Ten-second electron-spin coherence in isotopically engineered diamond
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | npj Quantum Information |
| Published | 2026-09-28 |
| Authors | T. Yamamoto, Hendrik Benjamin van Ommen, Kai-Niklas Schymik, Beer de Zoeten, Shinobu Onoda |
| DOI | 10.1038/s41534-026-01368-y |
| Citations | 2 |
A single nitrogen-vacancy (NV) centre in (111)-oriented, isotopically engineered diamond reached a dynamical decoupling coherence time of 11.2(8) s at 4 K. The measurement used 24,000 decoupling pulses with timing chosen to avoid mains-frequency noise. The same Fukuoka device reached a Hahn echo time of 6.8(1) ms after synchronized feedforward. Its homogeneous optical linewidth was 16.9(4) MHz. Yamamoto and colleagues report the spin result as the longest coherence time yet measured for a single solid-state electron spin.
Noise control raised Hahn echo coherence
Section titled “Noise control raised Hahn echo coherence”At the lowest isotope concentration, χ = 0.0013%, measured Ramsey times clustered near 300 μs. A simple fit gave an apparent (T_2^*) of about 383 μs. The authors found that these values were shorter than the expected nuclear-spin-bath limit. A Hahn echo was limited to about 2 ms, although the 13C bath alone could permit about 800 ms. The discrepancy led them to model 50 Hz magnetic interference from the electrical grid.
The authors synchronized each Hahn echo to a positive zero crossing of the 50 Hz waveform. They estimated the induced phase from X and Y signals acquired in 50 shots. They then shifted the final (\pi/2) pulse to compensate. This feedforward procedure extended (T_2^{\mathrm{Hahn}}) to 6.8(1) ms. The paper reports this as the longest Hahn echo time for a defect-centre electron-spin qubit. It exceeds the 1.5 to 2.5 ms range cited for several other defect-centre studies. Field-amplitude fluctuations over several seconds caused the remaining decay and spread.
The (111) growth process controlled isotopes and strain
Section titled “The (111) growth process controlled isotopes and strain”The authors used microwave-plasma chemical vapour deposition to grow homoepitaxial layers on low-dislocation (111) HPHT type-IIa substrates. Growth occurred at 120 Torr and (915 \pm 25,^\circ\mathrm{C}), with microwave power below 500 W. The CH4/H2 ratio was 0.03% for Fukuoka, 0.08% for Kagawa, and 0.01 to 0.08% for LD3. These low ratios supported step-flow growth at rates below 0.7 μm/h and suppressed secondary nucleation.
Independent mass-flow controllers mixed natural-abundance and 12C-enriched methane, setting the 13C concentration. SIMS gave 1.0937(86)% for natural methane and 0.0013(3)% for enriched methane. The latter equals 13(3) ppm. The calibrated flow relation covered 0.0139 to 1.0456% 13C. Isotope concentration sets both the nuclear-spin-bath strength and the number of 13C spins available as memory qubits. The (111) orientation supports preferential NV alignment along the surface-normal [111] axis, easier external-field alignment, and improved single-mode-fibre optical detection.
Nitrogen estimates separated SIMS limits from process bounds
Section titled “Nitrogen estimates separated SIMS limits from process bounds”SIMS placed hydrogen below 200 ppb, nitrogen below 20 ppb, and boron below 10 ppb. The present samples had roughly two orders of magnitude less mass than the tens-of-milligrams bulk samples used for sub-ppb nitrogen ESR. The authors therefore estimated solid nitrogen from reactor leakage, source-gas impurities, and an incorporation efficiency fitted with doped samples.
| Sample | 13C concentration | Nitrogen result | NV-centre preparation |
|---|---|---|---|
| Fukuoka | 0.0013(3)% | 1 to 4 ppb lower estimate; below 26 ppb upper estimate | As-grown centres, preferentially aligned along [111] |
| Kagawa | 0.0442%, 1.0937%, 0.1949%, and 0.0013% in four approximately 10 μm layers | 0.1 to 0.5 ppb lower estimate; below 3.2 ppb upper estimate | 2 MeV electrons at (5 \times 10^{11}\ \mathrm{cm}^{-2}), followed by 1000 °C annealing for 2 h in vacuum |
| LD3 | Multiple calibration layers | 0.2 to 2.3 ppb lower model; 2.9 to 20 ppb upper model | Isotope calibration only |
Kagawa had no detectable naturally occurring NV centres. The coherence measurements placed an upper bound of about 21 ppb on the electron-spin-bath concentration. These nitrogen results are process-derived bounds, not direct concentration measurements.
Dynamical decoupling exposed a change in the noise limit
Section titled “Dynamical decoupling exposed a change in the noise limit”CPMG data through 32 pulses fitted a mains model containing 50 Hz harmonics through 450 Hz. The 50 Hz component had a field amplitude of 2.95(3) mG and shifted the NV frequency by 8.28(8) kHz. The fit reproduced collapses and revivals, confirming mains interference as the dominant noise at low pulse counts.
