Negatively charged boron vacancy () spin defects are stable in nanoscale hexagonal boron nitride (hBN) flakes, which can be easily integrated into two-dimensional materials and devices to serve as both sensors and protective materials. Ion irradiation is frequently employed to create spin defects in hBN. However, the optimal ion irradiation parameters remain unclear, even though they play a crucial role in determining the depth and density of the defects, which in turn affect sensing sensitivity. In this work, we optimize the carbon ion irradiation parameters for creating spin defects by varying the irradiation dose and the incident angle. For 30 keV carbon ion irradiation, the optimal irradiation dose to create a ensemble is determined to be 4 × 1013 ions/cm2, and both continuous and pulsed optically detected magnetic resonance measurements are used to estimate the magnetic sensitivity and spin coherence properties. Moreover, the incident angle of energetic ions is found to influence both the depth and density distributions of the ensemble, a factor that is often overlooked. These results pave the way for improving the performance of quantum sensors based on hBN spin defects by optimizing the irradiation parameters.
ARTICLE HIGHLIGHTS
The optimal carbon ion irradiation dose is used to create spin defects in hBN.
ODMR measurements are used to estimate magnetic sensitivity and spin coherence properties.
The incident angle plays a key role in determining the depth of spin defects.
I. INTRODUCTION
An increasing number of luminescent defects, commonly known as color centers, have been observed in wide-bandgap materials. Among these, optically active spin defects are attracting growing attention owing to their advantages in quantum sensing, offering high spatial resolution and sensitivity at room temperature.1–10 In particular, negatively charged nitrogen vacancy color centers in diamond3–6 and negatively charged silicon vacancy and the neutral divacancy color centers in silicon carbide7–9 have opened up a wide range of applications in condensed matter physics, energy engineering, materials science,and biology. For instance, spin defects in diamond and silicon carbide can be used to determine the structure and dynamics of biomolecules3,6 and the magnetism of various materials at the nanoscale.5,9 However, it is challenging to fabricate near-surface spin defects in diamond and the associated quantum sensing devices at low cost.10,11
Recently, multiple color centers have been found in hexagonal boron nitride (hBN). Experimental and theoretical investigations indicate that vacancies, antisite defects, carbon- or oxygen-doped defects, or complexes of these point defects might serve as single-photon emitters or spin defects at room temperature.12–18 Currently, only negatively charged boron vacancy () color centers in hBN have been accurately characterized, which has been done by combining photoluminescence (PL) and optically detected magnetic resonance (ODMR) spectra with density functional theory (DFT) calculations.19–24 The ensemble of color centers possess spin S = 1 and display broad emission in the range of 700–1000 nm, which can be applied for the measurement of magnetic and electric fields, temperature, pressure, strain, radiofrequency signals, and paramagnetic spins in liquids at room temperature.13 Owing to the weak van der Waals interactions between adjacent hBN layers, hBN can be readily exfoliated into a desired thickness, and spin defects in atomically thin flakes are stable.25 This allows ultrathin hBN sensors based on color centers to be positioned close to the surface of a probed sample at the atomic scale, particularly facilitating integration into two-dimensional materials or devices. Nevertheless, the low photoluminescent quantum yield,23 the wide ODMR linewidth induced by strong hyperfine interaction between electron spins and nuclear spins,24 and the short spin coherence time32,33 of centers limit the sensitivity of such sensors.
To improve the sensing sensitivity of ensembles, plasmonic cavities have recently been employed to enhance photoluminescence (PL) intensity.13 Additionally, defects in isotopically purified h10B15N nanosheets exhibit well-resolved hyperfine structure and narrower ODMR linewidth.24 Advanced dynamical decoupling techniques have also been utilized to extend the coherence time of centers,32,33 which is essential for the detection of radiofrequency signals and coherence control of spin qubits.13 Besides, optimizing fabrication methods is a simpler and more direct strategy for enhancing the sensing sensitivity of ensembles. Initially, thermal neutron irradiation was employed to create centers;19 however, ion irradiation has now become the most commonly used method, owing to its high efficiency and low cost. For example, energetic ions such as hydrogen,26–29 helium,30–39 carbon,34,40–43 and nitrogen34,44–47 are frequently used to create centers for spin property research and quantum sensing applications. Nevertheless, the optimal irradiation dose required for different ions with varying energies, as well as the irradiation parameters such as the incident angle, need to be further investigated.
