be written in terms of the dimensionless normal coordinates xix_i as follows:

V∓(xi)=∓Δ2+∑i(ℏωi2xi2+fixi)=∓Δ2+∑iℏωi2(xi∓fiℏωi)2−fi2(ℏωi)2.(2)V_{\mp}(x_i)=\mp\frac{\Delta}{2}+\sum_i\left(\frac{\hbar\omega_i}{2}x_i^2+f_i x_i\right)=\mp\frac{\Delta}{2}+\sum_i\frac{\hbar\omega_i}{2}\left(x_i\mp\frac{f_i}{\hbar\omega_i}\right)^2-\frac{f_i^2}{(\hbar\omega_i)^2}. \tag*{(2)}

where Δ\Delta is the bare electronic excitation energy. The quantity fif_i is the derivative of the total energy with respect to the normal coordinate, i.e., the generalized force experienced by the system upon electronic excitation. By completing the square, one finds that the dimensionless displacement is given by αi=fi/(ℏωi)\alpha_i=f_i/(\hbar\omega_i). In ab initio calculations, however, the static displacement is typically obtained in Å units, which allows the shift in generalized coordinates to be computed as follows [35]:

Next, the generalized coordinates qiq_i provide a convenient way to define the Huang–Rhys factors SiS_i (or, equivalently, the dimensionless displacements αi\alpha_i):

Si=ωiqi22ℏor2αi=ωiℏ qi.(3)S_i=\frac{\omega_i q_i^2}{2\hbar}\quad\text{or}\quad2\alpha_i=\sqrt{\frac{\omega_i}{\hbar}}\,q_i. \tag*{(3)}

These quantities allow the zero-temperature lineshape function to be written as follows (see the Supplementary Material of Ref. [38] for details):

F(ℏω,T=0 K)=∑i,νi∣⟨0i(g)∣νi(e)⟩∣2δ(ℏω−∑iνiℏωi)=e−Stot∑n=0∞Stotnn!In(ℏω).(4)F(\hbar\omega,T=0\,\mathrm{K})=\sum_{i,\nu_i}\left|\langle0_i^{(g)}\mid\nu_i^{(e)}\rangle\right|^2\delta\left(\hbar\omega-\sum_i\nu_i\hbar\omega_i\right)=e^{-S_{\mathrm{tot}}}\sum_{n=0}^{\infty}\frac{S_{\mathrm{tot}}^n}{n!}I_n(\hbar\omega). \tag*{(4)}

Here, ∣νi(e)⟩\lvert\nu_i^{(e)}\rangle denotes the vibrational wavefunction of the excited state with νi\nu_i phonon quanta, whereas ⟨0i(g)∣\langle0_i^{(g)}\rvert denotes the vibrational wavefunction of the ground state with zero phonon quanta. The functions In(ℏω)I_n(\hbar\omega) are defined recursively by convolution,

In(ℏω)=∫d(ℏω′) In−1(ℏω′) S(ℏω−ℏω′).(5)I_n(\hbar\omega)=\int d(\hbar\omega')\,I_{n-1}(\hbar\omega')\,S(\hbar\omega-\hbar\omega'). \tag*{(5)}

Finally,

S(ℏω)=I1(ℏω)=∑iSiδ(ℏω−ℏωi)(6)S(\hbar\omega)=I_1(\hbar\omega)=\sum_i S_i\delta(\hbar\omega-\hbar\omega_i) \tag*{(6)}

is the Huang–Rhys spectral function, which determines the first phonon sideband, whereas the zero-phonon line is given by

I0(ℏω)=δ(ℏω),(7)I_0(\hbar\omega)=\delta(\hbar\omega), \tag*{(7)}

and the total Huang–Rhys factor is

Stot=∑iSi.(8)S_{\mathrm{tot}} = \sum_i S_i . \tag*{(8)}

At finite temperature (T>0T > 0), it is convenient to use the generating-function formalism based on the Fourier transform pair S(ℏω)↔S(t)S(\hbar\omega) \leftrightarrow S(t). The generating function is then defined as

G(t)=eS(t)−S(0),(9)G(t) = e^{S(t)-S(0)} , \tag*{(9)}

where

S(t)=∫−∞+∞[n(ℏω)e+iωt+(n(ℏω)+1)e−iωt]S(ℏω) d(ℏω),(10)S(t) = \int_{-\infty}^{+\infty} \left[n(\hbar\omega)e^{+i\omega t} + \left(n(\hbar\omega)+1\right)e^{-i\omega t}\right] S(\hbar\omega)\,\mathrm{d}(\hbar\omega) , \tag*{(10)}

and

n(ℏω)=[exp⁡(ℏωkBT)−1]−1(11)n(\hbar\omega) = \left[\exp\left(\frac{\hbar\omega}{k_{\mathrm{B}}T}\right)-1\right]^{-1} \tag*{(11)}

is the Bose–Einstein occupation factor for the vibrational quanta. The temperature-dependent lineshape function is then given by

F(ℏω,T)=12π∫−∞+∞G(t)eiωt−γ∣t∣ dt(12)F(\hbar\omega,T) = \frac{1}{2\pi}\int_{-\infty}^{+\infty}G(t)e^{i\omega t-\gamma|t|}\,\mathrm{d}t \tag*{(12)}

where γ\gamma may be chosen to reproduce the experimental broadening of the ZPL.

One must still distinguish between the cases of photoluminescence (PL) and absorption (A):

IPL(ℏω,T)=(ℏω)3F(ΔZPL−ℏω,T)andIA(ℏω,T)=(ℏω)F(ΔZPL+ℏω,T),(13)I_{\mathrm{PL}}(\hbar\omega,T) = (\hbar\omega)^3 F(\Delta_{\mathrm{ZPL}}-\hbar\omega,T) \quad\text{and}\quad I_{\mathrm{A}}(\hbar\omega,T) = (\hbar\omega) F(\Delta_{\mathrm{ZPL}}+\hbar\omega,T) , \tag*{(13)}

where ΔZPL\Delta_{\mathrm{ZPL}} is the zero-phonon transition energy. Thus, the PL and absorption lineshapes are weighted by factors of (ℏω)3(\hbar\omega)^3 and (ℏω)1(\hbar\omega)^1, respectively.

