betaFeedback

WorksPhysics and Astronomy

C2CNC_2C_N as a 2 eV Single-Photon Emitter Candidate in Hexagonal Boron Nitride

Abstract

The generation of single-photon emitters in hexagonal boron nitride around 2 eV emission is experimentally well-recognized; however the atomic nature of these emitters is unknown. In this paper, we use first-principles calculations to demonstrate that C2CNC_2C_N is a possible source of 2 eV single-photon emitter. We showcase the calculations of a complete set of static and dynamical properties related to defects, including exciton-defect couplings and electron-phonon interactions. In particular, we show it is critical to consider nonradiative processes when comparing with experimental lifetime for known 2 eV emitters. We find that C2CNC_2C_N has several key physical properties matching the ones of experimentally observed single-photon emitters. These include the zero-phonon line (2.13eV), Huang-Rhys factor (1.35), photoluminescence lifetime (2.19 ns), phonon-sideband energy (180meV), and photoluminescence spectrum. The identification of defect candidates for 2 eV emission paves the way for controllable single-photon emission generation.

Introduction

A single-photon emitter is a crucial component in building quantum information technologies, such as linear-optic quantum information processing [1], quantum simulation [2] and quantum communication [3]. Hexagonal boron nitride (hBN) is a two-dimensional (2D) material with a wide bandgap (∼6\sim6 eV) [4, 5], and can host stable and bright color centers which possess single-photon emission [6]. Meanwhile, the manipulation of the optoelectronic properties of singe-photon emitters has garnered special interest [7, 8]. These properties imply a great potential of hBN in developing quantum applications. Being able to select and purify single-photon emitters is critical for generating controllable and narrow line width single-photon emission. Hence, the identification of the atomic origin of the single-photon emitters is crucial to the development of this field.

Since 2016 [9], numerous experimental results have been reported on the photoluminescent properties of the single-photon emitters at ∼2\sim2 eV in hBN [27] [7, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26]. In TABLE I, we summarize the range and averaged values of photoluminescence properties observed in experiments, including the zero-phonon line (ZPL), phonon-sideband energy (PSE), Huang-Rhys factor (HR factor), and the photoluminescence lifetime (τPL\tau^{PL}). Typically, a ∼2\sim2 eV single-photon emitter in hBN exhibits a photoluminescence (PL) spectrum comprised of a sharp ZPL and one or two moderate phonon-sidebands (PSB) at room temperature. The PSE is the energy separation between

the ZPL and the first PSB peak. The electron-phonon coupling is estimated by either HR factor (SS) or Debye-Waller (DW factor =e−S=e^{-S}) [28]. The HR factor is typically ∼1\sim1, which indicates weak electron-phonon coupling. Furthermore, the PL lifetime is on the order of a few ns, and quantum yield (QY, η\eta) is reported to be 6–12% [8]. PL lifetime (τPL\tau^{PL}) reflects the lifetime of an excited state, determined by radiative (τR\tau^R) and nonradiative recombination together (τNR\tau^{NR}, τPL=1/(1/τR+1/τNR)\tau^{PL}=1/(1/\tau^R+1/\tau^{NR})). On the other hand, QY (η\eta) reflects the proportion of radiative recombination rates with respect to the total recombination rates, and is related to PL lifetime through τPL=τR∗η\tau^{PL}=\tau^R*\eta. PL lifetime and QY underscore the importance of studying both radiative and nonradiative recombination lifetimes. Finally, many of the reported single-photon emitters ∼2\sim2 eV possess linearly polarized excitation and emission [29, 31, 32] [9, 13, 15, 16, 17, 18, 19, 23], indicating the anisotropic structural symmetry of the corresponding defects [33, 34], i.e. possibly belonging to C2vC_{2v}, C2C_2 or CsC_s groups.

In terms of the atomic origin of these single-photon emitters, annular dark-field images have shown that carbon substitutions are abundant in hBN [35]. Recently, Mendelson et al. [13] has identified that various techniques of incorporating carbon into hBN yield ∼2\sim2 eV single-photon emitters. Moreover, X-ray photoelectron spectroscopy measurements show that more C-B bonds exist than that of C-N bonds [13]. This is an evidence that carbon substitution of nitrogen is more likely than carbon substitution of boron.

Theoretically, several defects have been proposed to be possibly responsible for single-photon emission, such as CB_\mathrm{B}VN_\mathrm{N} [36] [6], boron dangling bonds [37], NB_\mathrm{B}VN_\mathrm{N} [38], and carbon trimers [39]. Based on experimental obser-

ZPL (eV)PSE (meV)HR factorτPL\tau^{\mathrm{PL}} (ns)
Range(exp.)1.56–2.24a^{a}30–200b^{b}0.63–1.93c^{c}0.38–19.7d^{d}
Mean(exp.)2.00 ± 0.19a^{a}156 ± 33b^{b}1.19 ± 0.43c^{c}2.6 ± 1.5d^{d}
C2_2CN_\mathrm{N}(cal.)2.13e^{e}1801.352.19
C2_2CB_\mathrm{B}(cal.)1.42e^{e}1751.253.83 × 102^{2}

