Quantum sensing with a spin ensemble in a two-dimensional material

This paper presents a comprehensive experimental framework for quantum sensing using a spin ensemble in hexagonal boron nitride, achieving a record coherence time of 80 μs and sub-microtesla magnetic sensitivity at a 10 nm distance, thereby establishing a foundation for next-generation, atomically thin quantum sensors with ultrahigh sensitivity and tunable noise selectivity.

The Gist

Imagine you have a tiny, super-sensitive microphone that can hear the faintest whispers in a crowded room. In the world of quantum physics, scientists use "spin defects" (tiny imperfections in a crystal) as these microphones to measure magnetic and electric fields. Usually, these microphones are made of diamonds. But diamonds have a problem: if you try to put them very close to the thing you want to measure (like a tiny virus or a single molecule), the surface of the diamond gets "noisy" and the microphone stops working well.

This paper introduces a new, ultra-thin microphone made from a material called hexagonal boron nitride (hBN). Think of hBN as a sheet of paper so thin it's only a few atoms thick. Because it's so thin, you can place it right up against your target without the "surface noise" ruining the signal.

Here is a breakdown of what the scientists did, using simple analogies:

1. The "Central Spin" and its Neighbors

Inside this thin sheet of paper, there are tiny "defects" (missing atoms) that act as the sensor. Let's call the sensor the Central Spin.

• The Problem: The Central Spin isn't alone. It's surrounded by neighbors (other atoms with their own tiny magnetic spins). These neighbors are constantly chattering, which makes it hard for the Central Spin to hear the outside world.
• The Solution: The team didn't just ignore the neighbors; they learned to understand them perfectly. They mapped out exactly how the Central Spin talks to its three closest neighbors. It's like learning the exact dialect and rhythm of a specific group of people so you can tune out their chatter and focus on a specific conversation.

2. The "Switchable Radio"

One of the coolest things they discovered is that they can change what this sensor listens to just by turning a knob (a magnetic field).

• Magnetic Mode: When they point the magnetic field one way, the sensor becomes a radio tuned to magnetic noise. It ignores electric signals and only listens to magnetic ones.
• Electric Mode: When they point the field a different way (flat against the sheet), the sensor becomes a radio tuned to electric noise. It ignores magnetic signals and listens only to electric ones.
• Why it matters: This is like having a single radio that can instantly switch between FM and AM just by rotating the antenna, allowing the scientists to study different types of "noise" in the environment without changing the hardware.

3. The "Noise Map"

To make the sensor work perfectly, they had to figure out exactly what kind of noise was in the room.

• They used a special technique called dynamical decoupling. Imagine trying to hear a whisper in a storm. If you clap your hands in a specific rhythm, you can cancel out the wind noise and hear the whisper.
• By clapping (sending microwave pulses) in a very precise pattern, they filtered out the background noise and reconstructed a "map" of the noise in the material. They found that the noise followed a predictable pattern, which helps them understand how to make the sensor even better in the future.

4. The Results: A Record-Breaking Listen

• Long Memory: The sensor was able to "remember" its state for 80 microseconds. In the world of these tiny sensors, this is a very long time (like holding your breath for a long time underwater). This is a record for this type of material.
• Super Sensitivity: Because they could listen so clearly and for so long, they could detect magnetic fields that are incredibly weak (sub-microtesla) from a distance of just 10 nanometers (about the width of a large virus).
• Comparison: Their sensor is now just as good as the best diamond sensors, but because it's a thin sheet, it can get much closer to the target without losing its hearing.

Summary

The scientists took a very thin, atomically flat material and turned it into a high-tech sensor. They taught the sensor how to ignore its noisy neighbors, figured out how to switch between listening to magnetic and electric signals, and mapped out the background noise to get the clearest possible signal. This proves that these thin, 2D materials are ready to be the next generation of ultra-sensitive tools for measuring the tiny world around us.

Technical Summary

Technical Summary: Quantum Sensing with a Spin Ensemble in a van der Waals Material

Problem Statement
Solid-state spin defects, particularly nitrogen-vacancy (NV) centers in diamond, have revolutionized nanoscale metrology by offering sub-wavelength spatial resolution and high sensitivity. However, leading platforms face significant performance degradation when placed near surfaces or confined to nanoscale volumes, limiting their utility for proximity-based sensing. This degradation motivates the search for optically addressable spin sensors within atomically thin, two-dimensional (2D) van der Waals materials, which can maintain nanometric spatial resolution while preserving quantum coherence. While promising defects such as boron vacancies ($V^-_B$) in hexagonal boron nitride (hBN) have been identified, a comprehensive framework to fully characterize their effective Hamiltonian, coherent sensing dynamics, and noise environment has been lacking.