In this study, we examine the effects of ion irradiation parameters on the magnetic sensitivity and spin coherence time of spin defects in hBN. For 30 keV carbon ion irradiation, we obtain an optimal irradiation dose of 4 × 1013 ions/cm2. Notably, we discover that the incident angle of energetic ions plays a crucial role in influencing the depth of ensembles—this factor has typically been ignored in previous investigations. Continuous ODMR spectra reveal that the room-temperature magnetic sensitivity reaches 36.0 under a 6.28 mT bias magnetic field. Furthermore, the results of pulsed ODMR measurements indicate that the spin–lattice relaxation time (∼14 μs) and spin coherence time (∼100 ns) are relatively long. These results are essential for optimizing the irradiation process to enhance the spin properties of in hBN and for the development of nanoscale quantum sensors.
II. EXPERIMENTAL METHODS
A. Ion irradiation experiment
Bulk hBN with a thickness of several tens of micrometers was purchased from HQ Graphene. It was then carefully cleaved into several flakes, each exceeding 1 µm in thickness. These hBN flakes were subsequently adhered to silicon substrates by conductive tape for vertical implantation or affixed to a triangular silver frame for oblique implantation, as shown in Fig. 1(a). Carbon ion irradiation experiments were conducted using the 320 kV Platform for Multidisciplinary Research with Highly Charged Ions at the Institute of Modern Physics, CAS. The kinetic energy of carbon ions was fixed at 30 keV, and the corresponding irradiation dose rate was 1.43 × 1012 ions/(cm2 s). The depth distributions of carbon atoms and irradiation-induced vacancies, as simulated by the SRIM software package,48 are presented in Figs. 1(b) and 1(c). To create a moderate concentration of boron vacancy defects in hBN, the irradiation dose was set at 1 × 1013, 4 × 1014, and 1 × 1014 ions/cm2. Additionally, a carbon-implanted hBN flake with a dose of 1 × 1014 ions/cm2 at an angle of 30° was annealed for 1 h in an argon atmosphere at a temperature of 1297 K.
(a) An oblique angle of 30° is achieved using a triangular silver frame. (b) and (c) Depth distributions of carbon atoms and irradiation-induced vacancies, respectively, after 30 keV carbon ion irradiation, as simulated by SRIM.48
(a) An oblique angle of 30° is achieved using a triangular silver frame. (b) and (c) Depth distributions of carbon atoms and irradiation-induced vacancies, respectively, after 30 keV carbon ion irradiation, as simulated by SRIM.48
B. PL and ODMR experiments
After 30 keV carbon irradiation, the Raman and PL spectra of carbon-implanted hBN samples were measured using a confocal Raman spectrometer (Horiba iHR550) with a high-numerical-aperture (100×, NA = 0.9) microscope objective, under 532 nm laser excitation. The ODMR spectra of the carbon-implanted hBN flakes were obtained by sweeping the microwave frequency, and the total PL counts from 650 to 1000 nm were detected by an avalanche photodiode (APD). Our ODMR experiments were conducted using custom-built equipment from Guoyi Quantum (Hefei) Technology Co. Ltd. (Hefei, China), as shown in Fig. 2(a). A confocal microscope objective (60×, NA = 0.7) was employed to focus a 520 nm excitation laser onto the sample. The PL signal from the sample was separated from excitation light by a dichroic mirror, and the residual laser light was shielded by a 650 nm long-pass filter before arriving at the APD. To manipulate the spin states of the ensemble, the radiofrequency (RF) signal from the microwave source was amplified using a microwave amplifier, and then delivered to the hBN sample through a 20 μm copper wire. The hBN sample and copper wire were fixed on a printed circuit board, which was mounted on a piezoelectric nano-positioning stage. For the pulsed ODMR experiments, a multichannel arbitrary sequence generator (ASG) was used to control the excitation laser and RF signal, as well as the triggering of the APD. A permanent magnet mounted on a translation stage was used to provide an external magnetic field that could be tuned by adjusting the distance between the magnet and the hBN sample.