Table 1. Summary of the zero-phonon line (ZPL), phonon-sideband (PSE), Huang-Rhys (HR) factor and photoluminescence (PL) lifetime (τPL\tau^{\mathrm{PL}}) for ∼2 eV single-photon emitters in hh-BN from experimental measurements and our calculated two carbon defects.

vations [13], we focus on carbon defects in this work. Among the carbon defects, CB_\mathrm{B}VN_\mathrm{N} has high formation energy, 3∼6 eV higher than other carbon defects such as carbon dimer and trimers [40]. Carbon trimers including C2_2CN_\mathrm{N} and C2_2CB_\mathrm{B}, theoretically proposed by Jara et al. [39], were found energetically favorable and in good agreement with experimental PSE and PL [39]. However, some important information is still missing to confirm carbon trimers as a SPE candidate in Ref. 39. For example, only the ZPL for the lowest transition was calculated, about 0.4 eV smaller than the mean of ZPL in TABLE I. And other important properties such as HR factor, PL lifetime and QY, as the experimental characteristics of ∼2 eV single-photon emitters, were not calculated. Therefore, further study on these properties is desired to unveil the role of carbon trimers as experimentally observed SPEs.

In this paper, electronic and optical properties of the carbon trimers C2_2CN_\mathrm{N} and C2_2CB_\mathrm{B} are calculated from both first-principles many-body perturbation theory and DFT levels. We evaluate both static and dynamical properties of radiative and phonon-assisted nonradiative recombination, including exciton-defect and electron-phonon interactions. Our results demonstrate that C2_2CN_\mathrm{N} has remarkable agreement with experimental observations of 2 eV single photon emitters, including ZPL, PL lifetime, PSE, HR factor, and PL spectrum. We emphasize the importance of comparing all key signatures between theory and experiments for defect identification and validations.

Computational Methods

We employ open source plane-wave code Quantum ESPRESSO [41] for structural relaxation and phonon calculations of carbon defects in monolayer hBN. We use the optimized norm-conserving Vanderbilt (ONCV) pseudopotentials [42] and a 55 Ry wavefunction energy cutoff. We choose a supercell size of 6 × 6 which shows good convergence as tested in Ref. 38, 43, 44. Charged defect correction is included to eliminate the spurious electrostatic interaction by using the techniques developed in Refs. 45, 46 and implemented in the JDFTx code [47]. Total energy, defect formation energy and geometry are obtained with the Perdew-Burke-Ernzerhof (PBE) exchange correlation functional [48]. The charged defect formation energy Efq(d)E_{\mathrm{f}}^{q}(d) with the charge state qq is calculated by

Efq(d)=Etotq(d)−Etot(p)−∑iNiμi+qEFermi+Ecorr,(1)E_{\mathrm{f}}^{q}(d) = E_{\mathrm{tot}}^{q}(d) - E_{\mathrm{tot}}(p) - \sum_i N_i\mu_i + qE_{\mathrm{Fermi}} + E_{\mathrm{corr}}, \tag*{(1)}

where Etotq(d)E_{\mathrm{tot}}^{q}(d) is the total energy of the charged defect, Etot(p)E_{\mathrm{tot}}(p) is the total energy of the pristine system, NiN_i is the number of atoms of atomic species ii that is added (Ni>0N_i > 0) or removed (Ni<0N_i < 0), μi\mu_i is the chemical potential of the atomic species ii, EFermiE_{\mathrm{Fermi}} is the electron chemical potential, and EcorrE_{\mathrm{corr}} is the charged defect correction. The chemical potentials of B, N and C are obtained as follows. In the N-rich condition, μNN−rich=1/2Etot(N2)\mu_{\mathrm{N}}^{\mathrm{N-rich}} = 1/2 E_{\mathrm{tot}}(\mathrm{N}_2) where Etot(N2)E_{\mathrm{tot}}(\mathrm{N}_2) is the total energy of N2_2 molecule. In the N-poor condition, μBN−poor=Etot(B)\mu_{\mathrm{B}}^{\mathrm{N-poor}} = E_{\mathrm{tot}}(\mathrm{B}) where Etot(B)E_{\mathrm{tot}}(\mathrm{B}) is one atom’s total energy in a boron crystal. μNN−rich\mu_{\mathrm{N}}^{\mathrm{N-rich}} and μBN−rich\mu_{\mathrm{B}}^{\mathrm{N-rich}} are calculated according to the constraint μN+μB=μBN\mu_{\mathrm{N}} + \mu_{\mathrm{B}} = \mu_{\mathrm{BN}}, where μBN\mu_{\mathrm{BN}} is the total energy of BN with one unit cell. μC\mu_{\mathrm{C}}, on the other hand, is one atom’s total energy in graphene.

We evaluate electronic structures under GW approximation with the PBE eigenvalues and wavefunctions as the starting point by using the Yambo code [49]. The starting wavefunctions at PBE are sufficient for accurate descriptions of current spsp defect wavefunctions by comparing with the ones at hybrid functionals (see SI FIG. S1 for wavefunction comparison [44, 50, 51]). In single-shot GW (G0W0G_0W_0) calculations in this work, the Godby-Needs plasmon-pole approximation (PPA) [52, 53] is used to calculate the dielectric matrices with the plasmon frequency ωp=27.2\omega_p = 27.2 eV. Then Bethe-Salpeter Equation (BSE) is further solved on top of the GW approximation to include the electron-hole interaction in absorption spectra. Coulomb truncation for 2D systems [54] is applied to the out-of-plane direction, and k-point sampling is set to 3 × 3 × 1 for 6 × 6 × 1 supercells in all of the G0W0G_0W_0 and BSE calculations. More details on convergence tests can be found in SI Section II.