We additionally note that the Stokes shift, i.e., the difference between the vertical absorption and vertical emission energies, can be written as follows [34]:

EStokes=V+(αi)−V−(αi)⏟vertical absorption−[V+(−αi)+V−(−αi)]⏟vertical emission=(Δ+∑iℏωiSi)−(Δ−∑iℏωiSi)≈2Stotℏ⟨ωi⟩,(14)E_{\mathrm{Stokes}} = \underbrace{V_{+}(\alpha_i)-V_{-}(\alpha_i)}_{\text{vertical absorption}}-\underbrace{\left[V_{+}(-\alpha_i)+V_{-}(-\alpha_i)\right]}_{\text{vertical emission}} = \left(\Delta+\sum_i\hbar\omega_iS_i\right)-\left(\Delta-\sum_i\hbar\omega_iS_i\right)\approx2S_{\mathrm{tot}}\hbar\langle\omega_i\rangle, \tag*{(14)}

where ⟨ωi⟩\langle\omega_i\rangle denotes the SiS_i-weighted mean vibrational frequency,

⟨ωi⟩=∑iSiωiStot.(15)\langle\omega_i\rangle= \frac{\sum_i S_i\omega_i}{S_{\mathrm{tot}}} . \tag*{(15)}

Hyperfine tensors

Calculated values of the hyperfine coupling constants for nitrogen and boron nuclear spins are presented in Supplementary Tables 1 and 2, respectively, for the most significant neighboring sites of VB, see Supplementary Figure 2(a). Since the hyperfine interaction with farther nuclear spins is much weaker, we only consider the nearest nuclear spins. The spin density of the excited state shows a reduction of symmetry as shown in Supplementary Figure 2(b).

Determination of ZFS tensors

Determination of ZFS tensors

The zero-field-splitting (ZFS) tensor for a two-particle wavefunction

∣ψiψj⟩=ψi(r1)ψj(r2)−ψj(r1)ψi(r2)2(16)\lvert\psi_i\psi_j\rangle= \frac{\psi_i(r_1)\psi_j(r_2)-\psi_j(r_1)\psi_i(r_2)}{\sqrt{2}} \tag*{(16)}

Figure Supplementary Figure 1. Partial Huang-Rhys factors individually distributed into totally symmetric A1A_1 and Jahn-Teller active EE modes at 0 K temperature.

Calculated spin density for the ground and excited states

Figure Supplementary Figure 2. Calculated spin density. (a) Ground state. (b) Excited state. The vacant boron site located in the middle layer of rBN is depicted as a semitransparent green ball.

constructed from orbitals ψi\psi_i and ψj\psi_j with the same spin projection (e.g. both ↑\uparrow) can be calculated as follows, including exchange effects [52, 60]:

D(ij)=32Dzz(ij)=⟨ψiψj∣32fzz∣ψiψj⟩=32(⟨ψi(r1)ψj(r2)|fzz(r1−r2)|ψi(r1)ψj(r2)⟩−⟨ψi(r1)ψj(r2)|fzz(r1−r2)|ψj(r1)ψi(r2)⟩),(17)D^{(ij)} = \frac{3}{2}D_{zz}^{(ij)} = \left\langle\psi_i\psi_j \left| \frac{3}{2}f_{zz} \right| \psi_i\psi_j \right\rangle= \frac{3}{2}\left(\left\langle\psi_i(r_1)\psi_j(r_2)\middle|f_{zz}(r_1-r_2)\middle|\psi_i(r_1)\psi_j(r_2)\right\rangle- \left\langle\psi_i(r_1)\psi_j(r_2)\middle|f_{zz}(r_1-r_2)\middle|\psi_j(r_1)\psi_i(r_2)\right\rangle\right), \tag*{(17)}

where

fab(r)=r2δab−3rarbr5(18)f_{ab}(r) = \frac{r^2\delta_{ab}-3r_a r_b}{r^5} \tag*{(18)}

is the magnetic dipole–dipole kernel, with a,b∈{x,y,z}a,b \in\{x,y,z\} denoting Cartesian coordinates and r=x2+y2+z2r = \sqrt{x^2+y^2+z^2}.

At this point, we consider a C3vC_{3v}-symmetric ZFS tensor, which has only one independent parameter:

Dab=(DxxDxyDxzDxyDyyDyzDxzDyzDzz)=13(−D000−D0002D).(19)D_{ab} = \begin{pmatrix} D_{xx} & D_{xy} & D_{xz} \\ D_{xy} & D_{yy} & D_{yz} \\ D_{xz} & D_{yz} & D_{zz} \end{pmatrix} = \frac{1}{3}\begin{pmatrix} -D & 0 & 0 \\ 0 & -D & 0 \\ 0 & 0 & 2D \end{pmatrix}. \tag*{(19)}

Therefore, it is sufficient to evaluate the DzzD_{zz} tensor element. For simplicity, the physical constants μB\mu_B and gg are omitted.

This simplification becomes especially relevant for the ∣3E⟩\lvert{}^3E\rangle excited state, where orbital degeneracy can in principle induce off-diagonal elements in DabD_{ab}. However, these effects are expected to be absent even at low temperatures for VB−\mathrm{V}_{\mathrm{B}}^{-} in rBN because of strong electron-phonon interaction (see the small Ham reduction factor). As a consequence, only the splitting between the mS=±1m_S=\pm1 and mS=0m_S=0 sublevels remains, which can then be parametrized by a single ZFS constant DD.

siteAxxA_{xx}AyyA_{yy}AzzA_{zz}AxyA_{xy}
N1N_180.56657.86548.30419.670
N1N_180.58457.87348.329-19.669
N1N_146.50591.92448.3030.005
N2N_26.6994.4844.3720.594
N2N_25.5535.6284.3711.256
N2N_26.6944.4804.364-1.255
N2N_26.6944.4804.364-0.593
N2N_24.5226.6594.3710.660
N2N_24.5246.6594.372-0.662
N3N_30.8300.172-0.3980.000
N3N_30.3350.665-0.3990.286
N3N_30.3360.666-0.398-0.285
N4N_40.3150.3670.2460.101
N4N_40.4390.2370.2420.029
N4N_40.2620.4140.242-0.073
N4N_40.2660.4150.2460.073
N4N_40.4420.2400.247-0.029
N4N_40.3160.3670.247-0.102