(a) Schematic of the main components of the ODMR equipment. A 520 nm laser and a microwave (MW) system were coordinated to manipulate and read the spin states of the ensemble in the hBN flake. The hBN flake (white) was placed on a printed circuit board (green). The radiofrequency (RF) signal from the microwave source was amplified using a microwave amplifier, and then delivered to the hBN sample through a 20 μm copper wire. For the pulsed ODMR experiments, a multichannel arbitrary sequence generator (ASG was used to control the excitation laser and RF signal, as well as the triggering of the avalanche photodiode (APD). (b) The upper part shows schematic representations of the atomic structure of defects, which possess a spin triplet (S = 1) with a quantization axis perpendicular to the hBN crystal, as indicated by the purple arrow. The lower part depicts the energy level structure of spin defects in the wide bandgap of hBN, with a measured zero-field splitting D of 3.464 GHz.
(a) Schematic of the main components of the ODMR equipment. A 520 nm laser and a microwave (MW) system were coordinated to manipulate and read the spin states of the ensemble in the hBN flake. The hBN flake (white) was placed on a printed circuit board (green). The radiofrequency (RF) signal from the microwave source was amplified using a microwave amplifier, and then delivered to the hBN sample through a 20 μm copper wire. For the pulsed ODMR experiments, a multichannel arbitrary sequence generator (ASG was used to control the excitation laser and RF signal, as well as the triggering of the avalanche photodiode (APD). (b) The upper part shows schematic representations of the atomic structure of defects, which possess a spin triplet (S = 1) with a quantization axis perpendicular to the hBN crystal, as indicated by the purple arrow. The lower part depicts the energy level structure of spin defects in the wide bandgap of hBN, with a measured zero-field splitting D of 3.464 GHz.
III. RESULTS AND DISCUSSION
A. PL and ODMR spectra of spin defects in carbon-irradiated hBN
Figure 2(b) illustrates the atomic structure and energy level structure of defects, which possess a spin triplet (S = 1) with a quantization axis that is perpendicular to the hBN crystal. The energy levels of defects are all located in the wide bandgap of hBN, and the defect levels with ground and excited states allow laser absorption and PL emission. Consequently, the origin of the PL spectra can be explained by the energy level structure of defects in hBN. Moreover, the defect levels with ground and excited states are all spin triplets with ms = 0 and ±1. There is zero-field splitting (ZFS) D between ms = 0 and ms = ±1 due to spin–spin interaction, which can be excited by microwaves, and there exists a tiny splitting E between ms = −1 and ms = +1, even in the absence of an external field, due to the local electric field surrounding charged defects.25,32,49 The PL intensity decreases under green laser excitation when the electrons around defects are excited by the resonant microwaves from ms = 0 to ms = ±1. This is because the electrons in the excited ms = ±1 states can relax to ground states by nonradiative transition through metastable states. As a result, we can record the ratio of the spin-dependent PL intensity change of defects under laser excitation as the microwave frequency is varied, thus obtaining ODMR spectra.
As shown in Fig. 3(a), the PL spectra of carbon-implanted hBN flakes indicate that only color centers can be produced directly. The characteristic PL spectra of the ensemble span from 700 to 1000 nm, with a peak at 818 nm. The PL intensity reaches its maximum when the implantation dose is in the range of 4 × 1013–1 × 1014 ions/cm2. Moreover, an oblique impact at 30° incident angle reduces the PL intensity when the implantation dose is fixed at 1 × 1014 ions/cm2; a schematic of the incident angle can be seen in Fig. 1(a). Monte Carlo simulations [Figs. 1(b) and 1(c)] indicate that implantation at 30° decreases the depth distribution of carbon atoms and vacancies by ∼14 nm, and thus oblique implantation may lead to excessive point defects and even defect clusters, which further decrease the density. A recent study32 has also found that only a small portion of the irradiation-induced boron vacancy defects are negatively charged state spin defects. Figure 3(b) presents Raman spectra of carbon-implanted hBN flakes, which exhibit two obvious peaks at 449 and 1295 cm−1 attributed to irradiation-induced defects, as well as the characteristic phonon mode E2g at 1366 cm−1.37,38
PL spectra (a) and Raman spectra (b) of bulk hBN flakes, subjected to 30 keV carbon ion irradiation with different doses and incident angles. All four samples were excited by a 532 nm laser with the same power.