The method in Ref. 39 based on Fermi’s golden rule and solving BSE is applied for calculating radiative lifetime of intra-defect transitions that include exciton-defect coupling. Nonradiative recombination lifetime is also calculated with electron-phonon matrix elements in the static coupling approximation. The details of the nonradiative recombination can be found in Ref. 38, 55. The generating function approach, which is detailed in Ref. 56, is applied to the calculation of PL lineshape. More benchmark and convergence tests on PL calculations can be found in SI Section V.

Results and Discussions

Our results and discussions are organized as follows. First, we discuss the defect stability and their stable charge states through defect formation energies and charge transition levels. Then, we present the electronic structure, optical spectra, and radiative lifetime of the defects. Next, we show ZPL and nonradiative recombination processes of the intra-defect transitions (transition between two states localized on one isolated defect). We then discuss competition between radiative and nonradiative recombination processes and the resulting PL lifetime. Moreover, the PL lineshape and spectral function with the dominant phonon mode are discussed. Finally, we show a full picture comprised of ZPL, PL lifetimes, PSE, HR factor, and PL spectrum of C2_2CN in comparison with experiments.

Charged Defect Formation Energy

When nitrogen and boron atoms are unevenly substituted by atomic impurities, the formation energy of the defects depends on the elemental chemical potentials, i.e. different at the N-rich (B-poor) and N-poor (B-rich) conditions. In the N-rich condition, C2_2CB is more likely to form than C2_2CN because of the smaller formation energy of C2_2CB, as shown in Figure 1(b). On the other hand, C2_2CN is more likely to form in the N-poor condition. It is also found that both defects can be stable in charge states of q=0,±1q = 0, \pm1 at a range of electron chemical potentials within the electronic band gap. Without losing generality, we investigate both C2_2CN and C2_2CB in all possible charge states of q=0,±1q = 0, \pm1, and we find that the q=±1q = \pm1 charge states have ZPL either too large or too small, away from the experimental range (details found in SI TABLE S1). Thus, we focus on the neutral state of the defects. In particular, we find that the neutral C2_2CN have good agreement with experiments on all properties of ∼2\sim2 eV single-photon emitter, so we mainly discuss the neutral C2_2CN in the following. The results of C2_2CB can be found in SI Section III and summarized in Table I. We find the major discrepancy of C2_2CB to experiments is in ZPL and PL lifetime.

Charged defect formation energies of C$_2$CN and C$_2$CB as a function of Fermi level

Figure 1. Charged defect formation energy of defects C2_2CN and C2_2CB as a function of Fermi level at (a) N-poor and (b) N-rich conditions.

Electronic Structure and Optical Properties

The single-particle diagram in Figure 2(a) shows the defect-related electronic energy levels with their wavefunctions, and the host monolayer hBN band edges at the GW approximation, referenced to the vacuum level. One unoccupied defect state (2a22a_2) and three occupied defect states (2b12b_1, 1a21a_2, 1b11b_1) in the spin-down channel can lead to intra-defect transitions via radiative recombination or nonradiative recombination. To obtain optical transitions (or two-particle excitations), we calculate the absorption spectra by solving the BSE, which uses GW quasiparticle energies as inputs and includes excitonic effect. Details of numerical convergence tests can be found in SI Section II. Here only spin-conserved transitions are considered. We then calculate the radiative lifetimes of the intra-defect recombination via Fermi’s golden rule by the Eq. (2) (with dielectric constant equal to unity for monolayer 2D systems),

Single-particle diagram of the ground state and absorption spectra of C2CN

Figure 2. (a) Single-particle diagram of the ground state 2A2{}^{2}A_{2} at the level of G0_{0}W0_{0}@PBE, and (b) G0_{0}W0_{0}/BSE@PBE of C2_{2}CN_{\mathrm{N}}. VBM and CBM are -7.431 eV and -0.400 eV, respectively [60]. The defect states in the band gap are denoted by the irreducible representations of C2v_{2v} symmetry group based on corresponding wavefunction symmetry. The isosurface of the wavefunctions (PBE) is 3% of the maximum. In the G0_{0}W0_{0}/BSE@PBE spectra, the absorption peaks are labeled by the corresponding intra-defect transitions. x and y are the in-plane directions that are perpendicular and parallel to the C2_2 axis, respectively, and z is the out-of-plane direction. The spectral broadening is 0.02 eV.

