Exploring Overlapping Mechanisms of Dynamic Nuclear Polarization in Type 1b HPHT Diamond

TL;DR

The study shows that in type 1b HPHT diamond with inhomogeneous P1 centers, multiple DNP mechanisms (solid effect, cross effect, and truncated cross effect) compete within the same crystal, yielding complex spectra whose signs can depend on buildup time and crystal orientation. By applying monochromatic and frequency-modulated (chirped) MW excitation to single-crystal and powder samples at 3.34 T and 7.05 T, the authors demonstrate selective enhancement or suppression of specific DNP pathways, including up to sixfold modulation enhancements and sign inversions in powder spectra. EPR analyses reveal different extents of electron spectral diffusion and exchange coupling between powder and single-crystal samples, underpinning the modulation-dependent behavior. Overall, frequency-modulated DNP emerges as a powerful tool to control polarization transfer mechanisms and to probe electron spin dynamics in diamond, with implications for optimizing $^{13}$C NMR hyperpolarization in heterogeneous spin systems.

Abstract

The inhomogeneous distribution of P1 centers in type 1b HPHT diamond samples allows multiple DNP mechanisms to occur within the same crystal, resulting in complex DNP spectra. At some crystal orientations, different DNP mechanisms can compete to drive hyperpolarization with different signs at the same applied microwave frequency. We perform microwave-irradiated DNP using both monochromatic and frequency-modulated microwave excitation to explore the competition between these DNP mechanisms in diamond at room temperature. We demonstrate that frequency-modulated DNP is a tool for suppressing certain DNP mechanisms while enhancing others in a single-crystal diamond sample. Frequency modulation also enables higher enhancement of the NMR signal beyond traditional monochromatic DNP under some conditions. In a powder sample, competing enhancement mechanisms can also arise from different crystallite orientations in the powder. We observe that at certain microwave frequencies the DNP signal changes sign during the polarization build-up, even with monochromatic microwave irradiation. We do not observe this phenomenon in any single-crystal spectrum. We discuss both methods of investigating competing mechanisms of DNP as a means of selectively enhancing different DNP mechanisms driving $^{13}$C NMR signal enhancement.

Figures (8)

• Figure 1: DNP spectrum and spectral decomposition for two orientations of a single-crystal diamond sample under monochromatic MW irradiation. (a) Pulse sequences used in this work for monochromatic MW irradiation (top) and frequency-modulated (chirped) MW irradiation (bottom). The chirp can be described by the central frequency of the chirp ($\omega_{MW}$), the modulation amplitude $\Delta\omega$, and the modulation frequency (slope) $f_m$. For all experiments a saturation train of 30 $\uppi/2$ pulses was used, with a saturation delay between each pulse in the train of $\Delta = 50~\upmu$s. (b) Orientation 1 shows a spectrum containing multiple frequencies with competition between the positive and negative enhancements provided by different DNP mechanisms. This spectrum has a MW buildup time of 100 s. (c) Orientation 2 shows multiple DNP mechanisms summing constructively to generate large DNP enhancement, though few MW frequencies include competition in the sign of the signal. Here, the MW buildup time was 200 s. The top line on each panel is the EPR line, simulated with EasySpin stollEasySpinComprehensiveSoftware2006, with the largest intensity peak normalized to one.
• Figure 2: DNP spectra using frequency- and amplitude-modulated microwaves, measured on Orientation 1. (a) A waterfall plot of the DNP spectrum using modulated microwaves. Each trace was measured using a modulation amplitude of $\Delta\omega = 150.75$ MHz and a different value of $f_m$ (colorbar). (b) A waterfall plot at constant $f_m$ (500 Hz) and varying $\Delta\omega$ (colorbar). In both panels each trace was measured using a MW buildup time of 100 s.
• Figure 3: Colormaps of DNP signal enhancement for varying modulation frequency and amplitude for Orientation 1. Panels (a), (b), and (c) reproduce the DNP spectrum from Figure \ref{['fig:crystaldecomp']}(b) with central MW irradiation frequency shown in solid pink. The dashed pink lines represent the edges of MW irradiation when the modulation amplitude $\Delta\omega$ equals the $^{13}$C Larmor frequency $\omega_C$ (35.8 MHz). Panels (d), (e), and (f) show the corresponding colormap of modulation frequency and amplitude for the central frequency shown in (a), (b), and (c), respectively. In each colormap the two-timescale $1/T_{1e}$ relaxation rates, measured at 2.5 GHz, are drawn as vertical dashed black lines.
• Figure 4: Colormaps of DNP signal enhancement for varying modulation frequency and amplitude for Orientation 2. Panels (a), (b), and (c) reproduce the DNP spectrum from Figure \ref{['fig:crystaldecomp']}(a) with central MW irradiation frequency shown in solid pink. The dashed pink lines represent the edges of MW irradiation when the modulation amplitude $\Delta\omega$ equals the $^{13}$C Larmor frequency $\omega_C$ (35.8 MHz). Panels (d), (e), and (f) show the corresponding colormap of modulation frequency and amplitude for the central frequency shown in (a), (b), and (c), respectively. In each colormap the two-timescale $1/T_{1e}$ relaxation rates, measured at 2.5 GHz, are drawn as vertical dashed black lines.
• Figure 5: X-band EPR spectra of powder and single-crystal diamond at room temperature. (a) Twice-integrated EPR signal as a function of microwave power for the powder sample (black) and single-crystal (blue). Lines are fits of the data to a saturation curve. The spectra measured at lowest power, shown with an "x", for each sample (0.034 $\upmu$W) are shown in panels (b) (powder) and (c) (single-crystal). (b) Integrated EPR spectrum of the powder sample at X-band shows a typical powder pattern of the P1 center albeit with a broad background visible upon integration. (c) Integrated EPR spectrum of the single-crystal diamond P1 center shows clear hyperfine satellites, but without the broad background feature. Simulations were performed with EasySpin stollEasySpinComprehensiveSoftware2006. In (b) and (c) the largest intensity peak was normalized to one.