Table 1. Supplementary Table 1. Ab initio hyperfine parameters for 14N{}^{14}\mathrm{N} nuclei I=1I = 1 nuclear spins in the negatively charged boron-vacancy in rBN. All hyperfine values are in MHz unit.

siteAxxA_{xx}AyyA_{yy}AzzA_{zz}AxyA_{xy}
B1B_1-1.7405.358-4.3260.899
B1B_1-1.7405.358-4.326-0.899
B1B_12.8050.813-4.3263.523
B1B_12.8050.813-4.326-3.523
B1B_14.362-0.744-4.3262.624
B1B_14.362-0.744-4.326-2.624
B2B_2-1.2130.167-1.6510.000
B2B_2-0.178-0.868-1.6510.597
B2B_2-0.178-0.868-1.651-0.597
B3B_32.9202.0571.725-0.446
B3B_31.8853.0901.7240.151
B3B_31.8863.0901.725-0.151
B3B_32.9202.0571.7250.446
B3B_32.6582.3171.7240.597
B3B_32.6592.3181.725-0.597
B4B_4-0.977-0.272-1.1960.000
B4B_4-0.447-0.800-1.195-0.306
B4B_4-0.448-0.801-1.1960.305

Table 2. Supplementary Table 2. Ab initio hyperfine parameters for 11B{}^{11}\mathrm{B} nuclei with I=3/2I = 3/2 nuclear spins in the negatively charged boron-vacancy in rBN. All hyperfine values are in MHz unit.

More generally, for an NN-electron system, the ZFS tensor can be evaluated as a sum over occupied spin orbitals. For example, for the ∣3A2gnd⟩\lvert{}^{3}A_{2}^{\mathrm{gnd}}\rangle state of VB−\mathrm{V}_{\mathrm{B}}^{-},

Dgnd=1S(2S−1)∑i,jocc.D(ij),(20)D_{\mathrm{gnd}} = \frac{1}{S(2S-1)} \sum_{i,j}^{\mathrm{occ.}} D^{(ij)}, \tag*{(20)}

where the sum runs over all occupied orbitals in both spin channels. In the present case,

i,j∈{VB orbitals,a1↑,a1↓,a1↑∗,a1↓∗,ex↑,ex↓,ey↑,ey↓,ex↓∗,ey↓∗},(21)i,j \in\{\mathrm{VB\ orbitals}, a_{1}^{\uparrow}, a_{1}^{\downarrow}, a_{1}^{\uparrow*}, a_{1}^{\downarrow*}, e_{x}^{\uparrow}, e_{x}^{\downarrow}, e_{y}^{\uparrow}, e_{y}^{\downarrow}, e_{x}^{\downarrow*}, e_{y}^{\downarrow*}\}, \tag*{(21)}

as depicted in Figure 1b of the main text. We emphasize that the sum must also include all valence-band (VB) orbitals. By contrast, the orbitals {ex↑∗,ey↑∗}\{e_{x}^{\uparrow*}, e_{y}^{\uparrow*}\} are unoccupied and therefore do not enter the sum. For triplet states (S=1S = 1), the prefactor 1/[S(2S−1)]1/[S(2S - 1)] is equal to unity.

We compute the ZFS tensor using the implementation in VASP 5.4.1 as implemented by Martijn Marsman. For the two-hole configuration

∣3A2gnd,mS=+1⟩=∣ex⋆ey⋆⟩,(22)|{}^{3}A_{2}^{\mathrm{gnd},m_{S}=+1}\rangle= |e_{x}^{\star}e_{y}^{\star}\rangle, \tag*{(22)}

accessible within our DFT approach, we obtain

DZFS[∣ex⋆ey⋆⟩]=2627 MHz.(23)D_{\mathrm{ZFS}}\left[|e_{x}^{\star}e_{y}^{\star}\rangle\right] = 2627\ \mathrm{MHz}. \tag*{(23)}

As noted in the main text, we employ the spin-decontamination scheme proposed in Ref. 54, which requires evaluation of the ZFS for the broken-symmetry singlet configuration:

D[∣ey⋆ex⋆⟩]=−4250 MHz.(24)D\left[|e_{y}^{\star}e_{x}^{\star}\rangle\right] = -4250\ \mathrm{MHz}. \tag*{(24)}

In this case, the prefactor 1/[S(2S−1)]1/[S(2S-1)] must be removed from the VASP source code to avoid division by zero. The corrected ground-state ZFS is then obtained by averaging the ferromagnetic and antiferromagnetic solutions, taking into account the sign change of the latter:

Dtheorygnd=12(D[∣ex⋆ey⋆⟩]−D[∣ex⋆ey⋆⟩])=12(2627 MHz+4250 MHz)=3438 MHz.(25)D_{\mathrm{theory}}^{\mathrm{gnd}} = \frac{1}{2}\left(D\left[|e_{x}^{\star}e_{y}^{\star}\rangle\right]-D\left[|e_{x}^{\star}e_{y}^{\star}\rangle\right]\right) = \frac{1}{2}\left(2627\ \mathrm{MHz}+4250\ \mathrm{MHz}\right)=3438\ \mathrm{MHz}. \tag*{(25)}

This value is remarkably close to the experimentally observed room-temperature value,

Dexpt.gnd≈3.45 GHz,(26)D_{\mathrm{expt.}}^{\mathrm{gnd}} \approx3.45\ \mathrm{GHz}, \tag*{(26)}

which supports the plausibility of the correction scheme for VB−\mathrm{V}_{\mathrm{B}}^{-}.