τR=3c34E03μe−h2,(2)\tau^{R} = \frac{3c^{3}}{4E_{0}^{3}\mu_{e-h}^{2}}, \tag*{(2)}

where E0E_0 is the excitation energy, and μe−h2\mu_{e-h}^{2} is modulus square of exciton dipole moment in atomic units. The derivation detail of this equation can be found in Ref. 33, 57. We find two strong absorption peaks due to the optically allowed intra-defect transitions in Figure 2(b) and summarize the optical properties in TABLE II. In particular, only the transition 2a2↓→1a2↓2a_{2\downarrow} \to1a_{2\downarrow} shows a relatively short radiative lifetime that possibly falls into the range of experimental values. The other transition has an order of magnitude longer radiative lifetime. Therefore, we mainly focus on the former transition. Its shorter lifetime (i.e. 51.9 ns) is because of stronger oscillator strength and higher excitation energy (shown in Table II). Note that the excitonic effect at the defect is comparable or stronger than its host materials [58, 59], e.g. over one eV exciton binding energy EbE_b for both transitions in Table II. The exciton wavefunction in SI FIG. S3 shows that the exciton is localized and bound to the carbon defect C2_2CN and a few neighboring BN atoms, consistent with its large exciton binding energy. In order to compare with the experimental PL lifetimes in TABLE I, we also need to compute nonradiative lifetime as follows.

TransitionE0E_0 (eV)μe−h2\mu_{e-h}^{2} (a.u.)τR\tau^{\mathrm{R}} (ns)EbE_b (eV)
2a2↓→2b1↓2a_{2\downarrow} \rightarrow2b_{1\downarrow}1.344.90∗10−14.90 \ast10^{-1}7.95∗1027.95 \ast10^{2}2.21
2a2↓→1a2↓2a_{2\downarrow} \rightarrow1a_{2\downarrow}2.331.435.19∗1015.19 \ast10^{1}1.76

Table 2. Radiative recombination of C2_2CN_{\mathrm{N}}. Excitation energy (E0E_0), square modulus of dipole moment (μe−h2\mu_{e-h}^{2}), radiative lifetime (τR\tau^{\mathrm{R}}), and exciton binding energy (EbE_b) of the C2_2CN_{\mathrm{N}} defect for the two transitions that are prominent in the optical excitations below the optical gap.

Nonradiative Recombination

Nonradiative lifetime (τNR\tau^{\mathrm{NR}}) is a measure of how fast the nonradiative recombination happens between the final state ∣fi⟩\lvert f_i\rangle and initial state ∣i⟩\lvert i\rangle. The phonon-assisted nonradiative recombination is influenced by several factors and also evaluated via Fermi’s golden rule as below,

1τifNR=2πℏg∑n,mpin∣⟨fm∣He−ph∣in⟩∣2δ(Efm−Ein)(3)\frac{1}{\tau_{if}^{\mathrm{NR}}} = \frac{2\pi}{\hbar} g \sum_{n,m} p_{in} \left| \langle f_m \lvert H^{e-ph} \rvert i_n \rangle\right|^2 \delta(E_{fm}-E_{in}) \tag*{(3)}

where He−phH^{e-ph} is the electron-phonon coupling Hamiltonian, gg is the degeneracy factor of the final state that depends on the number of equivalent atomic configurations, and pinp_{in} is the occupation number of the vibronic state ∣in⟩\lvert i_n\rangle following Boltzmann distribution.

Under the static coupling approximation with one-dimensional (1D) phonon approximation [38, 55], we can rewrite Eq. (3) as

1τifNR=2πℏg∣Wif∣2Xif(T)(4)\frac{1}{\tau_{if}^{\mathrm{NR}}} = \frac{2\pi}{\hbar}g\lvert W_{if}\rvert^{2}X_{if}(T) \tag*{(4)}
Wif=⟨ψi(r,R)∣∂H∂Q∣ψf(r,R)⟩R=Ra(5)W_{if} = \left\langle\psi_i(\mathbf{r},\mathbf{R}) \left\lvert\frac{\partial H}{\partial Q} \right\rvert\psi_f(\mathbf{r},\mathbf{R}) \right\rangle_{\mathbf{R}=\mathbf{R}_a} \tag*{(5)}
Xif=∑n,mpin∣⟨ϕfm(R)∣Q−Qa∣ϕin(R)⟩∣2×δ(mℏωf−nℏωi+ΔEif).(6)X_{if} = \sum_{n,m}p_{in}\left\lvert\left\langle\phi_{fm}(\mathbf{R}) \left\lvert Q-Q_a \right\rvert\phi_{in}(\mathbf{R}) \right\rangle\right\rvert^{2} \times\delta(m\hbar\omega_f-n\hbar\omega_i+\Delta E_{if}). \tag*{(6)}

Eq. 4 is separated into the electronic term (WifW_{if}) that depends on the electronic wavefunction (ψ\psi) overlap, and the phonon term (XifX_{if}) that describes the strength of phonon contribution. The phonon term includes the energy conservation between initial and final vibronic states with vibrational frequencies of ωi\omega_i and ωf\omega_f, and ϕ\phi is the phonon wavefunction. The detailed derivation can be found in Ref. 38, 55.