For long sequences, the authors chose interpulse delays so the total decoupling time was a multiple of 20 ms. These revival times suppress most mains harmonics. With 24,000 pulses, the device reached 11.2(8) s. Ignoring revivals produced (T_2^{\mathrm{DD}}=T_0N^\eta) with (\eta\approx1) and (T_0=1.8(1)) ms. Sampling revivals gave (\eta=0.67(2)) and (T_0=16(2)) ms. The change occurred near 1,000 pulses, where the 450 Hz harmonic became relevant. If the two-thirds scaling represents nitrogen-limited coherence, the authors extrapolate a 16 ms Hahn echo and about 10 ppb nitrogen. That nitrogen value remains an inference.
Near-lifetime optical coherence survived, but charge noise set the limit
Section titled “Near-lifetime optical coherence survived, but charge noise set the limit”PLE measurements used a 515 nm repump at 20 μW for 50 μs. The same NV showed a transverse electric splitting of 2.1 GHz, indicating low lateral strain. In check-probe spectroscopy, a resonant check pulse prepared the charge state and frequency. A detuned probe then measured the (E_y) response. Both pulses lasted 400 μs at 1 nW, and each probe frequency used 50,000 repetitions.
Without a heralding threshold, the inhomogeneous linewidth was 222(3) MHz after 515 nm repumping. Thresholding at 30 counts revealed a homogeneous linewidth of 16.9(4) MHz. This was close to the approximately 13 MHz lifetime limit. The check success probability was 1.4%, although charge-resonance checks supported 92.5% single-shot spin readout.
Two check blocks bracketed a laser-driven diffusion period. Each check block applied (E_y) and (E_{\pm1}) light for 100 μs at 5 nW. An Ornstein-Uhlenbeck model fitted the backward-time count correlations with reduced chi-squared of 2.1. The diffusion timescale scaled as (P_d^{-1.05(4)}), and the model predicted (\tau_c>0.8) ms at 15 nW. No significant diffusion appeared within one second without applied laser light.
Under 637 nm diffusion excitation, the fitted inhomogeneous linewidth was 117(2) MHz. The comparison NV in non-isotopically purified (100) diamond had 42(9) MHz and a diffusion parameter about 30 times lower. The (111) sample was therefore stable but less stable than that reference under red excitation. Diffusion was also about an order of magnitude faster under 515 nm at the same power. The authors suggest vacancy and vacancy-complex photoionization, or hydrogen-related defects, at 637 nm. They associate faster 515 nm diffusion with remaining nitrogen-related defects. Identifying those defects requires broader frequency scans or direct spin probes.
Remaining limits guide material and control choices
Section titled “Remaining limits guide material and control choices”The 50 Hz field came from both drive currents and ambient laboratory fields. Inner and outer DC blocks improved coherence slightly but did not extend it beyond 3 ms. The authors identify magnetic shielding, relocation of noise sources, and active field stabilization as requirements for further gains.
Deeper isotope purification has a limited benefit once coherence is no longer controlled by the 13C bath. Removing more 13C reduces available memory qubits without a corresponding coherence gain. The detailed CPMG and optical analysis used one SIL NV, so its diffusion values do not describe a device-to-device distribution. The 1.4% check success probability also makes charge-resonance preparation relevant to repeated remote-entanglement attempts. Reducing diffusion could permit more repetitions before re-preparation and raise the overall entanglement rate.
Materials and support
Section titled “Materials and support”For NV-centre spin-photon devices, the paper needs single-crystal MPCVD diamond with a (111) surface, controlled 13C, low nitrogen, and gold striplines. The 6CCVD diamond materials catalog lists MPCVD-grown single-crystal material from 0.1 to 500 μm thick and plates or wafers up to 125 mm, with Au metallization. It does not specify (111) orientation or the paper’s isotope specification, so engineering support should confirm those requirements before a catalog plate is treated as suitable.
View Original Abstract
Abstract Solid-state spin defects are a promising platform for quantum networks. A key requirement is to combine long ground-state spin-coherence times with a coherent optical transition for spin-photon entanglement. Here, we investigate the spin and optical coherence of single nitrogen-vacancy (NV) centres in (111)-grown isotopically engineered diamond. Our diamond-growth process yields a precisely controlled 13 C concentration and low-ppb nitrogen concentrations. Combined with the mitigation of 50 Hz noise using a real-time feedforward scheme and tailored decoupling sequences, this enables record defect-electron-spin coherence times of T 2 = 6.8(1) ms for a Hahn echo and of $${T}_{2}^{DD}=11.2(8)$$ T 2 D D = 11.2 ( 8 ) s under dynamical decoupling. In addition, we observe coherent optical transitions with a near-lifetime-limited homogeneous linewidth of 16.9(4) MHz (222(3) MHz inhomogeneous broadening under 515nm illumination) and characterize the spectral diffusion dynamics. These results provide new avenues to investigate the incorporation of impurities in diamond and new opportunities for improved spin-qubit control for quantum networks and other quantum technologies.
Boron Doped Single Crystal Diamond Plate
View Specs →
Boron Doped Diamond Plate
View Specs →