PL spectra (a) and Raman spectra (b) of bulk hBN flakes, subjected to 30 keV carbon ion irradiation with different doses and incident angles. All four samples were excited by a 532 nm laser with the same power.
After 30 keV carbon ion implantation with a dose of 4 × 1013 ions/cm2, the total PL intensity of the ensemble can reach 5.5 × 106 photons/s under 520 nm laser excitation with a power of 5.6 mW, as shown in Fig. 4(a). The integral PL intensity I of spin defects increases with increasing laser power P, and the experimental data are well fitted by , where Psat is the saturation laser power and Isat is the saturation PL count rate.30
(a) Integral PL intensity in the range 650–1000 nm as a function of laser power. The saturation PL intensity and saturation laser power are 21.3 MHz and 16.1 mW, respectively. (b) ODMR spectra of ensemble in carbon-ion-irradiated hBN under different external magnetic fields. The solid lines are data fits with two Lorentzian peaks. (c) Magnetic field dependence of resonance frequencies due to the Zeeman effect corresponding to transitions from ms = 0 to ms = −1 and from ms = 0 to ms = +1. The solid lines were fitted by Eq. (1). The irradiation dose is 4 × 1013 ions/cm2, the power of the 520 nm excitation laser is 5.6 mW, and all experiments were carried out at 295 K.
(a) Integral PL intensity in the range 650–1000 nm as a function of laser power. The saturation PL intensity and saturation laser power are 21.3 MHz and 16.1 mW, respectively. (b) ODMR spectra of ensemble in carbon-ion-irradiated hBN under different external magnetic fields. The solid lines are data fits with two Lorentzian peaks. (c) Magnetic field dependence of resonance frequencies due to the Zeeman effect corresponding to transitions from ms = 0 to ms = −1 and from ms = 0 to ms = +1. The solid lines were fitted by Eq. (1). The irradiation dose is 4 × 1013 ions/cm2, the power of the 520 nm excitation laser is 5.6 mW, and all experiments were carried out at 295 K.
Figure 5(a) shows that the integral PL count of an hBN flake irradiated with a dose of 1 × 1014 ions/cm2 reaches 6.8 × 106 photons/s under 520 nm laser excitation with a power of 5.6 mW, slightly higher than the count for an hBN flake irradiated with a dose of 4 × 1013 ions/cm2. This is consistent with the PL spectra presented in Fig. 3(a). However, for the hBN flake irradiated with a dose of 1 × 1014 ions/cm2, the ODMR contrast is only 4.7% in the absence of an external magnetic field, and 2.5% under a 9.6 mT bias magnetic field, which are much lower than those for the hBN flake irradiated with a dose of 4× 1013 ions/cm2. As a result, the optimal dose for creating spin defects in hBN is 4 × 1013 ions/cm2 using 30 keV carbon ion irradiation. Additionally, the incident angle is crucial, since it influences the depth distribution and density of vacancy defects, as illustrated in Figs. 1(c) and 3(a).
(a) Integral PL intensity in the range 650–1000 nm as a function of trace time. (b) ODMR spectra of ensemble in carbon-ion-irradiated hBN flake. The irradiation dose is 1 × 1014 ions/cm2, the power of the 520 nm excitation laser is 5.6 mW, and all the experiments were carried out at 295 K.
(a) Integral PL intensity in the range 650–1000 nm as a function of trace time. (b) ODMR spectra of ensemble in carbon-ion-irradiated hBN flake. The irradiation dose is 1 × 1014 ions/cm2, the power of the 520 nm excitation laser is 5.6 mW, and all the experiments were carried out at 295 K.
B. Spin coherence properties of spin defects in carbon-irradiated hBN
We also conducted pulsed ODMR measurements to evaluate the spin coherence properties of spin defects in carbon-irradiated hBN with an optimal dose of 4 × 1013 ions/cm2. The Rabi oscillation shown in Fig. 6(a) demonstrates that the spin states of a ensemble can be coherently manipulated and read using an excitation laser and microwave pulses at room temperature. Moreover, this approach can also be used to evaluate the spin dephasing time by fitting the Rabi oscillation with , where A and B are fitting constants, fR is the Rabi frequency, τ is the length of the microwave pulse, and the fitting spin dephasing time is ∼30 ns.