We summarize the nonradiative recombination lifetime of the intra-defect transitions, along with ZPL, HR factor of the final state (SfS_f(1D)) in TABLE III. Note that the HR factor from the nonradiative recombination calculation is with 1D effective phonon approximation. Its comparison with the full-phonon HR factor (SfS_f, including all phonon eigenmodes) can be found in SI Section V. The calculated 1D HR factor is close to the full-phonon HR factor as can be seen by comparing Table III (1D) with Figure 3 (full phonon). ZPL is evaluated by two methods: one is using the vertical neutral excitation energy obtained from BSE with subtracting the Frank-Condon shift in the excited state (ZPL(BSE)), and another is from constrained DFT (CDFT) at the PBE level with geometry optimization (ZPL(CDFT)). The ZPL from two methods has 0.1 and 0.4 eV energy difference for the first two transitions, respectively. We used the CDFT in the nonradiative lifetime calculation to be consistent with other quantities in the equation. The nonradiative recombination of 2A2→02A2{}^{2}A_{2} \to{}^{2}_{0}A_{2} is fast due to the following reason. The electronic term WifW_{if} of 2A2→02A2{}^{2}A_{2} \to{}^{2}_{0}A_{2} is large because it is symmetry-allowed, about four orders of magnitude larger than the other two transitions. Since τNR\tau^{NR} is inversely proportional to the square of WifW_{if} according to Eq. (4), difference of WifW_{if} dominates over phonon contribution in XifX_{if}. The nonradiative lifetime of 2A2→02A2{}^{2}A_{2} \to{}^{2}_{0}A_{2} (2.29 ns) is several orders shorter than other two transitions.

Multiplet structure and important physical parameters of C2CN in hBN

Figure 3. Multiplet structure and important physical parameters of C2_2CN in hBN. 02A2{}^{2}_{0}A_{2} is the ground state, and 12B1{}^{2}_{1}B_{1} and 22A2{}^{2}_{2}A_{2} are the excited states. The red solid lines denote radiative recombination and the blue dashed lines denote nonradiative recombination. SS is the HR factor with full phonon calculations. τR\tau^{R} and τNR\tau^{NR} are the radiative lifetime and nonradiative lifetime, respectively. η\eta and τPL\tau^{PL} are the QY and PL lifetime, respectively.

TransitionEFCE_{\mathrm{FC}} (eV)ZPL(BSE) (eV)ZPL(CDFT) (eV)ΔQ\Delta Q (amu1/2^{1/2}Å)ℏωf\hbar\omega_f (meV)Sf(1D)S_f(1D)WifW_{if} (eV/(amu1/2^{1/2}Å))XifX_{if} (amu·Å2^2/eV)τNR\tau^{\mathrm{NR}}
1B1→2A2{}^{1}B_1 \rightarrow{}^{2}A_20.131.211.160.261220.951.34×10−41.34 \times10^{-4}7.17×10−67.17 \times10^{-6}∼1\sim1 ms
2A2→2A2{}^{2}A_2 \rightarrow{}^{2}A_20.202.131.700.271451.263.13×10−13.13 \times10^{-1}4.69×10−74.69 \times10^{-7}2.29 ns
2A2→1B1{}^{2}A_2 \rightarrow{}^{1}B_10.01–0.550.191280.675.58×10−55.58 \times10^{-5}1.05×10−21.05 \times10^{-2}3.20 μ\mus

Table III. Properties of nonradiative recombination of the intra-defect transitions. The transitions are denoted using the multielectron wavefunction notations. The ZPL(BSE) are evaluated by subtracting the Franck-Condon shift (EFCE_{\mathrm{FC}}) from the excitation energies from G0W0G_0W_0/BSE calculations. The ZPL(CDFT) are obtained by constrained DFT calculations at PBE level, which are used as the energy input for the nonradiative lifetime. ΔQ\Delta Q is the nuclear coordinate change between the initial and final states. ℏωf\hbar\omega_f is the phonon energy of the final state. Sf(1D)S_f(1D) is the HR factor with 1D effective phonon approximation. WifW_{if} and XifX_{if} are the electronic and phonon terms, respectively. The nonradiative lifetimes are calculated with 6×66 \times6 supercell at 300 K at PBE level. Here, 1B1→2A2{}^{1}B_1 \rightarrow{}^{2}A_2 is related to the transition 2a2↓→2b1↓2a_{2\downarrow} \rightarrow2b_{1\downarrow} in the single particle picture; transition 2A2→2A2{}^{2}A_2 \rightarrow{}^{2}A_2 is related to the transition 2a2↓→1a2↓2a_{2\downarrow} \rightarrow1a_{2\downarrow}; transition 2A2→1B1{}^{2}A_2 \rightarrow{}^{1}B_1 is related to the transition 2b1↓→1a2↓2b_{1\downarrow} \rightarrow1a_{2\downarrow}.

PL Lifetime and Quantum Yield

Next we will compute PL lifetime and quantum yield [37] using radiative and nonradiative recombination lifetime from TABLE II and Table III, respectively. We list them along with the full-phonon HR factors calculated with the method in Ref. 56, and quantum yield (η\eta) computed by Eq. (7) [61] in the multiplet diagram Figure 3.

