Mesoscopic Spin Coherence in a Disordered Dark Electron Spin Ensemble

Harnessing dipolar spin environments as controllable quantum resources is a central challenge in solid-state quantum technologies. Here, we report the observation of a coherent mesoscopic spin state in a disordered ensemble of substitutional nitrogen (P1) centers in diamond. An iterative Hartmann-Hahn protocol transfers polarization from dense nitrogen-vacancy (NV) centers to a P1 ensemble, yielding a 740-fold enhancement over room-temperature thermal equilibrium as revealed by differential readout. The resulting mesoscopic P1 spin ensemble exhibits collective Rabi oscillations and long-lived spin-lock and Hahn-echo coherences. We identify a crossover in the saturation polarization arising from the competition between coherent driving and local disorder, providing a quantitative measure of the system’s intrinsic disorder. These results establish a foundation for utilizing dark electron spin ensembles as robust resources for quantum sensing and quantum many-body simulation.

💡 Research Summary

In this work the authors demonstrate that a dense ensemble of optically dark substitutional‑nitrogen (P1) electron spins in diamond can be transformed into a coherent mesoscopic quantum state by repeatedly transferring polarization from a dense network of nitrogen‑vacancy (NV) centers. The sample contains ~6.3 ppm P1 and ~2.4 ppm NV, corresponding to roughly 10⁵ spins within the optical detection volume. A static magnetic field of 446 G aligns the NV axis and separates the |0⟩↔|−1⟩ transition, allowing the NV electronic spin to be treated as an effective spin‑½ system. Because the Zeeman splittings of NV and P1 differ by several hundred MHz, the dipolar interaction reduces to a secular Ising term. To bridge this energy mismatch the authors employ spin‑locking (SL) under the Hartmann‑Hahn (HH) condition, i.e., they drive both species at equal Rabi frequencies Ω_NV = Ω_P1 = Ω. In the rotating frame the Ising coupling becomes a flip‑flop term, enabling resonant exchange of polarization.

The experimental protocol consists of many cycles. Each cycle begins with a 532 nm laser pulse that optically polarizes the NV ensemble (≈75 % polarization). This is followed by a 5 µs double‑spin‑locking (DSL) period where both NV and a selected P1 subgroup (the 3/12 hyperfine/Jahn‑Teller component) are driven at the HH condition, allowing polarization to flow from NV to P1. After the DSL block a second laser pulse re‑polarizes the NVs, and the cycle repeats. By varying the number of cycles N the authors monitor the buildup of P1 polarization using a differential readout scheme: they prepare the P1 bath either parallel (↑↑) or antiparallel (↑↓) to the NV spin, perform a final DSL readout, and compute ΔC = C↑↑ − C↑↓, which directly reflects the net P1 polarization.

After 32 cycles the differential signal saturates at A ≈ 0.143. Using angular‑momentum conservation in the rotating frame and the known NV‑to‑P1 density ratio (≈2.6), the authors infer a P1 polarization of ≈7.4 %, i.e. a 740‑fold enhancement over the thermal equilibrium value (~0.01 % at room temperature). This corresponds to an effective spin temperature of ~405 mK. The rapid approach to saturation (characteristic cycle number N_sat ≈ 3, i.e. ~30 µs) is much faster than the measured rotating‑frame relaxation time T₁ρ^P1 ≈ 0.43 ms, indicating that additional loss channels are active during the laser illumination. A semiclassical master‑equation model attributes this to photo‑ionization of P1 centers and charge‑state dynamics of nearby NVs, yielding an effective T₁ρ^laser ≈ 32 µs.

To verify that the polarized P1 ensemble exhibits genuine collective coherence, the authors perform a second HH drive after the polarization cycles and observe collective Rabi oscillations in the differential signal. Fourier analysis shows a clear frequency component for the polarized state (N = 16) that is absent for the unpolarized thermal state (N = 0), confirming that the ensemble behaves as a coherent macroscopic spin rather than a set of independent fluctuators. The rotating‑frame relaxation time T₁ρ^P1 (0.43 ms) is comparable to the longitudinal relaxation time T₁^P1 (0.67 ms) measured in the dark, indicating that the decay is dominated by intrinsic dipolar processes rather than rapid diffusion of polarization into the bulk. Hahn‑echo measurements yield a dephasing time T₂^P1 ≈ 4.1 µs, roughly a factor of 2.6 shorter than the NV T₂ (≈10.7 µs), consistent with the relative defect densities and suggesting that inter‑group Ising couplings are refocused while intra‑group flip‑flop interactions dominate decoherence.

A key part of the study is the exploration of the competition between coherent driving strength Ω and the intrinsic disorder scale W of the spin network. By measuring the saturation amplitude A_sat as a function of Ω, the authors find that A_sat follows a Lorentzian‑like dependence A_sat(Ω) = A_∞ Ω²/(Ω² + W²). Fitting yields W ≈ 1.36 MHz, which quantifies the effective inhomogeneous broadening arising from the random spatial distribution of P1 centers. In the weak‑driving regime (Ω < W) energy mismatches suppress flip‑flop interactions, limiting polarization transfer; in the strong‑driving regime (Ω > W) the disorder is effectively averaged out, enabling efficient transport of polarization across the ensemble.

Numerical simulations of spin diffusion in the disordered network give a diffusion coefficient D ≈ 0.22 nm²/µs, leading to a diffusion length L_D ≈ 6.2 nm over the saturation timescale. Although the mean nearest‑neighbor distance for the addressed NV subgroup (~11.7 nm) exceeds L_D, Poisson statistics indicate that about 57 % of NVs have at least one P1 within the diffusion radius, providing sufficient connectivity for the observed polarization buildup.

Overall, the paper establishes three major advances: (1) a scalable protocol for polarizing a dense, optically dark spin bath at room temperature, achieving >700‑fold enhancement; (2) direct observation of collective coherent dynamics (Rabi oscillations, long T₁ρ and T₂) in the polarized bath, confirming its utility as a mesoscopic quantum resource; and (3) quantitative characterization of the disorder‑driven crossover that governs the efficiency of polarization transfer, providing a clear metric (W ≈ 1.36 MHz) for the intrinsic disorder of the system. These results open pathways to harness dark spin ensembles for quantum sensing, quantum memories, and analog quantum simulation of many‑body physics in solid‑state platforms.