Spin coherence properties of spin defects in carbon-irradiated hBN with an optimal dose of 4×1013 ions/cm2. (a) Rabi oscillation of spin defects. (b) and (c) Spin–lattice relaxation time T1 at ms = 0 and −1, respectively. (d) Spin–spin relaxation time T2. The power of the 520 nm excitation laser is 1.4 mW, and all the experiments were carried out at room temperature without an external magnetic field. The laser and microwave pulses are illustrated in the insets, where green, blue, and red rectangles represent the laser, microwave, and PL detection pulses, respectively.
Spin coherence properties of spin defects in carbon-irradiated hBN with an optimal dose of 4×1013 ions/cm2. (a) Rabi oscillation of spin defects. (b) and (c) Spin–lattice relaxation time T1 at ms = 0 and −1, respectively. (d) Spin–spin relaxation time T2. The power of the 520 nm excitation laser is 1.4 mW, and all the experiments were carried out at room temperature without an external magnetic field. The laser and microwave pulses are illustrated in the insets, where green, blue, and red rectangles represent the laser, microwave, and PL detection pulses, respectively.
We employed two kinds of pulses to estimate the spin–lattice relaxation time T1, as shown in Figs. 6(b) and 6(c). At room temperature, T1 is ∼14 μs, which is comparable to the value for hBN samples irradiated with thermal neutrons52,53 or other ions.32–34, Figure 6(d) shows that the spin coherence time of spin defects induced by 30 keV carbon implantation with the optimal dose is ∼102 ns, which is similar to previously measured values.32,33,43,52 These pulsed ODMR measurements are significant for enabling more complex sensing protocols, such as the detection of paramagnetic spins in liquids and RF signals.13
A comparison between different irradiation parameters and the spin properties of the ensemble in hBN flakes is presented in Table I. The fabrication methods and parameters significantly influence spin properties, including ODMR contrast, sensitivity, and coherence properties. It is noteworthy that microcavities can substantially enhance ODMR contrast and sensitivity,13 although this is not included in Table I. Additionally, for a given irradiated hBN sample, the ODMR contrast and sensitivity also depend on the microwave waveguide configuration, as well as the powers of the microwave and laser. For instance, for a ensemble in hBN flakes with a thickness of 100 nm created by 10 keV carbon ion implantation at a given dose, the ODMR contrast can be improved from 1.5% to 20% by substituting a 10 μm-wide gold waveguide by a gold coplanar waveguide with a width of 50 μm.40,43 Thermal neutrons and MeV electrons can penetrate through hBN flakes with a thickness of a few micrometers, producing an almost uniform density of defects. The depth distributions of defects in ion-irradiated hBN flakes are dependent on ion species, energy, and incident angle, as estimated by SRIM48 simulations.
Comparison of the depth, contrast, magnetic field sensitivity, spin–lattice relaxation time T1, spin-echo coherence time T2, and spin-dephasing time of VB– in hBN generated with various irradiation parameters.