η=1τR(1τR+∑i1τiNR)−1(7)\eta= \frac{1}{\tau^{R}}\left(\frac{1}{\tau^{R}}+\sum_i\frac{1}{\tau_i^{NR}}\right)^{-1} \tag*{(7)}

where ii denotes the ithi^{\mathrm{th}} nonradiative pathway for the transition from the initial state. In addition, we calculate the PL lifetime which is the inverse of the total recombination rate by Eq. (8)

τPL=1/(1τR+∑i1τiNR)=τR∗η.(8)\tau^{PL} = 1/\left(\frac{1}{\tau^{R}}+\sum_i\frac{1}{\tau_i^{NR}}\right)=\tau^{R} * \eta. \tag*{(8)}

For transition 2A2→02A2{}^{2}A_{2} \to{}^{2}_{0}A_{2}, the nonradiative lifetime is nearly ten times shorter than the radiative lifetime. By including the nonradiative recombination 2A2→12B1{}^{2}A_{2} \to{}^{2}_{1}B_{1}, the overall recombination leads to the quantum yield of 4.23%, about 2% lower than the experimental quantum yield range 6∼12% [8]. More importantly, the PL lifetime of transition 2A2→02A2{}^{2}A_{2} \to{}^{2}_{0}A_{2} from Eq. (8) is calculated to be 2.19 ns, exactly within the experimental range of PL lifetimes (TABLE I and Figure 5(a)). In addition, the calculated full-phonon HR factor of this transition is 1.35, indicating weak electron-phonon interaction and good agreement with the experimental HR factor in TABLE I and Figure 5(b).

Image not extracted

Figure 5. The calculated properties of C2_2CN_\mathrm{N} along with the experimental values which are summarized in TABLE. I. (a) PL lifetime vs ZPL; (b) HR factor vs ZPL; (c) PSE vs ZPL; (d) comparison between the theoretical PL spectrum (this work) and the experimental PL spectrum (Exp) [13]. In (a), (b) and (c), the blue dots are for the experimental values of ∼\sim2 eV single-photon emitters, the green dot is for the experimental PL lifetime of ∼\sim20 ns, and the red dots are for the calculated values. The calculated ZPL here is from G0_0W0_0/BSE@PBE calculation. The comparison shows that ZPL, PL lifetime, HR factor and PSE are all in the experimental range. The calculated PL spectrum also matches well with the experimental PL of the ∼\sim2 eV single-photon emission [13].

Phonon Modes of C2CN in hBN

To get more insights into electron-phonon coupling of C2_2CN in hBN, we calculate the spectral function (S(ℏω)S(\hbar\omega)) of 22A2→02A2{}^{2}_{2}A_{2} \to{}^{2}_{0}A_{2} with the partial HR factor (SkS_k) of phonon mode kk as shown in Figure 4. The phonon modes spread over the range 0∼190 meV, and the phonon modes around 180 meV have the largest contribution to the spectral function. In particular, the 187 meV phonon mode has the largest partial HR factor among these modes. By visualizing this phonon mode shown by the inset in Figure 4, we find it to be an out-of-phase phonon mode. The atoms involved in this phonon mode include not only defect-related atoms but also many B and N atoms around and away from the defect. Hence, the 187 meV phonon mode is quasi-local. To quantitatively characterize the localization of dominant phonon modes, we estimate the number of atoms involved in the vibration of a phonon mode, by projection of inverse participation ratio (IPR) on this phonon mode kk [56],

Image not extracted

Figure 4. Spectral function that shows the distribution of phonon modes and the contribution of phonon modes to the electron-phonon interaction of C2_2CN_\mathrm{N} defect in hBN. The left vertical axis and black solid line are for the spectral function, and the right vertical axis and red dots are for the partial HR factor as a function of phonon energy. The inset figures are the low energy phonon mode (60 meV) and dominant phonon mode (187 meV) of C2_2CN_\mathrm{N} for transition 2A2→2A2{}^{2}A_2 \rightarrow{}^{2}A_2. The red arrows show the atom displacement of the corresponding phonon modes.

IPRk=1∑α(∑iΔrk;α,i2)2(9)\mathrm{IPR}_k = \frac{1}{\sum_{\alpha}(\sum_i \Delta r_{k;\alpha,i}^{2})^{2}} \tag*{(9)}

where Δrk;α,i\Delta r_{k;\alpha,i} is a normalized vector that describes the displacement of the atom α\alpha along the direction ii in the phonon mode kk between the initial and final states. We then estimate the localization of the phonon mode by the localization ratio β\beta,

βk=NAT/IPRk,(10)\beta_k = \mathrm{NAT}/\mathrm{IPR}_k, \tag*{(10)}

where NAT stands for the total number of atoms in the supercell. The larger the localization ratio is, the more localized the phonon mode is. We find that the IPR of the 187 meV phonon mode is 46 for the 6x6 supercell calculation, and the localization ratio β\beta is ∼\sim2. The IPR of this mode increases as a function of supercell sizes, but the localization ratio keeps constant, as can be seen in the SI TABLE S4. This again reflects the nature of this mode being a mixture of defect and bulk phonons or a “quasi-local” mode. Besides the 187 meV phonon, the 60 meV phonon is pronounced in the low-energy range of the spectral function, as an in-phase phonon mode (see the inset in Figure 4). It also has a large IPR of 48 and β\beta of ∼\sim2. Comparing the calculated phonon mode energies with the experimental ones for 2eV single-photon emitters [62] [7], the 187 meV phonon mode is in good agreement with the ∼\sim180 meV longitudinal optical phonon that leads to the pronounced PSB.

