Magnetic resonance tracking of fluorescent nanodiamond fabrication

Magnetic resonance techniques (electron paramagnetic resonance (EPR) and nuclear magnetic resonance (NMR)) are used for tracking the multi-stage process of the fabrication of fluorescent nanodiamonds (NDs) produced by high-energy electron irradiation, annealing, and subsequent nano-milling. Pristine commercial high pressure and high temperature microdiamonds (MDs) with mean size 150 μm contain ∼5 × 10¹⁸ spins/g of singlet (S = 1/2) substitutional nitrogen defects P1, as well as sp³ C-C dangling bonds in the crystalline lattice. The half-field X-band EPR clearly shows (by the appearance of the intense 'forbidden' g = 4.26 line) that high-energy electron irradiation and annealing of MDs induce a large amount (∼5 × 10¹⁷ spins/g) of triplet (S = 1) magnetic centers, which are identified as negatively charged nitrogen vacancy defects (NV^-). This is supported by EPR observations of the 'allowed' transitions between Zeeman sublevels of the triplet state. After progressive milling of the fluorescent MDs down to an ultrasubmicron scale (≤100 nm), the relative abundance of EPR active NV^- defects in the resulting fluorescent NDs (FND) substantially decreases and, vice versa, the content of C-inherited singlet defects correlatively increases. In the fraction of the finest FNDs (mean particle size <20 nm), which are contained in the dried supernatant of ultracentrifuged aqueous dispersion of FNDs, the NV^- content is found to be reduced by one order of magnitude whereas the singlet defects content increases up to ∼2 × 10¹⁹ spins/g. In addition, another triplet-type defect, which is characterized by the g = 4.00 'forbidden' line, appears. On reduction of the particle size below the 20 nm limit, the 'allowed' EPR lines become practically unobservable, whereas the 'forbidden' lines remain as a reliable fingerprint of the presence of NV^- centers in small ND systems. The same size reduction causes the disappearance of the characteristic hyperfine satellites in the spectra of the P1 centers. We discuss the mechanisms that cause both the strong reduction of the peak intensity of the 'allowed' lines in EPR spectra of triplet defects and the transformation of the P1 spectra.