Methodology
The authors present a device-agnostic experimental framework to probe a 2D spin ensemble consisting of negatively charged boron vacancies ($V^-_B$) in an isotopically engineered hBN crystal (enriched with $^{15}$N). The system is treated as a central spin coupled to a bath of nuclear spins. Key methodological components include:

• Vector Magnetic Field Control: A tunable external vector magnet is used to rotate the quantization axis of the central spin. By applying fields at various angles ($\theta$), the researchers engineer the sensor's eigenstates to exhibit programmable sensitivity to distinct components of the hyperfine Hamiltonian and different types of external noise (magnetic vs. electric).
• Hamiltonian Learning: The team combines Optically Detected Magnetic Resonance (ODMR) spectroscopy with time-domain spin-echo dynamics. While ODMR provides initial spectral data, the authors utilize short-time spin-echo coherence modulations to extract the full hyperfine tensor components ($A_{\mu\nu}$) of the three nearest-neighbor $^{15}$N nuclear spins. This approach overcomes the resolution limitations of ODMR caused by power broadening and spectral diffusion.
• Switchable Sensing Modes: By aligning the external magnetic field parallel or orthogonal to the crystal's zero-field splitting (ZFS) axis, the researchers switch the sensor between a magnetic-field-sensitive mode (using $|0\rangle \leftrightarrow |-1\rangle$ transitions) and an electric-field-sensitive mode (using $|0\rangle \leftrightarrow |M_+\rangle$ transitions, where magnetic moments vanish to leading order).
• Noise Spectrum Reconstruction: To characterize the environmental noise limiting coherence, the authors employ the XY8 dynamical decoupling sequence. They utilize a finite-width filter function (FW-FF) formalism that explicitly accounts for finite pulse durations and control imperfections, enabling the numerical reconstruction of the noise power spectral density (PSD) from measured coherence decay profiles.

Key Results

• Hyperfine Hamiltonian Mapping: The study successfully reconstructs the full hyperfine interaction Hamiltonian for the $V^-_B$ defect. The extracted parameters show excellent agreement with experimental data and outperform previous literature values and density functional theory (DFT) predictions. Notably, the work reveals an in-plane gyromagnetic ratio ($\gamma_\perp/2\pi \approx 19.6$ GHz/T) that is approximately 30% smaller than the out-of-plane ratio ($\gamma_z/2\pi \approx 28$ GHz/T), indicating significant magnetic anisotropy in 2D hBN.
• Record Coherence Times: Under dynamical decoupling at low temperatures ($\approx 2$ K), the ensemble achieves a record electron-spin coherence time of $T_2 \approx 80$ $\mu$s with 2048 $\pi$ pulses. This is the longest coherence time measured in any van der Waals material to date. At room temperature, $T_2$ is limited by a shorter depolarization time ($T_1 \sim 10$ $\mu$s).
• Noise Characterization: The reconstructed noise PSD follows a power-law scaling of $1/\omega^\alpha$ with $\alpha \approx 0.9$ before rolling off near 10 MHz. This spectrum is consistent with theoretical models of spin diffusion in the nuclear bath and coupling to fluctuating two-level systems (TLS).
• Sensitivity Benchmarks: The authors estimate an AC magnetic field sensitivity of $\eta_{ac} \approx 138$ nT/$\sqrt{\text{Hz}}$ at 2 K and $\approx 290$ nT/$\sqrt{\text{Hz}}$ at room temperature. These values place the 2D hBN ensemble sensor on par with state-of-the-art NV center ensembles in diamond, despite the latter typically requiring complex surface treatments.

Significance and Claims
The paper claims to lay the foundation for next-generation quantum sensors based on atomically thin 2D van der Waals materials. By demonstrating a comprehensive framework for learning the Hamiltonian and noise environment, the work validates the potential of 2D spin defects to offer ultrahigh sensitivity and tunable noise selectivity. The ability to switch between magnetic and electric noise sensing modes via external field orientation highlights the versatility of these systems. The authors assert that these results, combined with the broad opportunities for defect engineering in 2D hosts, open pathways not only for advanced quantum sensing but also for scalable quantum simulation and networking, heralding a new era of integrated solid-state quantum technologies. The work emphasizes that the observed performance rivals diamond-based platforms while offering the distinct advantage of atomic thinness for proximity sensing.