Comparison between C2CN defect and Experiments

To further validate that C2_2CN_\mathrm{N} is a possible candidate of experimentally observed single-photon emitters, we calculate the PL of C2_2CN_\mathrm{N} for transition 2A2→2A2{}^{2}A_2 \rightarrow{}^{2}A_2 and compare it with the PL spectra of ∼\sim2 eV SPEs [13] in Figure 5. The details of the PL calculation can be found in SI Section V. The calculated PL shows similar PSB peaks to the experimental PL [13], as can be seen in Figure 5(d). A weak PSB peak next to the first PSB peak can be seen in both calculated and experimental PL spectra. This peak has been found to be more visible when the brightness is enhanced in an as-prepared array [62]. In addition, phonon-sideband energy can be a measure of the averaged phonon energy, found to be 180 meV, within the experimental range in TABLE I and Figure 5(c). The excellent agreement between the calculated and experimental physical parameters including ZPL, PL lifetime, HR factor, PSE and PL spectrum strongly suggest that the C2_2CN_\mathrm{N} defect is one possible source of ∼\sim2 eV single-photon emission. Finally, we note that the variation in experimental results shown in TABLE I can be from several sources of single-photon emitters and possible internal strains due to material preparation. Further defect identification including spin related properties can be through calculations of optically detected magnetic resonance (ODMR), which measures the spin-dependent optical contrast and is directly related to the information readout of spin qubits [63].

Conclusions

In conclusion, we investigated the possibilities of carbon defects for 2 eV single photon emission in hBN, by comparing key physical properties from first-principles calculations with experiments, including ZPL, PL lifetime and lineshape, as well as HR factor. We showcase the importance of considering multiple key signatures including both static and dynamical properties when identifying defects through theory and experimental comparison. We found C2_2CN_\mathrm{N} has the best agreement with experiments on all concerned properties. In particular, we show the radiative lifetime can be an order longer than the experimentally observed PL lifetime; but after including nonradiative processes, the agreement with experimental PL lifetime is excellent. We show the electron-phonon coupling in C2_2CN_\mathrm{N} is moderate with small HR factor and the dominant phonon mode at ∼\sim180 meV is quasi-local with significant participation of bulk atoms. Our work provides insight into the ∼\sim2 eV single-photon emission from the theoretical perspective, which is important for unraveling the unknown chemical nature of defects in hBN, manipulating defects and developing quantum applications.

Acknowledgments

We acknowledge the support by the National Science Foundation under grant no. DMR-1760260. This research used resources of the Scientific Data and Computing center, a component of the Computational Science Initiative, at Brookhaven National Laboratory under Contract No. DE-SC0012704, the lux supercomputer at UC Santa Cruz, funded by NSF MRI grant AST 1828315, the National Energy Research Scientific Computing Center (NERSC) a U.S. Department of Energy Office of Science User Facility operated under Contract No. DE-AC02-05CH11231, the Extreme Science and Engineering Discovery Environment (XSede) which is supported by National Science Foundation Grant No. ACI-

