Indirect Overgrowth as a Synthesis Route for Superior Diamond Nano Sensors

Quantum metrology using negatively charged nitrogen-vacancy (NV⁻) centers in diamond has gained significant traction in diverse fields, including information science, materials science, and life science. The NV⁻ center, composed of a substitutional nitrogen atom, a carbon vacancy, and an additional donor electron, is highly sensitive to various physical quantities like temperature, strain, electric and magnetic fields. Its ability to maintain coherence times up to milliseconds at room temperature makes it a promising candidate for quantum sensing. For high sensitivity in applications such as magnetic field sensing, the NV⁻ center needs to be positioned only a few nanometers below the diamond's surface. However, these shallow NV⁻ centers suffer from decoherence and charge instability due to surface defects. Increasing the depth of the NV⁻ centers mitigates this surface influence, resulting in longer coherence times. This paper focuses on optimizing the fabrication process of shallow NV⁻ centers to achieve a balance between sensitivity and coherence time, targeting a depth range of 10-30nm (the intermediate regime). Tailoring the NV⁻ center depth distribution and enhancing coherence times through increased average depth is crucial for many quantum sensing applications.

Previous research has demonstrated that low-energy implantation of NV⁻ centers followed by CVD diamond overgrowth enhances coherence times. However, the mechanisms behind significant NV⁻ center losses during overgrowth, such as etching and/or passivation by the hydrogen plasma used in CVD, remained unclear. A likely passivation mechanism is the conversion of NV⁻ to non-fluorescing NVH centers by reaction with hydrogen atoms from the plasma. While hydrogen diffusion in bulk diamond is well-understood, the dynamics of this process in shallow implanted layers, and the impact on NV⁻ center formation, needed further investigation. Confocal microscopy, offering optical detection of NV⁻ centers even in thin layers, is used here to overcome the limitations of electron-paramagnetic resonance (EPR) spectroscopy for studying shallow implanted layers.

The study compares two overgrowth methods: direct overgrowth (implantation, annealing, overgrowth) and indirect overgrowth (implantation, overgrowth, annealing). They used 15N nitrogen ions implanted at energies of 2.5 and 5 keV, with doses of 10⁹, 10¹¹, and 10¹² 15N+/cm² (low, medium, and high dose). Confocal microscopy with green laser excitation (519 nm) and pulsed optically detected magnetic resonance (pulsed ODMR) was used to characterize NV⁻ centers. The depth of NV⁻ centers was determined by analyzing the nuclear magnetic resonance (NMR) signal of H-spins in the immersion oil using the NV⁻ center as a sensor. The diamond growth rate was calibrated using the NV⁻ centers. The impact of etching was assessed by comparing depth distributions before and after overgrowth. The yield of NV⁻ centers after overgrowth and annealing was analyzed to study passivation kinetics as a function of implantation energy and dose. The study employed commercially available electronic-grade (100)-diamond substrates, a home-built CVD reactor, a low-energy ion implanter, and a home-built scanning confocal microscopy setup for experiments.

Direct overgrowth resulted in complete loss of NV⁻ centers, suggesting significant etching and/or passivation. Indirect overgrowth proved far more successful, yielding NV⁻ centers for medium and high doses. The depth distribution of NV⁻ centers after indirect overgrowth was uniformly shifted to deeper values, confirming successful burying below a diamond capping layer without significant etching. The average depth increased linearly with overgrowth time, with growth rates of 5 and 8 nm/h for 2.5 and 5 keV implantations, respectively. Coherence time (T₂) increased significantly with depth and with overgrowth, reaching up to 100 µs for T₂ and 20 µs for T₂* for 13nm overgrowth. The loss of NV⁻ centers during overgrowth was attributed to hydrogen passivation, forming NVH centers. The passivation rate decreased significantly with increasing implantation dose, indicating that implanted nitrogen hinders hydrogen diffusion. A two-step passivation mechanism was proposed: NV⁻ center formation followed by passivation. Higher nitrogen doses slowed down the NV⁻ to NVH conversion kinetics, suggesting nitrogen affects hydrogen diffusion. This nitrogen-dependent hydrogen mobility was hypothesized to be due to the nitrogen dopants altering the hydrogen's charge state and thereby reducing its mobility.

The results demonstrate that indirect overgrowth effectively addresses the trade-off between sensitivity and coherence time in shallow NV⁻ centers. By using low-energy implantation and precisely controlling the overgrowth thickness, this method allows for the creation of shallow NV⁻ centers with significantly longer coherence times compared to pure implantation. The observed dependence of passivation kinetics on the nitrogen dose provides valuable insights into the sub-surface diffusion of hydrogen in diamond during CVD growth, a critical factor for optimizing NV⁻ center fabrication. The finding that the implanted nitrogen impedes hydrogen diffusion suggests a possible charge-state modification of the hydrogen by the nitrogen dopants. This is a novel observation that could have significant implications for further research in controlling defect formation in diamond.

Indirect overgrowth offers a powerful method for fabricating high-quality shallow NV⁻ centers with enhanced coherence times and depth control. The method combines the advantages of low-energy implantation (narrow depth distribution, reduced implantation damage) with the ability to precisely tune the depth and improve coherence through CVD overgrowth. Future research should focus on further investigation of the hydrogen passivation mechanism and its dependence on various parameters, including hydrogen concentration, growth temperature and other dopants, to further optimize the fabrication process and to explore its implications for other color centers in diamond.

While the study provides strong evidence for hydrogen passivation, direct identification of NVH centers in the shallow implanted layer was not achieved. Further investigation using techniques sensitive to the NVH center could provide additional confirmation. The study focused on specific implantation energies and doses; further investigation exploring a wider range of parameters is needed for complete optimization. The precise mechanism through which nitrogen affects hydrogen diffusion needs additional study, perhaps through computational modeling or other experimental techniques.