⚛️ quantum physics

Direct Observation of Dipolar-Driven Anisotropic Quantum Projection Noise in a Solid-State Spin Ensemble

This study demonstrates quantum-projection-noise-resolved readout of a strongly-interacting 2D ensemble of nitrogen-vacancy centers in diamond, enabling the direct observation of how intrinsic dipolar interactions transform isotropic quantum noise into an anisotropic profile and paving the way for entanglement-enhanced solid-state sensing.

Original authors: Tasuku Ono, Weijie Wu, Haopu Yang, Lillian B. Hughes Wyatt, Benjamin Brenner, Che Liu, Collin Fan, Chris R. Laumann, Jonathan N. Hallén, Emily J. Davis, Ania C. Bleszynski Jayich, Norman Y. Yao

Published 2026-09-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Tasuku Ono, Weijie Wu, Haopu Yang, Lillian B. Hughes Wyatt, Benjamin Brenner, Che Liu, Collin Fan, Chris R. Laumann, Jonathan N. Hallén, Emily J. Davis, Ania C. Bleszynski Jayich, Norman Y. Yao

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the quest to build better sensors, scientists often look to the quantum world, where particles behave in ways that defy everyday intuition. One of the most promising tools for this work is a tiny defect found inside diamonds, known as a nitrogen-vacancy center. Imagine a diamond crystal where a carbon atom is missing and has been replaced by a nitrogen atom next to an empty space. This specific flaw acts like a microscopic magnet that can be read and controlled with light. Because these defects can hold their quantum state for a long time and are easy to manipulate with lasers, they have become a standard workhorse for measuring magnetic fields and temperatures at the scale of individual cells or materials.

To make these sensors even more powerful, researchers have long tried to use large groups, or ensembles, of these diamond defects at once. The logic is simple: if one sensor is good, a thousand should be a thousand times better. However, there is a catch. When you pack these sensors close together, they begin to talk to each other through magnetic forces, which usually scrambles the information they are trying to measure. Furthermore, reading the state of these sensors usually involves counting photons, or particles of light, which introduces a fundamental fuzziness known as noise. For a long time, it seemed impossible to read a dense group of these sensors without this noise drowning out the delicate quantum signals, or without the sensors' own interactions ruining the measurement.

A team of researchers has now successfully navigated this difficult territory. They demonstrated a way to read a dense, two-dimensional layer of these diamond defects with such precision that they could see the fundamental quantum noise of the group itself. More importantly, they watched how the sensors' natural magnetic interactions, which were previously seen as a problem, actually reshaped this noise in a predictable way. By using a clever method that involves reading the same sensor many times in rapid succession, they improved their ability to distinguish the signal from the noise by nearly ten times compared to standard methods. This allowed them to observe a specific quantum effect where the uncertainty of the group's state gets stretched and squeezed into an oval shape, rather than remaining a perfect circle.

The experiment took place in a laboratory where a thin sheet of diamond, only about seven nanometers thick, was prepared with a high concentration of these defects. The researchers focused on four different spots on this sheet, each containing a different number of defects, ranging from roughly 40 to nearly 230 in the tiny area they were observing. To read the state of these defects, they used a laser to excite the electrons, but they faced a challenge: the light used to read the state also tends to reset the system, limiting how many times they could check it. To get around this, they used the nitrogen atom's own nucleus as a memory bank. They would transfer the information from the electron to the nucleus, read the electron, and then transfer the information back, repeating this cycle over and over.

To make this repetitive reading work effectively, the team placed the diamond in a magnetic field of about 0.3 Tesla. This field was crucial because it slowed down the natural decay of the nuclear memory, allowing them to repeat the reading process more than 100 times in some cases. By averaging the results of these many readings, they could filter out the random flickering of the light itself and isolate the true quantum noise of the spin group. They confirmed that this noise behaved exactly as quantum theory predicts for a group of particles, oscillating in a specific pattern as they changed the orientation of the spins.

Once they could see this noise clearly, they turned their attention to what happened when the defects were allowed to interact with one another. They prepared the group in a specific state and then let it evolve naturally. Because the defects are packed so closely together in a flat layer, they interact strongly with each other. The researchers observed that this interaction caused the circular cloud of uncertainty representing the group's state to shear into an ellipse. In simpler terms, the noise became stronger in one direction and weaker in another, creating an anisotropic, or direction-dependent, profile. This is a direct signature of the spins twisting around each other.

To prove that this twisting was caused by the magnetic interactions between the defects and not by some flaw in their measurement equipment, they ran a control experiment. They applied a different sequence of pulses designed to cancel out the magnetic interactions between the defects while still protecting them from outside noise. When they did this, the anisotropy disappeared, and the noise remained perfectly circular. This confirmed that the distortion they saw was indeed driven by the defects talking to each other. They also compared spots with different densities and found that the denser the group, the faster this twisting effect occurred, which matched their theoretical expectations.

This work is significant because it shows that the very interactions that usually ruin sensitive measurements can be harnessed to reshape quantum noise. The ability to see this noise directly and to watch it evolve in real time opens the door to creating sensors that are more precise than the standard limits of physics would normally allow. While the researchers did not yet create a fully squeezed state that beats the standard quantum limit, they have provided the first direct observation of the mechanism that leads there. They have shown that by combining a dense group of sensors with a high-fidelity reading method, it is possible to move beyond simple noise reduction and into a regime where the collective behavior of the group can be engineered for better sensing.