Tentative derivation of the ZFS for the ∣3E⟩|{}^{3}E\rangle excited multiplet

We first construct the symmetry-adapted wavefunctions for the ground and excited multiplets:

∣3A2gnd⟩=∣ex⋆ey⋆⟩,|{}^{3}A_{2}^{\mathrm{gnd}}\rangle= |e_{x}^{\star}e_{y}^{\star}\rangle,
∣1A1⟩=(∣exex⋆⟩+∣eyey⋆⟩)/2,∣3Ex⟩=a∣a1ex⋆⟩+b(−∣exex⋆⟩+∣eyey⋆⟩)+c∣a1⋆ex⋆⟩,∣3A2⟩=(∣exey⋆⟩−∣eyex⋆⟩)/2,∣3Ey⟩=a∣a1ey⋆⟩+b(∣exey⋆⟩+∣eyex⋆⟩)+c∣a1⋆ey⋆⟩.(27)\begin{aligned} |{}^{1}A_{1}\rangle&= \left(|e_{x}e_{x}^{\star}\rangle+|e_{y}e_{y}^{\star}\rangle\right)/\sqrt{2}, \qquad|{}^{3}E_{x}\rangle= a|a_{1}e_{x}^{\star}\rangle+b\left(-|e_{x}e_{x}^{\star}\rangle+|e_{y}e_{y}^{\star}\rangle\right)+c|a_{1}^{\star}e_{x}^{\star}\rangle,\\ |{}^{3}A_{2}\rangle&= \left(|e_{x}e_{y}^{\star}\rangle-|e_{y}e_{x}^{\star}\rangle\right)/\sqrt{2}, \qquad|{}^{3}E_{y}\rangle= a|a_{1}e_{y}^{\star}\rangle+b\left(|e_{x}e_{y}^{\star}\rangle+|e_{y}e_{x}^{\star}\rangle\right)+c|a_{1}^{\star}e_{y}^{\star}\rangle. \tag*{(27)} \end{aligned}

For simplicity, the ↑\uparrow spin labels are omitted. According to our GW+BSE results, the expansion coefficients are a2=0.57a^{2}=0.57, 2b2=0.322b^{2}=0.32, and c2=0.07c^{2}=0.07, as shown in Figure 1c of the main text. Here, ∣3Ex,y⟩|{}^{3}E_{x,y}\rangle refers to the lower ∣3E⟩|{}^{3}E\rangle multiplet. The ∣1A1⟩|{}^{1}A_{1}\rangle level lies well above the upper ∣3E⟩|{}^{3}E\rangle state, is optically inactive, and appears at 2.35 eV.

Because we cannot evaluate the ZFS directly on top of the GW+BSE wavefunctions, we instead use Δ\DeltaSCF calculations. In this approach, we obtain the following two values entering the decontamination correction:

DΔSCFex.=12(D[ΨΔSCF↑↑]−D[ΨΔSCF↑↓])=12(−564 MHz+3579 MHz)=1508 MHz,(28)D_{\Delta\mathrm{SCF}}^{\mathrm{ex.}}=\frac{1}{2}\left(D\left[\Psi_{\Delta\mathrm{SCF}}^{\uparrow\uparrow}\right]-D\left[\Psi_{\Delta\mathrm{SCF}}^{\uparrow\downarrow}\right]\right)=\frac{1}{2}\left(-564\ \mathrm{MHz}+3579\ \mathrm{MHz}\right)=1508\ \mathrm{MHz}, \tag*{(28)}

where

ΨΔSCF↑↑≈∣aa1↑+bex↑+ca1⋆↑;ey⋆⟩(29)\Psi_{\Delta\mathrm{SCF}}^{\uparrow\uparrow}\approx|aa_{1}^{\uparrow}+be_{x}^{\uparrow}+ca_{1}^{\star\uparrow};e_{y}^{\star}\rangle \tag*{(29)}

and

ΨΔSCF↑↓≈∣aa1↑+bex↑+ca1⋆↑;ey⋆⟩(30)\Psi_{\Delta\mathrm{SCF}}^{\uparrow\downarrow}\approx|aa_{1}^{\uparrow}+be_{x}^{\uparrow}+ca_{1}^{\star\uparrow};e_{y}^{\star}\rangle \tag*{(30)}

are the single-Slater-determinant triplet and broken-symmetry singlet Δ\DeltaSCF solutions, respectively, approximately representing the ∣3Ey⟩|{}^{3}E_{y}\rangle triplet.

Thus, we identify the Δ\DeltaSCF component of ∣3Ey⟩|{}^{3}E_{y}\rangle as

∣3Ey⟩=∣aa1+ca1⋆+bex;ey⋆⟩+b∣eyex⋆⟩.(31)|{}^{3}E_{y}\rangle= |aa_{1}+ca_{1}^{\star}+be_{x};e_{y}^{\star}\rangle+ b|e_{y}e_{x}^{\star}\rangle. \tag*{(31)}

For simplicity, we neglect the multiconfigurational contribution b∣eyex⋆⟩b|e_{y}e_{x}^{\star}\rangle when estimating the excited-state ZFS.

Visualization of the broken-symmetry Kohn–Sham orbital

Figure S3. Visualization of the broken-symmetry Kohn-Sham orbital ∣aa1↑+bex↑+ca1↑⟩\lvert aa_1^{\uparrow} + be_x^{\uparrow} + ca_1^{\uparrow} \rangle that of (a) ΨΔSCF↑↓\Psi_{\Delta\mathrm{SCF}}^{\uparrow\downarrow} in Eq. (29) and (b) ΨΔSCF↑↑\Psi_{\Delta\mathrm{SCF}}^{\uparrow\uparrow} in Eq. (30).

SUPPLEMENTARY NOTE 4: EASYSPIN CODE

We generated Figure 3 of the main text by the following Easyspin code:

clear

Sys.S = 1; % two electrons spin with S=1/2
Sys.g = 2; % isotropic g for two electron spins
%Sys.lw= 10.0; % Gaussian Isotropic broadening
Sys.HStrain = [10 10 10]; %Gaussian FWHM Anisotropic broadening in all directions [MHz]
%(For hyperfine splittings)
Sys.D= [3450 10]; %MHz [D E]
Sys.A = [4.362 -0.744 -4.326; ... %11B hyperfine tensors: Axx, Ayy, Azz
         4.362 -0.744 -4.326; ... %11B.
         2.805  0.813 -4.326; ... %11B
         2.805  0.813 -4.326; ... %11B
        -1.74   5.358 -4.326; ... %11B
        -1.74   5.358 -4.326; ... %11B
        80.566 57.865 48.312; ... %14N
        46.506 91.925 48.304; ... %14N
        80.566 57.865 48.304];    %14N %MHz

Sys.Nucs = ’11B, 11B, 11B, 11B, 11B, 11B, 14N, 14N, 14N’;
Exp.Field = 9; %mT
Exp.Temperature = 4; %kelvin
Exp.mwRange = [3 4]; %GHz
%Exp.mwFreq = [];
Exp.CrystalSymmetry= ’C3v’;
Exp.MolFrame = [0 0 0];
Exp.SampleFrame = [0 0 0]; %sample aligned with lab frame
%Exp.CrystalOrientation = [0 0 1];
Exp.nPoints = 1500;
%Exp.Range = [0 100];
%Opt.Method = ’perturb1’;
%Opt.Verbosity = 1;

Opt.Method = ’hybrid’;
Opt.HybridCoreNuclei = [7 8 9]; % these atoms treated exactly and others perturbatively

pepper(Sys,Exp,Opt);

SUPPLEMENTARY NOTE 5: COMPARISON OF EVGW0 AND DFT-HSE LEVELS

In Supplementary Table 3 we list the quasiparticle and Kohn-Sham levels in the Γ\Gamma-point for the 216-atom 6×6×16 \times6 \times1 supercell. The results imply that the DFT-HSE spectrum is a good starting point for the excited state calculation by the Δ\DeltaSCF method.