Authors . | Material thickness . | Ions, dose . | Waveguide . | Depth (nm) . | Contrast . | Sensitivity () . | T1 (µs) . | T2 (ns) . | (ns) . |
---|---|---|---|---|---|---|---|---|---|
Fröch et al.26 | 290 nm | 15 keV H+, 5 × 1014 ions/cm2 | 20 μm copper wire | 250 | 3.6% | … | … | … | … |
Zhou et al.28 | 1 μm | 500 keV H+, 3 × 1016 ions/cm2 | 50 μm-wide gold stripline | <7 μm | ∼10% | 2.87 | … | … | … |
Hennessey et al.29 | 400 nm | 15 keV H+, 5 × 1015 ions/cm2 | … | 250 | ∼12% | … | … | … | … |
Gao et al.30 | 35 –180 μm | 200 eV–3 keV He+, ≤2 × 1014 ions/cm2 | 50 μm-wide gold stripline | 2.5–60 | 46% (4 W microwave) | 8 | 17 | 1100 | 120 |
Huang et al.31 | 98 nm | 5 keV He+, 5 × 1013 ions/cm2 | Gold microwave stripline | <100 | 10% | 8 pulsed | … | … | 105 |
Gong et al.32 | ∼100 nm | 3 keV He+, ≤1 × 1015 ions/cm2 | 50 μm-wide gold stripline | 60 | ∼4% | … | ≤13.3 | 70 | … |
Rizzato et al.33 | ∼100 nm | 3 keV He+, 3 × 1014 ions/cm2 | Gold stripline | 60 | ∼3% | ∼3 AC | ∼6 | ∼60 | … |
Guo et al.34 | 10–100 nm | 30 keV He+, ≤1015 ions/cm2 | 20 μm copper wire | <350 | ∼8% | … | ∼14 | … | … |
Sasaki et al.36 | 66 nm, 100 × 100 nm2 irradiated area | 30 keV He+, 1 × 1015 ions/cm2 | 20 μm copper wire, 3200 nm-wide gold wire | … | 20% | 73.6 | … | … | … |
Liang et al.39 | 200 nm | 500 keV He+, 3 × 1016 ions/cm2 | Copper waveguide | <1.8 μm | ∼7% | 2.55 (59.9 mW laser) | … | … | … |
Guo et al.34 | 10–100 nm | 30 keV C+, 1 × 1014 ions/cm2 | 20 μm copper wire | <150 | ∼4% | … | ∼11 | … | … |
Ramsay et al.43 | ∼100 nm | 10 keV C+, 1 × 1014 ions/cm2 | 10 μm-wide gold waveguide | 60 | 1.5% | … | 10 | <100 | <60 |
Baber et al.40 | ∼100 nm | 10 keV C+, 1 × 1014 ions/cm2 | Gold coplanar waveguide | 60 | 20% | … | 10 | … | 19 |
Guo et al.34 | 10–100 nm | 30 keV N+, 1 × 1014 ions/cm2 | 20 μm copper wire | <130 | ∼6% | … | ∼11 | … | … |
Healey et al.44 | 40 nm, 70 nm | 10, 15 keV N+, 1 × 1014 ions/cm2 | 50 μm copper wire loop | 50, 70 | ∼1% | ∼100, 2000 | 12 | … | … |
Haykal52 | ∼6 μm h10BN, h11BN | Thermal neutron, 2.6 × 1016 n/cm2, 2.6 × 1017 n/cm2 | Copper microwire | Whole sample | ∼3%, ∼4% | … | 16 | 62, 46 | ∼20 |
Gottscholl et al.53 | ∼1 μm | Thermal neutron, 2.3 × 1018 n/cm2 | 0.5 mm-wide copper stripline | Whole sample | ∼0.1% | … | 18 | 2000 | 100 |
Kumar et al.54 | Few tens of nm h10BN | Thermal neutron, 2.6 × 1016 n/cm2 | External loop antenna | Whole sample | ∼6% | 60, 110 (7 mT) | … | … | … |
Healey et al.55 | <1 μm hBN, powder-sized, 3–4 μm | 2 MeV electron, 5 × 1018 e/cm2 | Silver stripline (0.4 mm) | Whole sample | ∼10%, ∼8% | ∼1.7, ∼2.2 | ∼14, 20 | … | … |
This work | ∼1 μm thickness | 30 keV C+, 4 × 1013 ions/cm2 | 20 μm copper wire | 150 | 7.7% | 18.6 | 13.8 | 102 | ∼30 |
Authors . | Material thickness . | Ions, dose . | Waveguide . | Depth (nm) . | Contrast . | Sensitivity () . | T1 (µs) . | T2 (ns) . | (ns) . |
---|---|---|---|---|---|---|---|---|---|
Fröch et al.26 | 290 nm | 15 keV H+, 5 × 1014 ions/cm2 | 20 μm copper wire | 250 | 3.6% | … | … | … | … |
Zhou et al.28 | 1 μm | 500 keV H+, 3 × 1016 ions/cm2 | 50 μm-wide gold stripline | <7 μm | ∼10% | 2.87 | … | … | … |
Hennessey et al.29 | 400 nm | 15 keV H+, 5 × 1015 ions/cm2 | … | 250 | ∼12% | … | … | … | … |