Technical Summary: Direct Observation of Dipolar-Driven Anisotropic Quantum Projection Noise in a Solid-State Spin Ensemble

Problem Statement
The nitrogen-vacancy (NV) center in diamond is a leading platform for quantum sensing, offering robust room-temperature spin coherence and efficient optical control. However, achieving ultimate sensitivity requires optimizing two competing factors: the number of sensing spins ( $N$ ) and the spin readout noise ( $\sigma_R$ ). While increasing $N$ in an ensemble improves sensitivity via the standard quantum limit ( $\sim 1/\sqrt{N}$ ), high NV densities introduce strong intrinsic dipolar interactions that typically broaden linewidths and obscure external signals. Furthermore, conventional optical readout is dominated by photon shot noise, resulting in $\sigma_R \gg 1$ , which prevents the observation of quantum projection noise (QPN).

The central challenge addressed is the experimental inaccessibility of the regime where readout is limited by QPN ( $\sigma_R \approx 1$ ) in a dense, strongly interacting ensemble. Without reaching this regime, the beneficial effects of interactions—such as spin squeezing and entanglement-enhanced sensing—cannot be harnessed, as the readout noise masks the underlying quantum fluctuations.

Methodology
The authors utilized a solid-state system consisting of a thin, delta-doped layer ( $\lesssim 7$ nm) of NV centers in a (111)-oriented diamond crystal. Four distinct regions (labeled A, B, C, and D) with varying NV densities were created via electron-beam irradiation.

To overcome the photon shot noise limit, the team employed a repetitive readout protocol leveraging the NV center's intrinsic $^{15}$ N nuclear spin as a quantum memory. The protocol involves:

State Swapping: Transferring the electronic spin state to the nuclear spin.
Conditional Readout: Using a hyperfine-selective $\pi$ -pulse to implement a CNOT gate, mapping the nuclear state back to the electron for fluorescence detection.
Repetition: Repeating this cycle $k$ times (up to $k \gtrsim 100$ ) before the nuclear memory relaxes.
Magnetic Field Optimization: Applying a moderate external magnetic field ( $B \approx 0.3$ T) to suppress hyperfine flip-flop dynamics in the excited state, thereby extending the nuclear spin lifetime ( $T_{1n}$ ) and allowing for more repetitions.

The readout signal was processed using a differential scheme (pairing shots with complementary CNOT targets) to cancel long-timescale drifts. The total variance of the photon counts was measured, and the expected photon shot noise contribution was subtracted to isolate the residual variance ( $\sigma^2_{res}$ ), which contains the quantum projection noise and technical noise.

Key Contributions and Results

QPN-Resolved Readout in Dense Ensembles:
The authors demonstrated that repetitive readout improves the readout fidelity by nearly an order of magnitude compared to conventional single-shot readout. For the densest regions, the readout noise parameter $\sigma_R$ was reduced from $\sim 18$ (conventional) to $\sim 2.7$ –$6.8$. This reduction allowed for the direct resolution of the quantum projection noise of a coherent spin state, evidenced by the characteristic oscillation of the residual variance at twice the Rabi frequency as the spin state was rotated.
Quantitative Characterization of Spin Numbers:
By fitting the measured residual variance to numerical simulations that accounted for hyperfine linewidths, initialization timescales, and nuclear relaxation, the authors extracted the effective number of spins ( $N$ ) in each region. These values were cross-validated using independent $T_2$ coherence time measurements under XY-8 dynamical decoupling, showing consistency between the two methods.
Direct Observation of Dipolar-Driven Anisotropy:
The primary novel result is the observation of the intrinsic dipolar interactions reshaping the QPN. Starting with a coherent spin state polarized along the $x$ -axis, the ensemble was allowed to evolve under native dipolar interactions while extrinsic noise was suppressed via XY-8 dynamical decoupling.
- Anisotropic Shearing: The residual noise was measured along axes rotated by an angle $\phi$ in the $yz$-plane. The data showed that the initially isotropic (circular) uncertainty distribution sheared into an anisotropic (elliptical) profile over time. The noise was suppressed along one quadrature and amplified along the orthogonal one.
- Density Dependence: The rate of anisotropy development was observed to be faster in regions with higher NV density (stronger dipolar coupling), confirming the interaction-driven nature of the effect.
- Control via Decoupling: When the DROID-60 dynamical decoupling sequence was used instead of XY-8, the dipolar interactions were effectively SU(2)-symmetrized. Under this sequence, the anisotropy vanished, and the noise remained isotropic, confirming that the observed shearing originates specifically from the dipolar twisting dynamics rather than readout artifacts.

Significance and Claims
The paper claims to have opened the door to direct measurements of spin squeezing and entanglement-enhanced sensing in the solid state. By achieving QPN-resolved readout in a dense, interacting ensemble, the authors have demonstrated the experimental capability to observe the "twisting dynamics" where interactions reshape quantum noise.

The authors note that while the observed anisotropy is the immediate precursor to spin squeezing, realizing metrologically useful squeezing in a positionally disordered ensemble remains a challenge. However, this work establishes the necessary foundation: the ability to resolve fluctuations and correlations that are typically obscured by photon shot noise. The results suggest that combining repetitive readout with techniques to mitigate positional disorder could enable the direct observation of scalable spin squeezing and the study of non-Gaussian states and nonequilibrium many-body dynamics in solid-state spin ensembles.

More like this