↑\uparrow orb.value (eV)occ.↓\downarrow orb.value (eV)occ.
VBM0.89(1.13)1VBM0.89(1.13)1
a1↑a_1^{\uparrow}0.89(1.13)1a1↓a_1^{\downarrow}1.53(1.71)1
ex↑e_x^{\uparrow}1.48(1.29)1ex↓e_x^{\downarrow}1.47(1.72)1
ey↑e_y^{\uparrow}1.48(1.29)1ey↓e_y^{\downarrow}1.47(1.72)1
a1∗↑a_1^{*\uparrow}1.16(1.42)1a1∗↓a_1^{*\downarrow}2.32(1.81)1
ex∗↑e_x^{*\uparrow}1.26(1.50)1ex∗↓e_x^{*\downarrow}5.56(5.70)0
ey∗↑e_y^{*\uparrow}1.26(1.50)1ey∗↓e_y^{*\downarrow}5.56(5.70)0
CBM7.61(7.13)0CBM7.61(7.13)0

Table 3. Supplementary Table 3. The evGW0 quasiparticle (DFT-HSE) levels in the Γ\Gamma-point are given in eV unit as obtained in 216-atom supercell.

Finite Size Convergence

Both Reviewers asked whether the 6×6×16 \times6 \times1 (216-atom) supercell employed in the GW+BSE calculations is sufficiently converged, given that the ground-state properties reported in the main text were obtained in a 9×9×19 \times9 \times1 (486-atom) supercell. We address this here.

We first note a point of principle. The 1/r1/r Coulomb finite-size errors that make charged defects notoriously slowly convergent enter the formation energy and hence the charge transition levels, because there the charge state of the supercell changes. All quantities reported here—excitation energies, transition dipole moments, the phonon sideband, the ZFS and the hyperfine tensors—are instead evaluated within one and the same charge state, so the monopole and dipole image terms cancel between the initial and final state to leading order. What remains is the direct defect–defect coupling, which decays exponentially with the separation of the localized dangling-bond orbitals, and the residual dipole–dipole term, which decays as 1/r31/r^3. Convergence is therefore expected to be, and is found to be, much faster than for charge transition levels.

To make this quantitative we repeated the constrained-occupation (Δ\DeltaSCF) HSE calculation of the two lowest excited states in six supercells, spanning both in-plane and out-of-plane directions (Supplementary Fig. 4). Because the same Δ\DeltaSCF protocol can be run in every cell, including the 6×6×16 \times6 \times1 cell used for GW+BSE and the 9×9×19 \times9 \times1 cell used for the production runs, this provides a direct, like-for-like test of the effect of the cell size on the excited state, which is exactly the quantity the Reviewers ask about.

In-plane extension [Supplementary Fig. 4a; N×N×1N \times N \times1 with N=5N = 5, 6, 7, 9 hosting 149, 215, 293 and 485 atoms]. Going from 6×6×16 \times6 \times1 to 9×9×19 \times9 \times1 changes the excitation energy of the first (3E^3E) excited state by −18-18 meV and that of the second excited state by +16+16 meV. Over the entire series the total spread is below 50 meV for the second excited state and about 60 meV for the first, with a clean exponential approach to the asymptote (solid lines are exponential fits).

Out-of-plane extension [Supplementary Fig. 4b; 6×6×N6 \times6 \times N with N=1N = 1, 2, 3 hosting 215, 431 and 647 atoms]. Doubling the cell along the cc shifts the Δ\DeltaSCF excitation energy by only ∼20\sim20 meV, and tripling it by a further ∼2\sim2 meV. The rapid convergence reflects the fact that a single 6×6×16 \times6 \times1 rBN supercell already contains three BN sheets, so the local ABC stacking environment of the vacancy—which is what breaks the mirror symmetry and drives the entire effect reported in this work—is fully represented.

The transition dipole moments, which enter the radiative rate quadratically, are given next to each data point in Supplementary Fig. 4. They vary between 0.17 and 0.21 D for the first excited state and between 0.32 and 0.35 D for the second, i.e. by less than 20% across all cells, corresponding to less than a factor of 1.5 in the radiative rate. We emphasize that these are single-particle (Kohn–Sham) transition dipoles evaluated at the Δ\DeltaSCF level; they are used here purely as a convergence metric. The absolute dipole moments quoted in the main text (Figure 1d) are the excitonic dipoles from GW+BSE, which are larger because the correlated exciton is a coherent superposition of several electron–hole pairs (a1→e∗a_1 \rightarrow e^*, e→e∗e \rightarrow e^* and a1∗→e∗a_1^* \rightarrow e^*, see Figure 1c) whose contributions add constructively.

We conclude that the 6×6×16 \times6 \times1 supercell reproduces the 9×9×19 \times9 \times1 result to better than 20 meV in the excitation energy and better than 20% in the transition dipole moment. These residual finite-size errors are far smaller than the ∼\sim0.1 eV intrinsic accuracy of the underlying electronic-structure methods, and they are entirely negligible compared with the one- to two-orders-of-magnitude difference in radiative rate between rBN and hBN that constitutes the central claim of this work. The reason the larger 9×9×19 \times9 \times1 cell was nevertheless used for the ground-state properties is not convergence of the excitation energy but the phonon sideband, the hyperfine and the ZFS tensors: the Huang–Rhys analysis requires a dense set of supercell vibrational modes and the hyperfine tensors require several shells of neighbors, both of which benefit from the larger cell, and both of which are affordable at the HSE level but not at the GW+BSE level.