Gao et al.30 | 35 –180 μm | 200 eV–3 keV He+, ≤2 × 1014 ions/cm2 | 50 μm-wide gold stripline | 2.5–60 | 46% (4 W microwave) | 8 | 17 | 1100 | 120 |
Huang et al.31 | 98 nm | 5 keV He+, 5 × 1013 ions/cm2 | Gold microwave stripline | <100 | 10% | 8 pulsed | … | … | 105 |
Gong et al.32 | ∼100 nm | 3 keV He+, ≤1 × 1015 ions/cm2 | 50 μm-wide gold stripline | 60 | ∼4% | … | ≤13.3 | 70 | … |
Rizzato et al.33 | ∼100 nm | 3 keV He+, 3 × 1014 ions/cm2 | Gold stripline | 60 | ∼3% | ∼3 AC | ∼6 | ∼60 | … |
Guo et al.34 | 10–100 nm | 30 keV He+, ≤1015 ions/cm2 | 20 μm copper wire | <350 | ∼8% | … | ∼14 | … | … |
Sasaki et al.36 | 66 nm, 100 × 100 nm2 irradiated area | 30 keV He+, 1 × 1015 ions/cm2 | 20 μm copper wire, 3200 nm-wide gold wire | … | 20% | 73.6 | … | … | … |
Liang et al.39 | 200 nm | 500 keV He+, 3 × 1016 ions/cm2 | Copper waveguide | <1.8 μm | ∼7% | 2.55 (59.9 mW laser) | … | … | … |
Guo et al.34 | 10–100 nm | 30 keV C+, 1 × 1014 ions/cm2 | 20 μm copper wire | <150 | ∼4% | … | ∼11 | … | … |
Ramsay et al.43 | ∼100 nm | 10 keV C+, 1 × 1014 ions/cm2 | 10 μm-wide gold waveguide | 60 | 1.5% | … | 10 | <100 | <60 |
Baber et al.40 | ∼100 nm | 10 keV C+, 1 × 1014 ions/cm2 | Gold coplanar waveguide | 60 | 20% | … | 10 | … | 19 |
Guo et al.34 | 10–100 nm | 30 keV N+, 1 × 1014 ions/cm2 | 20 μm copper wire | <130 | ∼6% | … | ∼11 | … | … |
Healey et al.44 | 40 nm, 70 nm | 10, 15 keV N+, 1 × 1014 ions/cm2 | 50 μm copper wire loop | 50, 70 | ∼1% | ∼100, 2000 | 12 | … | … |
Haykal52 | ∼6 μm h10BN, h11BN | Thermal neutron, 2.6 × 1016 n/cm2, 2.6 × 1017 n/cm2 | Copper microwire | Whole sample | ∼3%, ∼4% | … | 16 | 62, 46 | ∼20 |
Gottscholl et al.53 | ∼1 μm | Thermal neutron, 2.3 × 1018 n/cm2 | 0.5 mm-wide copper stripline | Whole sample | ∼0.1% | … | 18 | 2000 | 100 |
Kumar et al.54 | Few tens of nm h10BN | Thermal neutron, 2.6 × 1016 n/cm2 | External loop antenna | Whole sample | ∼6% | 60, 110 (7 mT) | … | … | … |
Healey et al.55 | <1 μm hBN, powder-sized, 3–4 μm | 2 MeV electron, 5 × 1018 e/cm2 | Silver stripline (0.4 mm) | Whole sample | ∼10%, ∼8% | ∼1.7, ∼2.2 | ∼14, 20 | … | … |
This work | ∼1 μm thickness | 30 keV C+, 4 × 1013 ions/cm2 | 20 μm copper wire | 150 | 7.7% | 18.6 | 13.8 | 102 | ∼30 |
The coherence time of defects is limited by the nuclear spin bath of the hBN crystal, and also is influenced by the irradiation parameters. A spin coherence time of 102 ns is consistent with previously reported values.32,33,43,52 Coherence times of the order of microseconds have also been observed;30,53 however, the contrast of the spin-echo signal in these studies was quite weak, and experimental data at shorter time scales were not presented.52 The external magnetic field and advanced dynamical decoupling sequences can be used to extend the coherence time of defects,32,33,43 although this is not included in Table I. According to Table I, the optimal irradiation parameters are crucial for enhancing the spin properties. Our optimal carbon ion parameters demonstrate a relatively good ODMR contrast and coherence time compared with other irradiation parameters when a similar microwave waveguide is used.