References

  1. [27]G. Grosso, H. Moon, C. J. Ciccarino, J. Flick, N. Mendelson, L. Mennel, M. Toth, I. Aharonovich, P. Narang, and D. R. Englund, Low-temperature electron–phonon interaction of quantum emitters in hexagonal boron nitride, ACS Photonics 7, 1410 (2020).arxiv.org/abs/1910.02053
  2. [28]E. Neu, D. Steinmetz, J. Riedrich-Möller, S. Gsell, M. Fischer, M. Schreck, and C. Becher, Single photon emission from silicon-vacancy colour centres in chemical vapour deposition nano-diamonds on iridium, New J. Phys. 13, 025012 (2011).
  3. [29]K. Konthasinghe, C. Chakraborty, N. Mathur, L. Qiu, A. Mukherjee, G. D. Fuchs, and A. N. Vamivakas, Rabi oscillations and resonance fluorescence from a single hexagonal boron nitride quantum emitter, Optica 6, 542 (2019).
  4. [31]A. W. Schell, H. Takashima, T. T. Tran, I. Aharonovich, and S. Takeuchi, Coupling quantum emitters in 2d materials with tapered fibers, ACS Photonics 4, 761 (2017).
  5. [32]A. L. Exarhos, D. A. Hopper, R. N. Patel, M. W. Doherty, and C. L. Bassett, Magnetic-field-dependent quantum emission in hexagonal boron nitride at room temperature, Nat. Commun. 10, 1 (2019).
  6. [33]F. Wu, D. Rocca, and Y. Ping, Dimensionality and anisotropy dependence of radiative recombination in nanostructured phosphorene, J. Mater. Chem. C 7, 12891 (2019).
  7. [34]D. Yim, M. Yu, G. Noh, J. Lee, and H. Seo, Polarization and localization of single-photon emitters in hexagonal boron nitride wrinkles, ACS Appl. Mater. Interfaces 12, 36362 (2020).
  8. [35]O. L. Krivanek, M. F. Chisholm, V. Nicolosi, T. J. Pennycook, G. J. Corbin, N. Dellby, M. F. Murfitt, C. S. Own, Z. S. Szilagyi, M. P. Oxley, et al., Atom-by-atom structural and chemical analysis by annular dark-field electron microscopy, Nature 464, 571 (2010).
  9. [36]A. Sajid and K. S. Thygesen, VncB defect as source of single photon emission from hexagonal boron nitride, 2D Mater. 7, 031007 (2020).
  10. [37]M. E. Turiansky and C. G. Van de Walle, Impact of dangling bonds on properties of h-bn, 2D Mater. 8, 024002 (2021).
  11. [38]F. Wu, T. J. Smart, J. Xu, and Y. Ping, Carrier recombination mechanism at defects in wide band gap two-dimensional materials from first principles, Phys. Rev. B 100, 081407 (2019).
  12. [39]C. Jara, T. Rausch, S. Botti, M. A. Marques, A. Norambuena, R. Coto, J. Castellanos-Aguila, J. R. Maze, and F. Munoz, First-principles identification of single photon emitters based on carbon clusters in hexagonal boron nitride, J. Phys. Chem. A 125, 1325 (2021).
  13. [40]G. Cheng, Y. Zhang, L. Yan, H. Huang, Q. Huang, Y. Song, Y. Chen, and Z. Tang, A paramagnetic neutral cbvn center in hexagonal boron nitride monolayer for spin qubit application, Comput. Mater. Sci. 129, 247 (2017).
  14. [41]P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I. Dabo, A. Dal Corso, S. de Gironcoli, S. Fabris, G. Fratesi, R. Gebauer, U. Gerstmann, C. Gougoussis, A. Kokalj, M. Lazzeri, L. Martin-Samos, N. Marzari, F. Mauri, R. Mazzarello, S. Paolini, A. Pasquarello, L. Paulatto, C. Sbraccia, S. Scandolo, G. Sclauzero, A. P. Seitsonen, A. Smogunov, P. Umari, and R. M. Wentzcovitch, QUANTUM ESPRESSO: A Modular and Open-Source Project for Quantum Simulations of Materials., J. Phys.: Condens. Matter 21, 395502 (2009).DOI
  15. [42]D. R. Hamann, Optimized Norm-Conserving Vanderbilt Pseudopotentials, Phys. Rev. B 88, 085117 (2013).DOI
  16. [44]T. J. Smart, K. Li, J. Xu, and Y. Ping, Intersystem Crossing and Exciton–Defect Coupling of Spin Defects in Hexagonal Boron Nitride, npj Comput. Mater. 7, 59 (2021).arxiv.org/abs/2009.02830
  17. [47]R. Sundararaman, K. Letchworth-Weaver, K. A. Schwarz, D. Gunceler, Y. Ozhabes, and T. Arias, JDFTx: software for joint density-functional theory, SoftwareX 6, 278 (2017).DOI
  18. [48]J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77, 3865 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  19. [49]A. Marini, C. Hogan, M. Grüning, and D. Varsano, Yambo: An ab initio tool for excited state calculations, Comput. Phys. Commun. 180, 1392 (2009).DOI
  20. [50]F. Fuchs, J. Furthmüller, F. Bechstedt, M. Shishkin, and G. Kresse, Quasiparticle band structure based on a generalized kohn-sham scheme, Phys. Rev. B 76, 115109 (2007).DOI
  21. [51]F. Bechstedt, Many-Body Approach to Electronic Excitations (Springer-Verlag: Berlin, 2016).DOI
  22. [52]R. W. Godby and R. Needs, Metal-insulator transition in kohn-sham theory and quasiparticle theory, Phys. Rev. Lett. 22, 1169 (1989).pubmed.ncbi.nlm.nih.gov/10039594
  23. [53]A. Oschlies, R. Godby, and R. Needs, Gw self-energy calculations of carrier-induced band-gap narrowing in n-type silicon, Phys. Rev. B 51, 1527 (1995).pubmed.ncbi.nlm.nih.gov/9978867
  24. [54]C. A. Rozzi, D. Varsano, A. Marini, E. K. Gross, and A. Rubio, Exact coulomb cutoff technique for supercell calculations, Phys. Rev. B 73, 205119 (2006).DOI
  25. [55]A. Alkauskas, Q. Yan, and C. G. Van de Walle, First-principles theory of nonradiative carrier capture via multiphonon emission, Phys. Rev. B 90, 075202 (2014).DOI
  26. [56]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 J. Phys. 16, 073026 (2014).DOI
  27. [58]C. Guo, J. Xu, and Y. Ping, Substrate effect on excitonic shift and radiative lifetime of two-dimensional materials, J. Phys. Condens. Matter 33, 234001 (2021).DOI
  28. [59]C. Attaccalite, M. Bockstedte, A. Marini, A. Rubio, and L. Wirtz, Coupling of excitons and defect states in boron- nitride nanostructures. Phys. Rev. B 83, 144115 (2011).
  29. [60]T. J. Smart, F. Wu, M. Govoni, and Y. Ping, Fundamental principles for calculating charged defect ionization energies in ultrathin two-dimensional materials, Phys. Rev. Mater. 2, 124002 (2018).DOI
  30. [61]L. Novotny and B. Hecht, Principles of Nano-Optics (Cambridge university press, 2012).
  31. [62]J. A. Preuß, E. Rudi, J. Kern, R. Schmidt, R. Bratschitsch, and S. M. de Vasconcellos, Assembly of large hbn nanocrystal arrays for quantum light emission, 2D Mater. 8, 035005 (2021).DOI
  32. [63]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

Paper details

vlnc:W_2wq7gg4p
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