Figure 4. Finite-size convergence of the Δ\DeltaSCF excited states of VB−V_{\mathrm{B}}^{-} in rBN. Red symbols, first excited state 3A2(e⋆e⋆)↔3E(a1e⋆){}^{3}A_{2}(e^{\star}e^{\star}) \leftrightarrow{}^{3}E(a_{1}e^{\star}), polarized perpendicular to the C3 axis; blue symbols, second excited state 3A2(e⋆e⋆)↔3A2(ee⋆){}^{3}A_{2}(e^{\star}e^{\star}) \leftrightarrow{}^{3}A_{2}(ee^{\star}), polarized parallel to it. Solid lines are exponential fits to the asymptote. The number next to each point is the corresponding Kohn–Sham transition dipole moment, used here as a convergence metric only (see text). a In-plane enlargement of the BN layer, N×N×1N \times N \times1 with N=5N = 5, 6, 7, 9, hosting 149, 215, 293 and 485 atoms, respectively. b Stacking of supercells along the cc axis, 6×6×N6 \times6 \times N with N=1N = 1, 2, 3, hosting 215, 431 and 647 atoms, respectively.

References

  1. [1]W. Olovsson and M. Magnuson, Rhombohedral and turbostratic boron nitride polytypes investigated by x-ray absorption spectroscopy, The Journal of Physical Chemistry C 126, 21101 (2022).DOI
  2. [2]M. Zanfrognini, A. Plaud, I. Stenger, F. Fossard, L. Sponza, L. Schué, F. Paleari, E. Molinari, D. Varsano, L. Wirtz, et al., Distinguishing differing defect stackings in layered materials via luminescence spectroscopy, Physical Review Letters 131, 206902 (2023).arxiv.org/abs/2305.17554
  3. [3]M. Moret, A. Rousseau, P. Valvin, S. Sharma, L. Souqui, H. Pedersen, H. Högberg, G. Cassabois, J. Li, J. Edgar, et al., Rhombohedral and turbostratic boron nitride: X-ray diffraction and photoluminescence signatures, Applied Physics Letters 119 (2021).DOI
  4. [4]J. Qi, C. Ma, Q. Guo, C. Ma, Z. Zhang, F. Liu, X. Shi, L. Wang, M. Xue, M. Wu, P. Gao, H. Hong, X. Wang, E. Wang, C. Liu, and K. Liu, Stacking-controlled growth of rbn crystalline films with high nonlinear optical conversion efficiency up to 1%, Advanced Materials 36, 2303122 (2024).DOI
  5. [5]J. Iwański, K. P. Korona, M. Tokarczyk, G. Kowalski, A. K. Dąbrowska, P. Tatarczak, I. Rogala, M. Bilska, M. Wójcik, and S. Kret, Revealing polytypism in 2d boron nitride with uv photoluminescence, npj 2D Materials and Applications 8, 72 (2024).arxiv.org/abs/2405.19126
  6. [6]L. Wang, J. Qi, W. Wei, M. Wu, Z. Zhang, X. Li, H. Sun, Q. Guo, M. Cao, Q. Wang, C. Zhao, Y. Sheng, Z. Liu, C. Liu, M. Wu, Z. Xu, W. Wang, H. Hong, P. Gao, M. Wu, Z.-J. Wang, X. Xu, E. Wang, F. Ding, X. Zheng, K. Liu, and X. Bai, Bevel-edge epitaxy of ferroelectric rhombohedral boron nitride single crystal, Nature 629, 74 (2024).DOI
  7. [7]T. T. Tran, K. Bray, M. J. Ford, M. Toth, and I. Aharonovich, Quantum emission from hexagonal boron nitride monolayers, Nature Nanotechnology 11, 37 (2016).
  8. [8]Z. Shotan, H. Jayakumar, C. R. Considine, M. Mackoit, H. Fedder, J. Wracht­rup, A. Alkauskas, M. W. Doherty, V. M. Menon, and C. A. Meriles, Photoinduced Modification of Single-Photon Emitters in Hexagonal Boron Nitride, ACS Photonics 3, 2490 (2016).DOI
  9. [9]L. J. Martínez, T. Pelini, V. Waselowski, J. R. Maze, B. Gil, G. Cassabois, and V. Jacques, Efficient single photon emission from a high-purity hexagonal boron nitride crystal, Physical Review B 94, 121405 (2016).DOI
  10. [10]M. Bourrellier, S. Meuret, A. Tararan, C. Stéphan, M. Kociak, L. H. G. Tizei, and A. Zobelli, Bright UV Single Photon Emission at Point Defects in h-BN, Nano Letters 16, 4317 (2016).DOI
  11. [11]N. Chejanovsky, A. Mukherjee, J. Geng, Y.-C. Chen, Y. Kim, A. Denisenko, A. Finkler, T. Taniguchi, K. Watanabe, D. B. R. Dasari, P. Auburger, A. Gali, J. H. Smet, and J. Wrachtrup, Single-spin resonance in a van der Waals embedded paramagnetic defect, Nature Materials 20, 1079 (2021), number: 8.DOI
  12. [12]H. L. Stern, C. Qiu, G. J. Jarman, S. Eizagirre Barker, N. Mendelson, D. Chugh, S. Schott, H. H. Tan, H. Sirringhaus, I. Aharonovich, and M. Atatüre, Room-temperature optically detected magnetic resonance of single defects in hexagonal boron nitride, Nature Communications 13, 618 (2022).DOI
  13. [13]N.-J. Guo, S. Li, W. Liu, Y.-Z. Yang, X.-D. Zeng, S. Yu, Y. Meng, Z.-P. Li, Z.-A. Wang, L.-K. Xie, R.-C. Ge, J.-F. Wang, Q. Li, J.-S. Xu, Y.-T. Wang, J.-S. Tang, A. Gali, C.-F. Li, and G.-C. Guo, Coherent control of an ultra-bright single spin in hexagonal boron nitride at room temperature, Nature Communications 14, 2893 (2023).DOI
  14. [14]R. N. Patel, R. E. K. Fishman, T.-Y. Huang, J. A. Gusdorff, D. A. Fehr, D. A. Hopper, S. A. Breitweiser, B. Porat, M. E. Flatté, and L. C. Bassett, Room Temperature Dynamics of an Optically Addressable Single Spin in Hexagonal Boron Nitride, Nano Letters 24, 7623 (2024).DOI
  15. [15]H. L. Stern, C. M. Gilardoni, Q. Gu, S. Eizagirre Barker, O. F. J. Powell, X. Deng, S. A. Fraser, L. Follet, C. Li, A. J. Ramsay, H. H. Tan, I. Aharonovich, and M. Atatire, A quantum coherent spin in hexagonal boron nitride at ambient conditions, Nature Materials 23, 1379 (2024).