C. High-temperature annealing of carbon-irradiated hBN flake
To create carbon-related color centers in a carbon-irradiated hBN flake with a dose of 1 × 1014 ions/cm2 at a 30° incident angle, a thermal annealing experiment was performed at 1297 K in an argon atmosphere for 1 h. Figure 7 shows that the additional Raman peaks at 449 and 1295 cm−1 observed in Fig. 3(b), along with the characteristic PL spectra of the ensemble, have disappeared,37,38 which indicates that the most of the irradiation-induced defects were repaired after the thermal annealing. Besides, the annealed hBN flake exhibits a PL spectrum with a peak at ∼580 nm, which is very similar to recent experimental results.56 According to previous theoretical calculations, the PL signal with peak at ∼580 nm may originate from a carbon trimer or tetramer.57–59 In this work, we have focused on ensembles in hBN created by 30 keV carbon ion implantation.
(a) Raman and (b) PL spectra of carbon-implanted hBN after annealing at 1297 K for 1 h. The kinetic energy of carbon ions is 30 keV, and the irradiation dose is 1 × 1014 ions/cm2 at 30° incident angle.
(a) Raman and (b) PL spectra of carbon-implanted hBN after annealing at 1297 K for 1 h. The kinetic energy of carbon ions is 30 keV, and the irradiation dose is 1 × 1014 ions/cm2 at 30° incident angle.
IV. CONCLUSIONS
We have investigated the optimal ion irradiation parameters on the basis of PL intensity, ODMR contrast, magnetic sensitivity, and spin coherence time of spin defects in hBN. For 30 keV carbon ion irradiation, we have identified an optimal dose of 4 × 1013 ions/cm2. Importantly, we have found that the angle of energetic ions has a significant impact on the depth and density distributions of the ensemble—a factor often overlooked in earlier studies. Our continuous ODMR spectra have demonstrated a room-temperature magnetic sensitivity of 36.0 under a 6.28 mT bias magnetic field, and the pulsed ODMR results have shown that the spin–lattice relaxation time and spin coherence time are relatively long. Additionally, we have found that it is possible to generate carbon-related color centers in bulk hBN flakes by combining 30 keV carbon implantation with annealing at 1297 K for 1 h. These results are crucial for refining the irradiation process to enhance the spin properties of in hBN and for advancing the development of atomic-scale quantum sensors utilizing this van der Waals semiconductor.
ACKNOWLEDGMENTS
This work was supported by the National Key Research and Development Program Project (2024YFF0726104), Key Laboratory of Modern Optical Technologies of the Education Ministry of China, Soochow University (Grant No. KJS2135), a China Postdoctoral Science Foundation Funded Project (Grant No. 2024M751236), and the Jiangxi Provincial Natural Science Foundation (Grant No. 20232BAB211030).
AUTHOR DECLARATIONS
Conflict of Interest
The authors have no conflicts to disclose.
DATA AVAILABILITY
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
Fei Ren received his Master’s degree in Nuclear Technology and Applications from Lanzhou University in 2018. He is currently pursuing his doctorate at the State Key Laboratory of Precision Measuring Technology and Instruments, within the School of Precision Instrument and Opto-Electronics Engineering at Tianjin University, China. His research interests include quantum sensing based on spin defects in hexagonal boron nitride, and atomistic simulations of point defects in semiconductors.
Zongwei Xu is a Professor at Tianjin University and a Doctoral Supervisor. His research interests include ultrafast energy beam (ion and laser) processing, Raman and photoluminescence spectroscopy characterization, wide-bandgap semiconductor devices, microcutting tools, and nanocutting technology. He is the Chairman of the first Sino-German Symposium on Defect Engineering in SiC Device Manufacturing–Atomistic Simulations, Characterization and Processing and the Principal Investigator of the 2021–2023 Mobility Programme of the Sino-German Center for Research Promotion (M-0396).
Yiyuan Wu is a Lecturer and Master’s Supervisor at the East China University of Technology. He obtained his doctoral degree in Particle Physics and Nuclear Physics from Lanzhou University in 2020. His research interests include the development of novel semiconductor neutron detectors, and radiation effects on nuclear materials and semiconductors.