  16. [16]X. Gao, S. Vaidya, K. Li, Z. Ge, S. Dikshit, S. Zhang, P. Ju, K. Shen, Y. Jin, Y. Ping, and T. Li, Single nuclear spin detection and control in a van der Waals material, Nature 643, 943 (2025).DOI
  17. [17]P. Auburger and A. Gali, Towards ab initio identification of paramagnetic substitutional carbon defects in hexagonal boron nitride acting as quantum bits, Phys. Rev. B 104, 075410 (2021).DOI
  18. [18]S. Li, A. Pershin, and A. Gali, Quantum emission from coupled spin pairs in hexagonal boron nitride, Nature Communications 16, 5842 (2025).DOI
  19. [19]I. O. Robertson, B. Whitefield, S. C. Scholten, P. Singh, A. J. Healey, P. Reineck, M. Kianinia, G. Barcza, V. Ivády, D. A. Broadway, I. Aharonovich, and J.-P. Tetienne, A charge transfer mechanism for optically addressable solid-state spin pairs, Nature Physics 21, 1981 (2025).DOI
  20. [20]B. Whitefield, H. Z. J. Zeng, J. Liddle-Wesolowski, I. O. Robertson, A. Ganyecz, V. Ivády, K. Watanabe, T. Taniguchi, M. Toth, J.-P. Tetienne, I. Aharonovich, and M. Kianinia, Narrowband quantum emitters in hexagonal boron nitride with optically addressable spins, Nature Materials 25, 412 (2026).DOI
  21. [21]M. Abdi, J.-P. Chou, A. Gali, and M. B. Plenio, Color Centers in Hexagonal Boron Nitride Monolayers: A Group Theory and Ab Initio Analysis, ACS Photonics 5, 1967 (2018).DOI
  22. [22]A. Gottscholl, M. Kianinia, V. Soltamov, S. Orlinskii, G. Mamin, C. Bradac, C. Kasper, K. Krambrock, A. Sperlich, M. Toth, I. Aharonovich, and V. Dyakonov, Initialization and read-out of intrinsic spin defects in a van der waals crystal at room temperature, Nature materials 19, 540 (2020).DOI
  23. [23]P. Udvarhelyi, T. Clua-Provost, A. Durand, J. Li, J. H. Edgar, B. Gil, G. Cassabois, V. Jacques, and A. Gali, A planar defect spin sensor in a two-dimensional material susceptible to strain and electric fields, npj Computational Materials 9, 1 (2023), number: 1.arxiv.org/abs/2304.00492
  24. [24]J. E. Fröch, L. P. Spencer, M. Kianinia, D. D. Totonjian, M. Nguyen, A. Gottscholl, V. Dyakonov, M. Toth, S. Kim, and I. Aharonovich, Coupling Spin Defects in Hexagonal Boron Nitride to Monolithic Bullseye Cavities, Nano Letters 21, 6549 (2021).
  25. [25]J. R. Reimers, J. Shen, M. Kianinia, C. Bradac, I. Aharonovich, M. J. Ford, and P. Piecuch, Photoluminescence, photophysics, and photochemistry of the vB− defect in hexagonal boron nitride, Physical Review B 102, 144105 (2020).DOI
  26. [26]V. Ivády, G. Barcza, G. Thiering, S. Li, H. Hamdi, J.-P. Chou, Ö. Legeza, and A. Gali, Ab initio theory of the negatively charged boron vacancy qubit in hexagonal boron nitride, npj Computational Materials 6, 41 (2020).DOI
  27. [27]F. Libbi, P. M. M. de Melo, Z. Zanolli, M. J. Verstraete, and N. Marzari, Phonon-assisted luminescence in defect centers from many-body perturbation theory, Physical Review Letters 128, 167401 (2022).DOI
  28. [28]Z. Benedek, Á. Ganyecz, O. Bulancea-Lindvall, G. Barcza, and V. Ivády, An extended ab initio theory of the V_B center in hBN: excited states, Jahn-Teller distortion, and pressure dependence, arXiv preprint arXiv:2605.04283 10.48550/arXiv.2605.04283 (2026).
  29. [29]A. Gale, M. Kianinia, J. Horder, C. Tweedie, M. Singhal, D. Scognamiglio, J. Qi, K. Liu, C. Verdi, I. Aharonovich, and M. Toth, Quantum Emitters in Rhombohedral Boron Nitride, Advanced Optical Materials 13, e00593 (2025), eprint: https://advanced.onlinelibrary.wiley.com/doi/pdf/10.1002/adom.202
  30. [30]N. Estaji, I. A. Sarsari, G. Thiering, and A. Gali, Spin-phonon relaxation of boron vacancy centers in two-dimensional boron nitride polytypes, Physical Review B 112, L201407 (2025).DOI
  31. [31]A. Gali, T. Demján, M. Vörös, G. Thiering, E. Canuccia, and A. Marini, Electron-vibration coupling induced renormalization in the photoemission spectrum of diamondoids, Nature Communications 7, 10.1038/ncomms11327 (2016).DOI
  32. [32]K. Huang and A. Rhys, Theory of light absorption and non-radiative transitions in f-centres, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 204, 406 (1950).
  33. [33]R. Kubo and Y. Toyozawa, Application of the method of generating function to radiative and non-radiative transitions of a trapped electron in a crystal, Progress of Theoretical Physics 13, 160–182 (1955).DOI
  34. [34]M. de Jong, L. Seijo, A. Meijerink, and F. T. Rabouw, Resolving the ambiguity in the relation between stokes shift and huang–rhys parameter, Physical Chemistry Chemical Physics 17, 16959–16969 (2015).DOI
  35. [35]A. Alkauskas, B. B. Buckley, D. D. Awschalom, and C. G. Van de Walle, First-principles theory of the luminescence lineshape for the triplet transition in diamond nv centres, New Journal of Physics 16, 073026 (2014).DOI
  36. [36]Y. Jin, M. Govoni, G. Wolfowicz, S. E. Sullivan, F. J. Heremans, D. D. Awschalom, and G. Galli, Photoluminescence spectra of point defects in semiconductors: Validation of first-principles calculations, Physical Review Materials 5, 084603 (2021).DOI
  37. [37]M. Lax, The frank-condon principle and its application to crystals, The Journal of chemical physics 20, 1752 (1952).
  38. [38]P. Kehayias, M. W. Doherty, D. English, R. Fischer, A. Jarmola, K. Jensen, N. Leefer, P. Hemmer, N. B. Manson, and D. Budker, Infrared absorption band and vibronic structure of the nitrogen-vacancy center in diamond, Phys. Rev. B 88, 165202 (2013).DOI
  39. [39]G. Thiering and A. Gali, Ab initio calculation of spin-orbit coupling for an NV center in diamond exhibiting dynamic jahn-teller effect, Physical Review B 96, 081115 (2017).DOI
  40. [40]N. Mathur, A. Mukherjee, X. Gao, J. Luo, B. A. McCul-lian, T. Li, A. N. Vamivakas, and G. D. Fuchs, Excited-state spin-resonance spectroscopy of vB− defect centers in hexagonal boron nitride, Nature Communications 13, 10.1038/s41467-022-30772-z (2022).DOI
  41. [41]B. Tóth, Á. Gali, and G. Thiering, Jahn-Teller-dynamics: Python package to determine vibronic interaction demonstrated on molecules and trigonal defect qubits, Computer Physics Communications 327, 110299 (2026).
  42. [42]G. Thiering and A. Gali, Nuclear-spin relaxation in solid-state-defect quantum bits via electron-phonon coupling in the optically excited state, Phys. Rev. Appl. 24, 044027 (2025).DOI
  43. [43]S. Baber, R. N. E. Malein, P. Khatri, P. S. Keatley, S. Guo, F. Withers, A. J. Ramsay, and I. J. Luxmoore, Excited state spectroscopy of boron vacancy defects in hexagonal boron nitride using time-resolved optically detected magnetic resonance, Nano Letters 22, 461–467 (2021).DOI
  44. [44]P. Yu, H. Sun, M. Wang, T. Zhang, X. Ye, J. Zhou, H. Liu, C.-J. Wang, F. Shi, Y. Wang, and J. Du, Excited-state spectroscopy of spin defects in hexagonal boron nitride, Nano Letters 22, 3545–3549 (2022).DOI
  45. [45]Z. Mu, H. Cai, D. Chen, J. Kenny, Z. Jiang, S. Ru, X. Lyu, T. S. Koh, X. Liu, I. Aharonovich, and W. Gao, Excited-state optically detected magnetic resonance of spin defects in hexagonal boron nitride, Phys. Rev. Lett. 128, 216402 (2022).DOI
  46. [46]X. Gao, S. Vaidya, K. Li, P. Ju, B. Jiang, Z. Xu, A. E. L. Allcca, K. Shen, T. Taniguchi, K. Watanabe, S. A. Bhave, Y. P. Chen, Y. Ping, and T. Li, Nuclear spin polarization and control in hexagonal boron nitride, Nature Materials 21, 1024–1028 (2022).DOI
  47. [47]P. E. Blöchl, Projector augmented-wave method, Physical review B 50, 17953 (1994).pubmed.ncbi.nlm.nih.gov/9976227
  48. [48]G. Kresse and J. Hafner, Ab initio molecular-dynamics simulation of the liquid-metal–amorphous-semiconductor transition in germanium, Physical Review B 49, 14251 (1994).pubmed.ncbi.nlm.nih.gov/10010505
  49. [49]G. Kresse and J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Physical review B 54, 11169 (1996).pubmed.ncbi.nlm.nih.gov/9984901
  50. [50]A. Gali, E. Janzén, P. Deák, G. Kresse, and E. Kaxiras, Theory of spin-conserving excitation of the n-v-center in diamond, Physical review letters 103, 186404 (2009).DOI
  51. [51]K. Szász, T. Hornos, M. Marsman, and A. Gali, Hyperfine coupling of point defects in semiconductors by hybrid density functional calculations: The role of core spin polarization, Phys. Rev. B 88, 075202 (2013).DOI
  52. [52]Z. Brodgal and A. Gali, The spin–spin zero-field splitting tensor in the projector-augmented-wave method, Journal of Physics: Condensed Matter 26, 015305 (2013).
  53. [53]J. Heyd, G. E. Scuseria, and M. Ernzerhof, Hybrid functionals based on a screened coulomb potential, The Journal of chemical physics 118, 8207 (2003).DOI
  54. [55]S. Grimme, J. Antony, S. Ehrlich, and H. Krieg, A consistent and accurate ab initio parametrization of density functional dispersion correction (dft-d) for the 94 elements h-pu, The Journal of chemical physics 132 (2010).DOI
  55. [56]S. Stoll and A. Schweiger, Easyspin, a comprehensive software package for spectral simulation and analysis in epr, Journal of magnetic resonance 178, 42 (2006).pubmed.ncbi.nlm.nih.gov/16188474
  56. [57]C. Attaccalite, M. Bockstedte, A. Marini, A. Rubio, and L. Wirtz, Coupling of excitons and defect states in boron-nitride nanostructures, Physical Review B—Condensed Matter and Materials Physics 83, 144115 (2011).DOI
  57. [58]G. Onida, L. Reining, and A. Rubio, Electronic excitations: density-functional versus many-body green’s-function approaches, Reviews of modern physics 74, 601 (2002).DOI
  58. [59]L. Sponza, H. Amara, C. Attaccalite, S. Latil, T. Galvani, F. Paleari, L. Wirtz, and F. Ducastelle, Direct and indirect excitons in boron nitride polymorphs: A story of atomic configuration and electronic correlationDOI
  59. [60]M. J. Rayson and P. R. Briddon, First principles method for the calculation of zero-field splitting tensors in periodic systems, Phys. Rev. B 77, 035119 (2008).DOI

Paper details

vlnc:W_d2rbwvnh
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents